swa-latn-100mb-after-ppt-shuff-dyck-10mb-ckpt500_seed10/checkpoint-20000/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.654509471303364,
  "eval_steps": 1000,
  "global_step": 20000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.838997936248779,
      "epoch": 0.0014136273678258412,
      "grad_norm": 13.1875,
      "learning_rate": 2e-06,
      "loss": 13.9198,
      "mean_token_accuracy": 0.0,
      "num_tokens": 4340.0,
      "step": 5
    },
    {
      "entropy": 4.843051767349243,
      "epoch": 0.0028272547356516823,
      "grad_norm": 14.0625,
      "learning_rate": 4.5e-06,
      "loss": 13.831,
      "mean_token_accuracy": 0.0,
      "num_tokens": 8692.0,
      "step": 10
    },
    {
      "entropy": 4.923465204238892,
      "epoch": 0.004240882103477523,
      "grad_norm": 16.75,
      "learning_rate": 7e-06,
      "loss": 13.4913,
      "mean_token_accuracy": 0.0,
      "num_tokens": 12453.0,
      "step": 15
    },
    {
      "entropy": 5.232606410980225,
      "epoch": 0.005654509471303365,
      "grad_norm": 26.25,
      "learning_rate": 9.5e-06,
      "loss": 12.9423,
      "mean_token_accuracy": 0.0,
      "num_tokens": 16439.0,
      "step": 20
    },
    {
      "entropy": 7.618780517578125,
      "epoch": 0.007068136839129205,
      "grad_norm": 17.75,
      "learning_rate": 1.2e-05,
      "loss": 11.494,
      "mean_token_accuracy": 0.0,
      "num_tokens": 20515.0,
      "step": 25
    },
    {
      "entropy": 10.471516036987305,
      "epoch": 0.008481764206955046,
      "grad_norm": 4.34375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.526,
      "mean_token_accuracy": 0.0,
      "num_tokens": 24336.0,
      "step": 30
    },
    {
      "entropy": 10.700703144073486,
      "epoch": 0.009895391574780888,
      "grad_norm": 3.8125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.3432,
      "mean_token_accuracy": 0.013547292444854975,
      "num_tokens": 28465.0,
      "step": 35
    },
    {
      "entropy": 10.689900112152099,
      "epoch": 0.01130901894260673,
      "grad_norm": 3.578125,
      "learning_rate": 1.95e-05,
      "loss": 9.9111,
      "mean_token_accuracy": 0.046335773542523384,
      "num_tokens": 32217.0,
      "step": 40
    },
    {
      "entropy": 10.446650981903076,
      "epoch": 0.01272264631043257,
      "grad_norm": 3.4375,
      "learning_rate": 2.2e-05,
      "loss": 9.6264,
      "mean_token_accuracy": 0.040803880989551546,
      "num_tokens": 36369.0,
      "step": 45
    },
    {
      "entropy": 10.473936176300048,
      "epoch": 0.01413627367825841,
      "grad_norm": 2.703125,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.3506,
      "mean_token_accuracy": 0.04644691050052643,
      "num_tokens": 40227.0,
      "step": 50
    },
    {
      "entropy": 10.520126819610596,
      "epoch": 0.015549901046084252,
      "grad_norm": 2.21875,
      "learning_rate": 2.7e-05,
      "loss": 9.2423,
      "mean_token_accuracy": 0.060035499185323714,
      "num_tokens": 44266.0,
      "step": 55
    },
    {
      "entropy": 10.462940120697022,
      "epoch": 0.016963528413910092,
      "grad_norm": 2.3125,
      "learning_rate": 2.95e-05,
      "loss": 9.1768,
      "mean_token_accuracy": 0.05174140147864818,
      "num_tokens": 48187.0,
      "step": 60
    },
    {
      "entropy": 10.422643947601319,
      "epoch": 0.018377155781735936,
      "grad_norm": 2.203125,
      "learning_rate": 3.2e-05,
      "loss": 9.0365,
      "mean_token_accuracy": 0.04872686117887497,
      "num_tokens": 52252.0,
      "step": 65
    },
    {
      "entropy": 10.438050460815429,
      "epoch": 0.019790783149561775,
      "grad_norm": 2.40625,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 8.9607,
      "mean_token_accuracy": 0.05605476088821888,
      "num_tokens": 56421.0,
      "step": 70
    },
    {
      "entropy": 10.400554084777832,
      "epoch": 0.021204410517387615,
      "grad_norm": 2.0625,
      "learning_rate": 3.7e-05,
      "loss": 8.878,
      "mean_token_accuracy": 0.06232075840234756,
      "num_tokens": 60341.0,
      "step": 75
    },
    {
      "entropy": 10.250280380249023,
      "epoch": 0.02261803788521346,
      "grad_norm": 2.28125,
      "learning_rate": 3.95e-05,
      "loss": 8.7283,
      "mean_token_accuracy": 0.062391096726059914,
      "num_tokens": 64465.0,
      "step": 80
    },
    {
      "entropy": 10.2134352684021,
      "epoch": 0.0240316652530393,
      "grad_norm": 1.84375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 8.7126,
      "mean_token_accuracy": 0.06613181754946709,
      "num_tokens": 68514.0,
      "step": 85
    },
    {
      "entropy": 10.055916213989258,
      "epoch": 0.02544529262086514,
      "grad_norm": 1.84375,
      "learning_rate": 4.45e-05,
      "loss": 8.5801,
      "mean_token_accuracy": 0.06724648177623749,
      "num_tokens": 72661.0,
      "step": 90
    },
    {
      "entropy": 9.991343212127685,
      "epoch": 0.02685891998869098,
      "grad_norm": 1.8203125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.4773,
      "mean_token_accuracy": 0.0715904712677002,
      "num_tokens": 76698.0,
      "step": 95
    },
    {
      "entropy": 9.897418022155762,
      "epoch": 0.02827254735651682,
      "grad_norm": 1.8203125,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.3786,
      "mean_token_accuracy": 0.06964126601815224,
      "num_tokens": 80732.0,
      "step": 100
    },
    {
      "entropy": 9.81051197052002,
      "epoch": 0.029686174724342665,
      "grad_norm": 1.6875,
      "learning_rate": 5.2e-05,
      "loss": 8.2664,
      "mean_token_accuracy": 0.07291285954415798,
      "num_tokens": 84797.0,
      "step": 105
    },
    {
      "entropy": 9.750697708129882,
      "epoch": 0.031099802092168505,
      "grad_norm": 1.734375,
      "learning_rate": 5.45e-05,
      "loss": 8.1699,
      "mean_token_accuracy": 0.06931682229042054,
      "num_tokens": 88758.0,
      "step": 110
    },
    {
      "entropy": 9.548874855041504,
      "epoch": 0.032513429459994345,
      "grad_norm": 1.3671875,
      "learning_rate": 5.7e-05,
      "loss": 8.0496,
      "mean_token_accuracy": 0.07683553956449032,
      "num_tokens": 92722.0,
      "step": 115
    },
    {
      "entropy": 9.418696975708007,
      "epoch": 0.033927056827820185,
      "grad_norm": 1.515625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 7.8938,
      "mean_token_accuracy": 0.07645067945122719,
      "num_tokens": 96602.0,
      "step": 120
    },
    {
      "entropy": 9.192409229278564,
      "epoch": 0.035340684195646024,
      "grad_norm": 1.40625,
      "learning_rate": 6.2e-05,
      "loss": 7.831,
      "mean_token_accuracy": 0.07514876201748848,
      "num_tokens": 100598.0,
      "step": 125
    },
    {
      "entropy": 9.073713493347167,
      "epoch": 0.03675431156347187,
      "grad_norm": 1.3515625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.734,
      "mean_token_accuracy": 0.07320134155452251,
      "num_tokens": 104460.0,
      "step": 130
    },
    {
      "entropy": 8.958717918395996,
      "epoch": 0.03816793893129771,
      "grad_norm": 1.421875,
      "learning_rate": 6.7e-05,
      "loss": 7.6242,
      "mean_token_accuracy": 0.07055199854075908,
      "num_tokens": 108360.0,
      "step": 135
    },
    {
      "entropy": 8.655115032196045,
      "epoch": 0.03958156629912355,
      "grad_norm": 1.296875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.5081,
      "mean_token_accuracy": 0.07828053757548332,
      "num_tokens": 112176.0,
      "step": 140
    },
    {
      "entropy": 8.63590440750122,
      "epoch": 0.04099519366694939,
      "grad_norm": 1.484375,
      "learning_rate": 7.2e-05,
      "loss": 7.6179,
      "mean_token_accuracy": 0.06859740577638149,
      "num_tokens": 116192.0,
      "step": 145
    },
    {
      "entropy": 8.526045417785644,
      "epoch": 0.04240882103477523,
      "grad_norm": 1.2109375,
      "learning_rate": 7.45e-05,
      "loss": 7.5008,
      "mean_token_accuracy": 0.06685782484710216,
      "num_tokens": 120221.0,
      "step": 150
    },
    {
      "entropy": 8.428415966033935,
      "epoch": 0.04382244840260108,
      "grad_norm": 1.1953125,
      "learning_rate": 7.7e-05,
      "loss": 7.5019,
      "mean_token_accuracy": 0.06606126651167869,
      "num_tokens": 124364.0,
      "step": 155
    },
    {
      "entropy": 8.274092197418213,
      "epoch": 0.04523607577042692,
      "grad_norm": 1.1171875,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.3469,
      "mean_token_accuracy": 0.08038331940770149,
      "num_tokens": 128266.0,
      "step": 160
    },
    {
      "entropy": 8.154208850860595,
      "epoch": 0.04664970313825276,
      "grad_norm": 1.1015625,
      "learning_rate": 8.2e-05,
      "loss": 7.3261,
      "mean_token_accuracy": 0.07359547801315784,
      "num_tokens": 131985.0,
      "step": 165
    },
    {
      "entropy": 8.039358854293823,
      "epoch": 0.0480633305060786,
      "grad_norm": 1.109375,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.2318,
      "mean_token_accuracy": 0.08298368901014327,
      "num_tokens": 135892.0,
      "step": 170
    },
    {
      "entropy": 7.960572910308838,
      "epoch": 0.04947695787390444,
      "grad_norm": 1.21875,
      "learning_rate": 8.7e-05,
      "loss": 7.3743,
      "mean_token_accuracy": 0.07648940533399581,
      "num_tokens": 139937.0,
      "step": 175
    },
    {
      "entropy": 8.018472337722779,
      "epoch": 0.05089058524173028,
      "grad_norm": 1.1875,
      "learning_rate": 8.95e-05,
      "loss": 7.1981,
      "mean_token_accuracy": 0.08083997145295144,
      "num_tokens": 143659.0,
      "step": 180
    },
    {
      "entropy": 7.916880559921265,
      "epoch": 0.052304212609556124,
      "grad_norm": 1.2734375,
      "learning_rate": 9.2e-05,
      "loss": 7.3147,
      "mean_token_accuracy": 0.08173628598451614,
      "num_tokens": 147394.0,
      "step": 185
    },
    {
      "entropy": 7.840789651870727,
      "epoch": 0.05371783997738196,
      "grad_norm": 1.265625,
      "learning_rate": 9.45e-05,
      "loss": 7.1724,
      "mean_token_accuracy": 0.08405760079622268,
      "num_tokens": 151082.0,
      "step": 190
    },
    {
      "entropy": 7.727948999404907,
      "epoch": 0.0551314673452078,
      "grad_norm": 1.34375,
      "learning_rate": 9.7e-05,
      "loss": 7.1786,
      "mean_token_accuracy": 0.08063250184059143,
      "num_tokens": 155099.0,
      "step": 195
    },
    {
      "entropy": 7.822998285293579,
      "epoch": 0.05654509471303364,
      "grad_norm": 1.140625,
      "learning_rate": 9.95e-05,
      "loss": 7.3548,
      "mean_token_accuracy": 0.07048202715814114,
      "num_tokens": 159297.0,
      "step": 200
    },
    {
      "entropy": 7.713145303726196,
      "epoch": 0.05795872208085948,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000102,
      "loss": 7.2237,
      "mean_token_accuracy": 0.07353744581341744,
      "num_tokens": 163337.0,
      "step": 205
    },
    {
      "entropy": 7.855460119247437,
      "epoch": 0.05937234944868533,
      "grad_norm": 1.125,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.277,
      "mean_token_accuracy": 0.08301040753722191,
      "num_tokens": 167440.0,
      "step": 210
    },
    {
      "entropy": 7.675248670578003,
      "epoch": 0.06078597681651117,
      "grad_norm": 1.234375,
      "learning_rate": 0.000107,
      "loss": 7.1489,
      "mean_token_accuracy": 0.08562113344669342,
      "num_tokens": 171238.0,
      "step": 215
    },
    {
      "entropy": 7.707080364227295,
      "epoch": 0.06219960418433701,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001095,
      "loss": 7.2134,
      "mean_token_accuracy": 0.07360663190484047,
      "num_tokens": 175476.0,
      "step": 220
    },
    {
      "entropy": 7.5691710948944095,
      "epoch": 0.06361323155216285,
      "grad_norm": 1.484375,
      "learning_rate": 0.000112,
      "loss": 7.1273,
      "mean_token_accuracy": 0.08262931331992149,
      "num_tokens": 179398.0,
      "step": 225
    },
    {
      "entropy": 7.5606465339660645,
      "epoch": 0.06502685891998869,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001145,
      "loss": 7.1102,
      "mean_token_accuracy": 0.08324644044041633,
      "num_tokens": 183355.0,
      "step": 230
    },
    {
      "entropy": 7.605392646789551,
      "epoch": 0.06644048628781453,
      "grad_norm": 1.296875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.12,
      "mean_token_accuracy": 0.07809790521860123,
      "num_tokens": 187278.0,
      "step": 235
    },
    {
      "entropy": 7.519303512573242,
      "epoch": 0.06785411365564037,
      "grad_norm": 1.484375,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1402,
      "mean_token_accuracy": 0.0761954978108406,
      "num_tokens": 191136.0,
      "step": 240
    },
    {
      "entropy": 7.577075338363647,
      "epoch": 0.06926774102346621,
      "grad_norm": 1.78125,
      "learning_rate": 0.000122,
      "loss": 7.2122,
      "mean_token_accuracy": 0.08024812079966068,
      "num_tokens": 195096.0,
      "step": 245
    },
    {
      "entropy": 7.606031370162964,
      "epoch": 0.07068136839129205,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0001245,
      "loss": 7.2001,
      "mean_token_accuracy": 0.07799483351409435,
      "num_tokens": 199483.0,
      "step": 250
    },
    {
      "entropy": 7.643957281112671,
      "epoch": 0.0720949957591179,
      "grad_norm": 1.6328125,
      "learning_rate": 0.000127,
      "loss": 7.1263,
      "mean_token_accuracy": 0.08172452189028263,
      "num_tokens": 203515.0,
      "step": 255
    },
    {
      "entropy": 7.497114038467407,
      "epoch": 0.07350862312694374,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0001295,
      "loss": 7.1289,
      "mean_token_accuracy": 0.07867333814501762,
      "num_tokens": 207451.0,
      "step": 260
    },
    {
      "entropy": 7.548718070983886,
      "epoch": 0.07492225049476958,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000132,
      "loss": 7.0456,
      "mean_token_accuracy": 0.08255640640854836,
      "num_tokens": 211507.0,
      "step": 265
    },
    {
      "entropy": 7.439682149887085,
      "epoch": 0.07633587786259542,
      "grad_norm": 1.3125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1149,
      "mean_token_accuracy": 0.08738862499594688,
      "num_tokens": 215236.0,
      "step": 270
    },
    {
      "entropy": 7.497509336471557,
      "epoch": 0.07774950523042126,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.0841,
      "mean_token_accuracy": 0.08626129925251007,
      "num_tokens": 218934.0,
      "step": 275
    },
    {
      "entropy": 7.530088710784912,
      "epoch": 0.0791631325982471,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0001395,
      "loss": 7.031,
      "mean_token_accuracy": 0.08123192861676216,
      "num_tokens": 223185.0,
      "step": 280
    },
    {
      "entropy": 7.557523965835571,
      "epoch": 0.08057675996607294,
      "grad_norm": 1.96875,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.1551,
      "mean_token_accuracy": 0.07593904286623002,
      "num_tokens": 227044.0,
      "step": 285
    },
    {
      "entropy": 7.449309587478638,
      "epoch": 0.08199038733389878,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001445,
      "loss": 7.0628,
      "mean_token_accuracy": 0.09167415499687195,
      "num_tokens": 231359.0,
      "step": 290
    },
    {
      "entropy": 7.531680345535278,
      "epoch": 0.08340401470172462,
      "grad_norm": 1.3125,
      "learning_rate": 0.000147,
      "loss": 7.0471,
      "mean_token_accuracy": 0.08983734510838985,
      "num_tokens": 235224.0,
      "step": 295
    },
    {
      "entropy": 7.344621992111206,
      "epoch": 0.08481764206955046,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0001495,
      "loss": 6.905,
      "mean_token_accuracy": 0.09648748189210891,
      "num_tokens": 239119.0,
      "step": 300
    },
    {
      "entropy": 7.288154172897339,
      "epoch": 0.0862312694373763,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000152,
      "loss": 7.0089,
      "mean_token_accuracy": 0.0949060007929802,
      "num_tokens": 243275.0,
      "step": 305
    },
    {
      "entropy": 7.433500814437866,
      "epoch": 0.08764489680520215,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00015450000000000001,
      "loss": 7.0206,
      "mean_token_accuracy": 0.09643495455384254,
      "num_tokens": 247415.0,
      "step": 310
    },
    {
      "entropy": 7.416780138015747,
      "epoch": 0.089058524173028,
      "grad_norm": 1.453125,
      "learning_rate": 0.000157,
      "loss": 7.0159,
      "mean_token_accuracy": 0.08751917779445648,
      "num_tokens": 251575.0,
      "step": 315
    },
    {
      "entropy": 7.372589826583862,
      "epoch": 0.09047215154085383,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0001595,
      "loss": 6.9997,
      "mean_token_accuracy": 0.08504554927349091,
      "num_tokens": 255686.0,
      "step": 320
    },
    {
      "entropy": 7.3249767303466795,
      "epoch": 0.09188577890867967,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000162,
      "loss": 6.9852,
      "mean_token_accuracy": 0.09507602602243423,
      "num_tokens": 259850.0,
      "step": 325
    },
    {
      "entropy": 7.423571062088013,
      "epoch": 0.09329940627650551,
      "grad_norm": 1.421875,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.0103,
      "mean_token_accuracy": 0.09533059746026992,
      "num_tokens": 264043.0,
      "step": 330
    },
    {
      "entropy": 7.514725065231323,
      "epoch": 0.09471303364433135,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00016700000000000002,
      "loss": 7.0725,
      "mean_token_accuracy": 0.08970947004854679,
      "num_tokens": 268296.0,
      "step": 335
    },
    {
      "entropy": 7.174638795852661,
      "epoch": 0.0961266610121572,
      "grad_norm": 2.28125,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.8961,
      "mean_token_accuracy": 0.09778379574418068,
      "num_tokens": 272062.0,
      "step": 340
    },
    {
      "entropy": 7.25974817276001,
      "epoch": 0.09754028837998303,
      "grad_norm": 1.28125,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8514,
      "mean_token_accuracy": 0.10595791339874268,
      "num_tokens": 275936.0,
      "step": 345
    },
    {
      "entropy": 7.272134685516358,
      "epoch": 0.09895391574780887,
      "grad_norm": 1.640625,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.8802,
      "mean_token_accuracy": 0.10179490596055984,
      "num_tokens": 279526.0,
      "step": 350
    },
    {
      "entropy": 7.339320135116577,
      "epoch": 0.10036754311563471,
      "grad_norm": 1.34375,
      "learning_rate": 0.000177,
      "loss": 6.9552,
      "mean_token_accuracy": 0.0998102031648159,
      "num_tokens": 283577.0,
      "step": 355
    },
    {
      "entropy": 7.324524545669556,
      "epoch": 0.10178117048346055,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0001795,
      "loss": 6.7984,
      "mean_token_accuracy": 0.101173298060894,
      "num_tokens": 287377.0,
      "step": 360
    },
    {
      "entropy": 7.140333604812622,
      "epoch": 0.10319479785128641,
      "grad_norm": 1.28125,
      "learning_rate": 0.000182,
      "loss": 6.973,
      "mean_token_accuracy": 0.09467578828334808,
      "num_tokens": 291582.0,
      "step": 365
    },
    {
      "entropy": 7.392356729507446,
      "epoch": 0.10460842521911225,
      "grad_norm": 1.328125,
      "learning_rate": 0.0001845,
      "loss": 6.8489,
      "mean_token_accuracy": 0.0993436373770237,
      "num_tokens": 295608.0,
      "step": 370
    },
    {
      "entropy": 7.154901552200317,
      "epoch": 0.10602205258693809,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000187,
      "loss": 6.872,
      "mean_token_accuracy": 0.1050638273358345,
      "num_tokens": 299660.0,
      "step": 375
    },
    {
      "entropy": 7.05282769203186,
      "epoch": 0.10743567995476393,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0001895,
      "loss": 6.7846,
      "mean_token_accuracy": 0.11004708558321,
      "num_tokens": 303412.0,
      "step": 380
    },
    {
      "entropy": 7.143272113800049,
      "epoch": 0.10884930732258977,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000192,
      "loss": 6.8498,
      "mean_token_accuracy": 0.10617732927203179,
      "num_tokens": 307352.0,
      "step": 385
    },
    {
      "entropy": 7.2218762874603275,
      "epoch": 0.1102629346904156,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0001945,
      "loss": 6.9129,
      "mean_token_accuracy": 0.10999525636434555,
      "num_tokens": 311418.0,
      "step": 390
    },
    {
      "entropy": 7.1364624977111815,
      "epoch": 0.11167656205824145,
      "grad_norm": 1.578125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7597,
      "mean_token_accuracy": 0.11771088987588882,
      "num_tokens": 315150.0,
      "step": 395
    },
    {
      "entropy": 7.207963132858277,
      "epoch": 0.11309018942606729,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.9407,
      "mean_token_accuracy": 0.10095959007740021,
      "num_tokens": 318898.0,
      "step": 400
    },
    {
      "entropy": 7.126479911804199,
      "epoch": 0.11450381679389313,
      "grad_norm": 1.515625,
      "learning_rate": 0.000202,
      "loss": 6.7726,
      "mean_token_accuracy": 0.11357507780194283,
      "num_tokens": 322495.0,
      "step": 405
    },
    {
      "entropy": 7.224441242218018,
      "epoch": 0.11591744416171897,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8993,
      "mean_token_accuracy": 0.11550437733530998,
      "num_tokens": 326473.0,
      "step": 410
    },
    {
      "entropy": 7.140273571014404,
      "epoch": 0.1173310715295448,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000207,
      "loss": 6.8522,
      "mean_token_accuracy": 0.10760911256074905,
      "num_tokens": 330393.0,
      "step": 415
    },
    {
      "entropy": 7.134380578994751,
      "epoch": 0.11874469889737066,
      "grad_norm": 1.359375,
      "learning_rate": 0.0002095,
      "loss": 6.7692,
      "mean_token_accuracy": 0.11172740682959556,
      "num_tokens": 334449.0,
      "step": 420
    },
    {
      "entropy": 7.2949425220489506,
      "epoch": 0.1201583262651965,
      "grad_norm": 1.4375,
      "learning_rate": 0.000212,
      "loss": 6.9219,
      "mean_token_accuracy": 0.10026353597640991,
      "num_tokens": 338740.0,
      "step": 425
    },
    {
      "entropy": 7.066179800033569,
      "epoch": 0.12157195363302234,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002145,
      "loss": 6.7336,
      "mean_token_accuracy": 0.11876392140984535,
      "num_tokens": 342429.0,
      "step": 430
    },
    {
      "entropy": 7.117091083526612,
      "epoch": 0.12298558100084818,
      "grad_norm": 1.421875,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7741,
      "mean_token_accuracy": 0.12331236079335213,
      "num_tokens": 346218.0,
      "step": 435
    },
    {
      "entropy": 7.182254219055176,
      "epoch": 0.12439920836867402,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002195,
      "loss": 6.7754,
      "mean_token_accuracy": 0.11808006688952447,
      "num_tokens": 350310.0,
      "step": 440
    },
    {
      "entropy": 7.058240985870361,
      "epoch": 0.12581283573649985,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000222,
      "loss": 6.7434,
      "mean_token_accuracy": 0.12305478826165199,
      "num_tokens": 353952.0,
      "step": 445
    },
    {
      "entropy": 7.198861885070801,
      "epoch": 0.1272264631043257,
      "grad_norm": 1.875,
      "learning_rate": 0.0002245,
      "loss": 6.9148,
      "mean_token_accuracy": 0.11514305397868156,
      "num_tokens": 357667.0,
      "step": 450
    },
    {
      "entropy": 7.127170419692993,
      "epoch": 0.12864009047215155,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.7165,
      "mean_token_accuracy": 0.12080714330077172,
      "num_tokens": 361309.0,
      "step": 455
    },
    {
      "entropy": 6.963030576705933,
      "epoch": 0.13005371783997738,
      "grad_norm": 1.40625,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7651,
      "mean_token_accuracy": 0.13394678458571435,
      "num_tokens": 365099.0,
      "step": 460
    },
    {
      "entropy": 7.293617916107178,
      "epoch": 0.13146734520780323,
      "grad_norm": 1.53125,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.8263,
      "mean_token_accuracy": 0.11883056983351707,
      "num_tokens": 369141.0,
      "step": 465
    },
    {
      "entropy": 7.0113513469696045,
      "epoch": 0.13288097257562906,
      "grad_norm": 1.46875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.7673,
      "mean_token_accuracy": 0.11753190383315086,
      "num_tokens": 373128.0,
      "step": 470
    },
    {
      "entropy": 7.056188488006592,
      "epoch": 0.1342945999434549,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000237,
      "loss": 6.7551,
      "mean_token_accuracy": 0.12295122221112251,
      "num_tokens": 377103.0,
      "step": 475
    },
    {
      "entropy": 7.010499048233032,
      "epoch": 0.13570822731128074,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002395,
      "loss": 6.8054,
      "mean_token_accuracy": 0.11294787377119064,
      "num_tokens": 381573.0,
      "step": 480
    },
    {
      "entropy": 7.241208553314209,
      "epoch": 0.1371218546791066,
      "grad_norm": 1.328125,
      "learning_rate": 0.000242,
      "loss": 6.8248,
      "mean_token_accuracy": 0.12627912536263466,
      "num_tokens": 385759.0,
      "step": 485
    },
    {
      "entropy": 6.94762954711914,
      "epoch": 0.13853548204693242,
      "grad_norm": 1.421875,
      "learning_rate": 0.0002445,
      "loss": 6.7048,
      "mean_token_accuracy": 0.12667714580893516,
      "num_tokens": 389700.0,
      "step": 490
    },
    {
      "entropy": 7.064765691757202,
      "epoch": 0.13994910941475827,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000247,
      "loss": 6.6205,
      "mean_token_accuracy": 0.12440077215433121,
      "num_tokens": 393279.0,
      "step": 495
    },
    {
      "entropy": 7.031151533126831,
      "epoch": 0.1413627367825841,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002495,
      "loss": 6.725,
      "mean_token_accuracy": 0.12710191905498505,
      "num_tokens": 397379.0,
      "step": 500
    },
    {
      "entropy": 6.966974687576294,
      "epoch": 0.14277636415040995,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000252,
      "loss": 6.7762,
      "mean_token_accuracy": 0.12484551221132278,
      "num_tokens": 401349.0,
      "step": 505
    },
    {
      "entropy": 7.063830709457397,
      "epoch": 0.1441899915182358,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002545,
      "loss": 6.6911,
      "mean_token_accuracy": 0.12564558237791063,
      "num_tokens": 405361.0,
      "step": 510
    },
    {
      "entropy": 6.972732162475586,
      "epoch": 0.14560361888606163,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000257,
      "loss": 6.726,
      "mean_token_accuracy": 0.12524131089448928,
      "num_tokens": 409614.0,
      "step": 515
    },
    {
      "entropy": 6.847900104522705,
      "epoch": 0.14701724625388748,
      "grad_norm": 1.4375,
      "learning_rate": 0.0002595,
      "loss": 6.6623,
      "mean_token_accuracy": 0.13759828507900237,
      "num_tokens": 413529.0,
      "step": 520
    },
    {
      "entropy": 6.994072675704956,
      "epoch": 0.1484308736217133,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000262,
      "loss": 6.6178,
      "mean_token_accuracy": 0.12413196116685868,
      "num_tokens": 417365.0,
      "step": 525
    },
    {
      "entropy": 6.847810077667236,
      "epoch": 0.14984450098953916,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.6269,
      "mean_token_accuracy": 0.1389990709722042,
      "num_tokens": 421232.0,
      "step": 530
    },
    {
      "entropy": 7.125303459167481,
      "epoch": 0.151258128357365,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.7183,
      "mean_token_accuracy": 0.12456061840057372,
      "num_tokens": 425503.0,
      "step": 535
    },
    {
      "entropy": 6.852126216888427,
      "epoch": 0.15267175572519084,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7542,
      "mean_token_accuracy": 0.12805211022496224,
      "num_tokens": 429770.0,
      "step": 540
    },
    {
      "entropy": 7.0629345893859865,
      "epoch": 0.15408538309301667,
      "grad_norm": 1.625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6387,
      "mean_token_accuracy": 0.13042709827423096,
      "num_tokens": 433657.0,
      "step": 545
    },
    {
      "entropy": 6.941704559326172,
      "epoch": 0.15549901046084252,
      "grad_norm": 1.453125,
      "learning_rate": 0.0002745,
      "loss": 6.7511,
      "mean_token_accuracy": 0.1230709470808506,
      "num_tokens": 437732.0,
      "step": 550
    },
    {
      "entropy": 6.886470556259155,
      "epoch": 0.15691263782866835,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000277,
      "loss": 6.4963,
      "mean_token_accuracy": 0.13797944858670236,
      "num_tokens": 441866.0,
      "step": 555
    },
    {
      "entropy": 6.786446952819825,
      "epoch": 0.1583262651964942,
      "grad_norm": 1.46875,
      "learning_rate": 0.0002795,
      "loss": 6.5783,
      "mean_token_accuracy": 0.1337943233549595,
      "num_tokens": 445437.0,
      "step": 560
    },
    {
      "entropy": 6.786550521850586,
      "epoch": 0.15973989256432006,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.564,
      "mean_token_accuracy": 0.13779796212911605,
      "num_tokens": 449602.0,
      "step": 565
    },
    {
      "entropy": 6.764790821075439,
      "epoch": 0.16115351993214588,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002845,
      "loss": 6.4694,
      "mean_token_accuracy": 0.13991471529006957,
      "num_tokens": 453140.0,
      "step": 570
    },
    {
      "entropy": 6.97475233078003,
      "epoch": 0.16256714729997174,
      "grad_norm": 1.53125,
      "learning_rate": 0.000287,
      "loss": 6.7096,
      "mean_token_accuracy": 0.12624734565615653,
      "num_tokens": 457210.0,
      "step": 575
    },
    {
      "entropy": 7.009247064590454,
      "epoch": 0.16398077466779756,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0002895,
      "loss": 6.6826,
      "mean_token_accuracy": 0.12368027344346047,
      "num_tokens": 461534.0,
      "step": 580
    },
    {
      "entropy": 6.9301024913787845,
      "epoch": 0.16539440203562342,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000292,
      "loss": 6.4802,
      "mean_token_accuracy": 0.13896113261580467,
      "num_tokens": 465055.0,
      "step": 585
    },
    {
      "entropy": 6.829527807235718,
      "epoch": 0.16680802940344924,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002945,
      "loss": 6.5188,
      "mean_token_accuracy": 0.13359609320759774,
      "num_tokens": 469133.0,
      "step": 590
    },
    {
      "entropy": 6.923969316482544,
      "epoch": 0.1682216567712751,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000297,
      "loss": 6.6646,
      "mean_token_accuracy": 0.1328178010880947,
      "num_tokens": 473078.0,
      "step": 595
    },
    {
      "entropy": 6.906017303466797,
      "epoch": 0.16963528413910092,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002995,
      "loss": 6.5999,
      "mean_token_accuracy": 0.13147619739174843,
      "num_tokens": 476988.0,
      "step": 600
    },
    {
      "entropy": 6.845853567123413,
      "epoch": 0.17104891150692678,
      "grad_norm": 1.40625,
      "learning_rate": 0.000302,
      "loss": 6.4638,
      "mean_token_accuracy": 0.13877335488796233,
      "num_tokens": 481170.0,
      "step": 605
    },
    {
      "entropy": 6.808208322525024,
      "epoch": 0.1724625388747526,
      "grad_norm": 1.75,
      "learning_rate": 0.0003045,
      "loss": 6.4607,
      "mean_token_accuracy": 0.12855389714241028,
      "num_tokens": 484853.0,
      "step": 610
    },
    {
      "entropy": 6.694374704360962,
      "epoch": 0.17387616624257846,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000307,
      "loss": 6.5042,
      "mean_token_accuracy": 0.14368257224559783,
      "num_tokens": 488550.0,
      "step": 615
    },
    {
      "entropy": 6.714710140228272,
      "epoch": 0.1752897936104043,
      "grad_norm": 1.5,
      "learning_rate": 0.0003095,
      "loss": 6.5511,
      "mean_token_accuracy": 0.1393193356692791,
      "num_tokens": 492621.0,
      "step": 620
    },
    {
      "entropy": 6.823784923553466,
      "epoch": 0.17670342097823014,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000312,
      "loss": 6.4969,
      "mean_token_accuracy": 0.12763671800494195,
      "num_tokens": 496627.0,
      "step": 625
    },
    {
      "entropy": 6.75352110862732,
      "epoch": 0.178117048346056,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003145,
      "loss": 6.5534,
      "mean_token_accuracy": 0.13516684994101524,
      "num_tokens": 500794.0,
      "step": 630
    },
    {
      "entropy": 6.771178388595581,
      "epoch": 0.17953067571388182,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000317,
      "loss": 6.5715,
      "mean_token_accuracy": 0.1311728797852993,
      "num_tokens": 504672.0,
      "step": 635
    },
    {
      "entropy": 6.785078620910644,
      "epoch": 0.18094430308170767,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0003195,
      "loss": 6.4995,
      "mean_token_accuracy": 0.14259493052959443,
      "num_tokens": 508834.0,
      "step": 640
    },
    {
      "entropy": 6.843728113174438,
      "epoch": 0.1823579304495335,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000322,
      "loss": 6.4686,
      "mean_token_accuracy": 0.13577061742544175,
      "num_tokens": 512722.0,
      "step": 645
    },
    {
      "entropy": 6.774205303192138,
      "epoch": 0.18377155781735935,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.5824,
      "mean_token_accuracy": 0.13288535103201865,
      "num_tokens": 517270.0,
      "step": 650
    },
    {
      "entropy": 6.76011438369751,
      "epoch": 0.18518518518518517,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.5048,
      "mean_token_accuracy": 0.14277215600013732,
      "num_tokens": 521438.0,
      "step": 655
    },
    {
      "entropy": 6.762216997146607,
      "epoch": 0.18659881255301103,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.4822,
      "mean_token_accuracy": 0.1396136611700058,
      "num_tokens": 525890.0,
      "step": 660
    },
    {
      "entropy": 6.725051021575927,
      "epoch": 0.18801243992083685,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4927,
      "mean_token_accuracy": 0.14150392562150954,
      "num_tokens": 529775.0,
      "step": 665
    },
    {
      "entropy": 6.711924600601196,
      "epoch": 0.1894260672886627,
      "grad_norm": 1.515625,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4771,
      "mean_token_accuracy": 0.14005500972270965,
      "num_tokens": 533889.0,
      "step": 670
    },
    {
      "entropy": 6.584859466552734,
      "epoch": 0.19083969465648856,
      "grad_norm": 1.296875,
      "learning_rate": 0.000337,
      "loss": 6.402,
      "mean_token_accuracy": 0.1481585830450058,
      "num_tokens": 537701.0,
      "step": 675
    },
    {
      "entropy": 6.701951599121093,
      "epoch": 0.1922533220243144,
      "grad_norm": 1.3125,
      "learning_rate": 0.0003395,
      "loss": 6.4879,
      "mean_token_accuracy": 0.14182492196559907,
      "num_tokens": 542160.0,
      "step": 680
    },
    {
      "entropy": 6.702091026306152,
      "epoch": 0.19366694939214024,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000342,
      "loss": 6.4397,
      "mean_token_accuracy": 0.14451163560152053,
      "num_tokens": 546317.0,
      "step": 685
    },
    {
      "entropy": 6.670078468322754,
      "epoch": 0.19508057675996607,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.2943,
      "mean_token_accuracy": 0.15481528118252755,
      "num_tokens": 550304.0,
      "step": 690
    },
    {
      "entropy": 6.529588890075684,
      "epoch": 0.19649420412779192,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000347,
      "loss": 6.3855,
      "mean_token_accuracy": 0.1449325367808342,
      "num_tokens": 554276.0,
      "step": 695
    },
    {
      "entropy": 6.679137516021728,
      "epoch": 0.19790783149561775,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003495,
      "loss": 6.4109,
      "mean_token_accuracy": 0.144695745408535,
      "num_tokens": 558044.0,
      "step": 700
    },
    {
      "entropy": 6.46890025138855,
      "epoch": 0.1993214588634436,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000352,
      "loss": 6.398,
      "mean_token_accuracy": 0.15260268077254296,
      "num_tokens": 561667.0,
      "step": 705
    },
    {
      "entropy": 6.7185471534729,
      "epoch": 0.20073508623126943,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003545,
      "loss": 6.421,
      "mean_token_accuracy": 0.1382300853729248,
      "num_tokens": 565738.0,
      "step": 710
    },
    {
      "entropy": 6.691477727890015,
      "epoch": 0.20214871359909528,
      "grad_norm": 1.53125,
      "learning_rate": 0.000357,
      "loss": 6.4849,
      "mean_token_accuracy": 0.1443715550005436,
      "num_tokens": 569629.0,
      "step": 715
    },
    {
      "entropy": 6.537031507492065,
      "epoch": 0.2035623409669211,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0003595,
      "loss": 6.3245,
      "mean_token_accuracy": 0.14985155463218688,
      "num_tokens": 573272.0,
      "step": 720
    },
    {
      "entropy": 6.596857976913452,
      "epoch": 0.20497596833474696,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000362,
      "loss": 6.3822,
      "mean_token_accuracy": 0.14262079298496247,
      "num_tokens": 577224.0,
      "step": 725
    },
    {
      "entropy": 6.525388908386231,
      "epoch": 0.20638959570257281,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003645,
      "loss": 6.3224,
      "mean_token_accuracy": 0.15136756971478463,
      "num_tokens": 580919.0,
      "step": 730
    },
    {
      "entropy": 6.716662502288818,
      "epoch": 0.20780322307039864,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000367,
      "loss": 6.445,
      "mean_token_accuracy": 0.1410931281745434,
      "num_tokens": 584985.0,
      "step": 735
    },
    {
      "entropy": 6.596371221542358,
      "epoch": 0.2092168504382245,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003695,
      "loss": 6.3992,
      "mean_token_accuracy": 0.15255053117871284,
      "num_tokens": 588979.0,
      "step": 740
    },
    {
      "entropy": 6.564611911773682,
      "epoch": 0.21063047780605032,
      "grad_norm": 1.671875,
      "learning_rate": 0.000372,
      "loss": 6.3667,
      "mean_token_accuracy": 0.1505613371729851,
      "num_tokens": 592930.0,
      "step": 745
    },
    {
      "entropy": 6.536220741271973,
      "epoch": 0.21204410517387617,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003745,
      "loss": 6.4103,
      "mean_token_accuracy": 0.1515594244003296,
      "num_tokens": 596832.0,
      "step": 750
    },
    {
      "entropy": 6.828583717346191,
      "epoch": 0.213457732541702,
      "grad_norm": 1.8125,
      "learning_rate": 0.000377,
      "loss": 6.446,
      "mean_token_accuracy": 0.14928378835320472,
      "num_tokens": 600934.0,
      "step": 755
    },
    {
      "entropy": 6.329220914840699,
      "epoch": 0.21487135990952785,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003795,
      "loss": 6.3199,
      "mean_token_accuracy": 0.1495775043964386,
      "num_tokens": 604868.0,
      "step": 760
    },
    {
      "entropy": 6.744077539443969,
      "epoch": 0.21628498727735368,
      "grad_norm": 1.484375,
      "learning_rate": 0.000382,
      "loss": 6.4376,
      "mean_token_accuracy": 0.13565291613340377,
      "num_tokens": 608747.0,
      "step": 765
    },
    {
      "entropy": 6.577572679519653,
      "epoch": 0.21769861464517953,
      "grad_norm": 1.421875,
      "learning_rate": 0.0003845,
      "loss": 6.4215,
      "mean_token_accuracy": 0.1479615420103073,
      "num_tokens": 612627.0,
      "step": 770
    },
    {
      "entropy": 6.517498445510864,
      "epoch": 0.21911224201300536,
      "grad_norm": 1.40625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3368,
      "mean_token_accuracy": 0.1571616619825363,
      "num_tokens": 616455.0,
      "step": 775
    },
    {
      "entropy": 6.5876189231872555,
      "epoch": 0.2205258693808312,
      "grad_norm": 1.390625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.4168,
      "mean_token_accuracy": 0.1469177559018135,
      "num_tokens": 620617.0,
      "step": 780
    },
    {
      "entropy": 6.415558195114135,
      "epoch": 0.22193949674865707,
      "grad_norm": 1.5625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3376,
      "mean_token_accuracy": 0.1490316942334175,
      "num_tokens": 624863.0,
      "step": 785
    },
    {
      "entropy": 6.5751752853393555,
      "epoch": 0.2233531241164829,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3625,
      "mean_token_accuracy": 0.15206135734915732,
      "num_tokens": 629073.0,
      "step": 790
    },
    {
      "entropy": 6.595431232452393,
      "epoch": 0.22476675148430875,
      "grad_norm": 1.421875,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3723,
      "mean_token_accuracy": 0.15247093737125397,
      "num_tokens": 633550.0,
      "step": 795
    },
    {
      "entropy": 6.571492719650268,
      "epoch": 0.22618037885213457,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003995,
      "loss": 6.3611,
      "mean_token_accuracy": 0.15650528520345688,
      "num_tokens": 637459.0,
      "step": 800
    },
    {
      "entropy": 6.416607141494751,
      "epoch": 0.22759400621996043,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000402,
      "loss": 6.3272,
      "mean_token_accuracy": 0.14604957774281502,
      "num_tokens": 641587.0,
      "step": 805
    },
    {
      "entropy": 6.561601161956787,
      "epoch": 0.22900763358778625,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004045,
      "loss": 6.5026,
      "mean_token_accuracy": 0.14142623618245126,
      "num_tokens": 645726.0,
      "step": 810
    },
    {
      "entropy": 6.5540244579315186,
      "epoch": 0.2304212609556121,
      "grad_norm": 1.453125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2808,
      "mean_token_accuracy": 0.15965329706668854,
      "num_tokens": 649688.0,
      "step": 815
    },
    {
      "entropy": 6.466880655288696,
      "epoch": 0.23183488832343793,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004095,
      "loss": 6.3081,
      "mean_token_accuracy": 0.14743991270661355,
      "num_tokens": 653745.0,
      "step": 820
    },
    {
      "entropy": 6.281443119049072,
      "epoch": 0.23324851569126379,
      "grad_norm": 1.4375,
      "learning_rate": 0.000412,
      "loss": 6.2163,
      "mean_token_accuracy": 0.1635961800813675,
      "num_tokens": 657650.0,
      "step": 825
    },
    {
      "entropy": 6.540815114974976,
      "epoch": 0.2346621430590896,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004145,
      "loss": 6.3282,
      "mean_token_accuracy": 0.15691596940159797,
      "num_tokens": 661614.0,
      "step": 830
    },
    {
      "entropy": 6.437383794784546,
      "epoch": 0.23607577042691547,
      "grad_norm": 1.4375,
      "learning_rate": 0.000417,
      "loss": 6.2877,
      "mean_token_accuracy": 0.15950870811939238,
      "num_tokens": 665484.0,
      "step": 835
    },
    {
      "entropy": 6.302047681808472,
      "epoch": 0.23748939779474132,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004195,
      "loss": 6.3023,
      "mean_token_accuracy": 0.1515549197793007,
      "num_tokens": 669718.0,
      "step": 840
    },
    {
      "entropy": 6.494394445419312,
      "epoch": 0.23890302516256715,
      "grad_norm": 1.46875,
      "learning_rate": 0.000422,
      "loss": 6.2621,
      "mean_token_accuracy": 0.1587437778711319,
      "num_tokens": 673913.0,
      "step": 845
    },
    {
      "entropy": 6.465457439422607,
      "epoch": 0.240316652530393,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004245,
      "loss": 6.3427,
      "mean_token_accuracy": 0.14686428755521774,
      "num_tokens": 677911.0,
      "step": 850
    },
    {
      "entropy": 6.367018795013427,
      "epoch": 0.24173027989821882,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000427,
      "loss": 6.1437,
      "mean_token_accuracy": 0.16687410622835158,
      "num_tokens": 681629.0,
      "step": 855
    },
    {
      "entropy": 6.405222368240357,
      "epoch": 0.24314390726604468,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004295,
      "loss": 6.1789,
      "mean_token_accuracy": 0.15915078967809676,
      "num_tokens": 685675.0,
      "step": 860
    },
    {
      "entropy": 6.3251574516296385,
      "epoch": 0.2445575346338705,
      "grad_norm": 1.375,
      "learning_rate": 0.000432,
      "loss": 6.1061,
      "mean_token_accuracy": 0.16585489213466645,
      "num_tokens": 689630.0,
      "step": 865
    },
    {
      "entropy": 6.331725025177002,
      "epoch": 0.24597116200169636,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004345,
      "loss": 6.2086,
      "mean_token_accuracy": 0.15823635309934617,
      "num_tokens": 693580.0,
      "step": 870
    },
    {
      "entropy": 6.292018890380859,
      "epoch": 0.24738478936952218,
      "grad_norm": 1.53125,
      "learning_rate": 0.000437,
      "loss": 6.2824,
      "mean_token_accuracy": 0.15428755432367325,
      "num_tokens": 697531.0,
      "step": 875
    },
    {
      "entropy": 6.245914125442505,
      "epoch": 0.24879841673734804,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004395,
      "loss": 6.2452,
      "mean_token_accuracy": 0.16177606135606765,
      "num_tokens": 701421.0,
      "step": 880
    },
    {
      "entropy": 6.380286502838135,
      "epoch": 0.2502120441051739,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000442,
      "loss": 6.2034,
      "mean_token_accuracy": 0.158734093606472,
      "num_tokens": 705557.0,
      "step": 885
    },
    {
      "entropy": 6.279850482940674,
      "epoch": 0.2516256714729997,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004445,
      "loss": 6.1344,
      "mean_token_accuracy": 0.1670902654528618,
      "num_tokens": 709173.0,
      "step": 890
    },
    {
      "entropy": 6.265487957000732,
      "epoch": 0.25303929884082554,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000447,
      "loss": 6.0852,
      "mean_token_accuracy": 0.1611698791384697,
      "num_tokens": 713144.0,
      "step": 895
    },
    {
      "entropy": 6.158499622344971,
      "epoch": 0.2544529262086514,
      "grad_norm": 1.359375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.3942,
      "mean_token_accuracy": 0.15290624499320984,
      "num_tokens": 717187.0,
      "step": 900
    },
    {
      "entropy": 6.500699520111084,
      "epoch": 0.25586655357647725,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.2656,
      "mean_token_accuracy": 0.14998120591044425,
      "num_tokens": 721311.0,
      "step": 905
    },
    {
      "entropy": 6.283317804336548,
      "epoch": 0.2572801809443031,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.1865,
      "mean_token_accuracy": 0.1690343588590622,
      "num_tokens": 725193.0,
      "step": 910
    },
    {
      "entropy": 6.32297830581665,
      "epoch": 0.2586938083121289,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.1775,
      "mean_token_accuracy": 0.1630359873175621,
      "num_tokens": 729208.0,
      "step": 915
    },
    {
      "entropy": 6.252132749557495,
      "epoch": 0.26010743567995476,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2431,
      "mean_token_accuracy": 0.162384469807148,
      "num_tokens": 733364.0,
      "step": 920
    },
    {
      "entropy": 6.240159702301026,
      "epoch": 0.2615210630477806,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000462,
      "loss": 6.1907,
      "mean_token_accuracy": 0.16134183555841447,
      "num_tokens": 737549.0,
      "step": 925
    },
    {
      "entropy": 6.357678174972534,
      "epoch": 0.26293469041560646,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004645,
      "loss": 6.2012,
      "mean_token_accuracy": 0.16150422096252443,
      "num_tokens": 742009.0,
      "step": 930
    },
    {
      "entropy": 6.26593918800354,
      "epoch": 0.26434831778343226,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000467,
      "loss": 6.3095,
      "mean_token_accuracy": 0.1559640161693096,
      "num_tokens": 746258.0,
      "step": 935
    },
    {
      "entropy": 6.422447776794433,
      "epoch": 0.2657619451512581,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004695,
      "loss": 6.2134,
      "mean_token_accuracy": 0.16972213834524155,
      "num_tokens": 750027.0,
      "step": 940
    },
    {
      "entropy": 6.179395294189453,
      "epoch": 0.26717557251908397,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000472,
      "loss": 6.1788,
      "mean_token_accuracy": 0.15804249197244644,
      "num_tokens": 754030.0,
      "step": 945
    },
    {
      "entropy": 6.166366243362427,
      "epoch": 0.2685891998869098,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004745,
      "loss": 6.0744,
      "mean_token_accuracy": 0.1662062868475914,
      "num_tokens": 757663.0,
      "step": 950
    },
    {
      "entropy": 6.173876714706421,
      "epoch": 0.2700028272547357,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000477,
      "loss": 6.1816,
      "mean_token_accuracy": 0.1622084781527519,
      "num_tokens": 761806.0,
      "step": 955
    },
    {
      "entropy": 6.358310317993164,
      "epoch": 0.2714164546225615,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004795,
      "loss": 6.1687,
      "mean_token_accuracy": 0.1678529866039753,
      "num_tokens": 765900.0,
      "step": 960
    },
    {
      "entropy": 6.13313364982605,
      "epoch": 0.27283008199038733,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000482,
      "loss": 6.1376,
      "mean_token_accuracy": 0.16201107054948807,
      "num_tokens": 769842.0,
      "step": 965
    },
    {
      "entropy": 6.349887704849243,
      "epoch": 0.2742437093582132,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004845,
      "loss": 6.2531,
      "mean_token_accuracy": 0.16530761122703552,
      "num_tokens": 773717.0,
      "step": 970
    },
    {
      "entropy": 6.021614742279053,
      "epoch": 0.27565733672603904,
      "grad_norm": 1.3984375,
      "learning_rate": 0.000487,
      "loss": 5.9991,
      "mean_token_accuracy": 0.17551461309194566,
      "num_tokens": 777720.0,
      "step": 975
    },
    {
      "entropy": 6.14023642539978,
      "epoch": 0.27707096409386484,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004895,
      "loss": 6.027,
      "mean_token_accuracy": 0.16881906613707542,
      "num_tokens": 781482.0,
      "step": 980
    },
    {
      "entropy": 6.2495396614074705,
      "epoch": 0.2784845914616907,
      "grad_norm": 1.671875,
      "learning_rate": 0.000492,
      "loss": 6.1729,
      "mean_token_accuracy": 0.15896565914154054,
      "num_tokens": 785390.0,
      "step": 985
    },
    {
      "entropy": 5.967093658447266,
      "epoch": 0.27989821882951654,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004945,
      "loss": 6.0915,
      "mean_token_accuracy": 0.1750103384256363,
      "num_tokens": 789464.0,
      "step": 990
    },
    {
      "entropy": 6.431793308258056,
      "epoch": 0.2813118461973424,
      "grad_norm": 1.46875,
      "learning_rate": 0.000497,
      "loss": 6.179,
      "mean_token_accuracy": 0.15985391139984131,
      "num_tokens": 793229.0,
      "step": 995
    },
    {
      "entropy": 6.186357164382935,
      "epoch": 0.2827254735651682,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004995,
      "loss": 6.0844,
      "mean_token_accuracy": 0.1675830215215683,
      "num_tokens": 797490.0,
      "step": 1000
    },
    {
      "epoch": 0.2827254735651682,
      "eval_entropy": 6.063302199271112,
      "eval_loss": 6.372450828552246,
      "eval_mean_token_accuracy": 0.16418416006589595,
      "eval_num_tokens": 797490.0,
      "eval_runtime": 6.7832,
      "eval_samples_per_second": 1663.516,
      "eval_steps_per_second": 208.013,
      "step": 1000
    },
    {
      "entropy": 6.069096565246582,
      "epoch": 0.28413910093299405,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004999999849524451,
      "loss": 6.136,
      "mean_token_accuracy": 0.16406980752944947,
      "num_tokens": 801242.0,
      "step": 1005
    },
    {
      "entropy": 6.29629168510437,
      "epoch": 0.2855527283008199,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000499999923821757,
      "loss": 6.2518,
      "mean_token_accuracy": 0.15641706585884094,
      "num_tokens": 805285.0,
      "step": 1010
    },
    {
      "entropy": 6.121434211730957,
      "epoch": 0.28696635566864576,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000499999815667476,
      "loss": 6.0715,
      "mean_token_accuracy": 0.17298818677663802,
      "num_tokens": 809278.0,
      "step": 1015
    },
    {
      "entropy": 6.067972850799561,
      "epoch": 0.2883799830364716,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999996604896251,
      "loss": 6.198,
      "mean_token_accuracy": 0.17273716926574706,
      "num_tokens": 813519.0,
      "step": 1020
    },
    {
      "entropy": 6.159738492965698,
      "epoch": 0.2897936104042974,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004999994582882363,
      "loss": 6.1376,
      "mean_token_accuracy": 0.1695775046944618,
      "num_tokens": 817549.0,
      "step": 1025
    },
    {
      "entropy": 6.177052736282349,
      "epoch": 0.29120723777212326,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999992090633522,
      "loss": 6.0973,
      "mean_token_accuracy": 0.16485699787735938,
      "num_tokens": 821608.0,
      "step": 1030
    },
    {
      "entropy": 5.899837636947632,
      "epoch": 0.2926208651399491,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999989128150248,
      "loss": 6.0539,
      "mean_token_accuracy": 0.17530483603477479,
      "num_tokens": 825795.0,
      "step": 1035
    },
    {
      "entropy": 6.26076979637146,
      "epoch": 0.29403449250777497,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004999985695433159,
      "loss": 6.1328,
      "mean_token_accuracy": 0.1712009161710739,
      "num_tokens": 829693.0,
      "step": 1040
    },
    {
      "entropy": 6.17474045753479,
      "epoch": 0.29544811987560077,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999981792482973,
      "loss": 6.157,
      "mean_token_accuracy": 0.16908271610736847,
      "num_tokens": 833665.0,
      "step": 1045
    },
    {
      "entropy": 6.174765777587891,
      "epoch": 0.2968617472434266,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004999977419300507,
      "loss": 6.0853,
      "mean_token_accuracy": 0.16263003200292586,
      "num_tokens": 837107.0,
      "step": 1050
    },
    {
      "entropy": 6.165372848510742,
      "epoch": 0.2982753746112525,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999972575886673,
      "loss": 6.2114,
      "mean_token_accuracy": 0.16624142229557037,
      "num_tokens": 841182.0,
      "step": 1055
    },
    {
      "entropy": 6.30288724899292,
      "epoch": 0.29968900197907833,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004999967262242483,
      "loss": 6.1395,
      "mean_token_accuracy": 0.16506897360086442,
      "num_tokens": 845266.0,
      "step": 1060
    },
    {
      "entropy": 6.16644515991211,
      "epoch": 0.3011026293469042,
      "grad_norm": 1.453125,
      "learning_rate": 0.000499996147836905,
      "loss": 6.1257,
      "mean_token_accuracy": 0.16231662929058074,
      "num_tokens": 849118.0,
      "step": 1065
    },
    {
      "entropy": 6.304385757446289,
      "epoch": 0.30251625671473,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000499995522426758,
      "loss": 6.2201,
      "mean_token_accuracy": 0.16377525627613068,
      "num_tokens": 853101.0,
      "step": 1070
    },
    {
      "entropy": 5.996787118911743,
      "epoch": 0.30392988408255583,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004999948499939383,
      "loss": 6.0619,
      "mean_token_accuracy": 0.1750151202082634,
      "num_tokens": 856671.0,
      "step": 1075
    },
    {
      "entropy": 6.145863771438599,
      "epoch": 0.3053435114503817,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999941305385863,
      "loss": 6.009,
      "mean_token_accuracy": 0.1848408907651901,
      "num_tokens": 860639.0,
      "step": 1080
    },
    {
      "entropy": 6.0218729972839355,
      "epoch": 0.30675713881820754,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004999933640608521,
      "loss": 6.0525,
      "mean_token_accuracy": 0.16953370422124864,
      "num_tokens": 864270.0,
      "step": 1085
    },
    {
      "entropy": 6.129056644439697,
      "epoch": 0.30817076618603334,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004999925505608963,
      "loss": 6.0399,
      "mean_token_accuracy": 0.16365957260131836,
      "num_tokens": 868069.0,
      "step": 1090
    },
    {
      "entropy": 6.096409034729004,
      "epoch": 0.3095843935538592,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999916900388887,
      "loss": 6.07,
      "mean_token_accuracy": 0.16356224715709686,
      "num_tokens": 871943.0,
      "step": 1095
    },
    {
      "entropy": 6.21712646484375,
      "epoch": 0.31099802092168505,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999907824950093,
      "loss": 6.0956,
      "mean_token_accuracy": 0.17098572999238967,
      "num_tokens": 876168.0,
      "step": 1100
    },
    {
      "entropy": 5.9889732837677006,
      "epoch": 0.3124116482895109,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004999898279294475,
      "loss": 6.0648,
      "mean_token_accuracy": 0.1675517365336418,
      "num_tokens": 880134.0,
      "step": 1105
    },
    {
      "entropy": 6.075795602798462,
      "epoch": 0.3138252756573367,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004999888263424031,
      "loss": 6.0542,
      "mean_token_accuracy": 0.16933156102895736,
      "num_tokens": 883970.0,
      "step": 1110
    },
    {
      "entropy": 6.1346986293792725,
      "epoch": 0.31523890302516255,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004999877777340852,
      "loss": 6.1297,
      "mean_token_accuracy": 0.15770118832588195,
      "num_tokens": 887980.0,
      "step": 1115
    },
    {
      "entropy": 5.962949466705322,
      "epoch": 0.3166525303929884,
      "grad_norm": 1.296875,
      "learning_rate": 0.000499986682104713,
      "loss": 5.9096,
      "mean_token_accuracy": 0.18470141440629959,
      "num_tokens": 891867.0,
      "step": 1120
    },
    {
      "entropy": 5.931029558181763,
      "epoch": 0.31806615776081426,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999855394545157,
      "loss": 6.0265,
      "mean_token_accuracy": 0.17036193311214448,
      "num_tokens": 895769.0,
      "step": 1125
    },
    {
      "entropy": 6.057384490966797,
      "epoch": 0.3194797851286401,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004999843497837317,
      "loss": 5.9771,
      "mean_token_accuracy": 0.16771688610315322,
      "num_tokens": 899640.0,
      "step": 1130
    },
    {
      "entropy": 5.935440540313721,
      "epoch": 0.3208934124964659,
      "grad_norm": 1.3125,
      "learning_rate": 0.00049998311309261,
      "loss": 5.9876,
      "mean_token_accuracy": 0.1825812965631485,
      "num_tokens": 903598.0,
      "step": 1135
    },
    {
      "entropy": 6.0772843837738035,
      "epoch": 0.32230703986429177,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004999818293814089,
      "loss": 6.0469,
      "mean_token_accuracy": 0.17144874930381776,
      "num_tokens": 907720.0,
      "step": 1140
    },
    {
      "entropy": 5.91833529472351,
      "epoch": 0.3237206672321176,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999804986503966,
      "loss": 6.001,
      "mean_token_accuracy": 0.1728842318058014,
      "num_tokens": 911531.0,
      "step": 1145
    },
    {
      "entropy": 6.0705688953399655,
      "epoch": 0.3251342945999435,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004999791208998513,
      "loss": 5.9162,
      "mean_token_accuracy": 0.17554775178432463,
      "num_tokens": 915428.0,
      "step": 1150
    },
    {
      "entropy": 5.908070373535156,
      "epoch": 0.3265479219677693,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999776961300611,
      "loss": 5.957,
      "mean_token_accuracy": 0.1746842861175537,
      "num_tokens": 919499.0,
      "step": 1155
    },
    {
      "entropy": 6.098568487167358,
      "epoch": 0.3279615493355951,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999762243413236,
      "loss": 5.9827,
      "mean_token_accuracy": 0.176666696369648,
      "num_tokens": 923530.0,
      "step": 1160
    },
    {
      "entropy": 5.904607772827148,
      "epoch": 0.329375176703421,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004999747055339464,
      "loss": 5.9302,
      "mean_token_accuracy": 0.18254795223474501,
      "num_tokens": 927770.0,
      "step": 1165
    },
    {
      "entropy": 5.955020999908447,
      "epoch": 0.33078880407124683,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004999731397082469,
      "loss": 5.9771,
      "mean_token_accuracy": 0.17447073385119438,
      "num_tokens": 932008.0,
      "step": 1170
    },
    {
      "entropy": 6.0833601474761965,
      "epoch": 0.3322024314390727,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999715268645524,
      "loss": 5.9811,
      "mean_token_accuracy": 0.18211427927017212,
      "num_tokens": 936388.0,
      "step": 1175
    },
    {
      "entropy": 5.904801702499389,
      "epoch": 0.3336160588068985,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004999698670032,
      "loss": 6.0315,
      "mean_token_accuracy": 0.1786898285150528,
      "num_tokens": 940436.0,
      "step": 1180
    },
    {
      "entropy": 6.072543048858643,
      "epoch": 0.33502968617472434,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999681601245366,
      "loss": 6.0486,
      "mean_token_accuracy": 0.17283598110079765,
      "num_tokens": 944296.0,
      "step": 1185
    },
    {
      "entropy": 6.052107238769532,
      "epoch": 0.3364433135425502,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999664062289189,
      "loss": 6.0668,
      "mean_token_accuracy": 0.17393210679292678,
      "num_tokens": 948318.0,
      "step": 1190
    },
    {
      "entropy": 6.093935489654541,
      "epoch": 0.33785694091037605,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004999646053167134,
      "loss": 5.9547,
      "mean_token_accuracy": 0.17875370681285857,
      "num_tokens": 952147.0,
      "step": 1195
    },
    {
      "entropy": 6.063412761688232,
      "epoch": 0.33927056827820185,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004999627573882966,
      "loss": 6.0947,
      "mean_token_accuracy": 0.16490099132061004,
      "num_tokens": 956445.0,
      "step": 1200
    },
    {
      "entropy": 5.901909017562867,
      "epoch": 0.3406841956460277,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004999608624440546,
      "loss": 5.9295,
      "mean_token_accuracy": 0.17917849570512773,
      "num_tokens": 960327.0,
      "step": 1205
    },
    {
      "entropy": 6.053218269348145,
      "epoch": 0.34209782301385355,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004999589204843834,
      "loss": 6.0542,
      "mean_token_accuracy": 0.17569087892770768,
      "num_tokens": 964387.0,
      "step": 1210
    },
    {
      "entropy": 6.057649898529053,
      "epoch": 0.3435114503816794,
      "grad_norm": 1.375,
      "learning_rate": 0.000499956931509689,
      "loss": 5.9893,
      "mean_token_accuracy": 0.17533947229385377,
      "num_tokens": 968327.0,
      "step": 1215
    },
    {
      "entropy": 6.110534811019898,
      "epoch": 0.3449250777495052,
      "grad_norm": 1.4375,
      "learning_rate": 0.000499954895520387,
      "loss": 6.0285,
      "mean_token_accuracy": 0.17372053861618042,
      "num_tokens": 972358.0,
      "step": 1220
    },
    {
      "entropy": 5.764717531204224,
      "epoch": 0.34633870511733106,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004999528125169028,
      "loss": 5.9328,
      "mean_token_accuracy": 0.17071535885334016,
      "num_tokens": 976266.0,
      "step": 1225
    },
    {
      "entropy": 5.965855121612549,
      "epoch": 0.3477523324851569,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000499950682499672,
      "loss": 6.0183,
      "mean_token_accuracy": 0.17038709074258804,
      "num_tokens": 980521.0,
      "step": 1230
    },
    {
      "entropy": 6.097711992263794,
      "epoch": 0.34916595985298277,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004999485054691395,
      "loss": 5.9159,
      "mean_token_accuracy": 0.17354207187891008,
      "num_tokens": 984413.0,
      "step": 1235
    },
    {
      "entropy": 6.042867660522461,
      "epoch": 0.3505795872208086,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004999462814257604,
      "loss": 6.0334,
      "mean_token_accuracy": 0.16417137533426285,
      "num_tokens": 988080.0,
      "step": 1240
    },
    {
      "entropy": 6.060603237152099,
      "epoch": 0.3519932145886344,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999440103699996,
      "loss": 5.9772,
      "mean_token_accuracy": 0.18074256777763367,
      "num_tokens": 992010.0,
      "step": 1245
    },
    {
      "entropy": 5.730267524719238,
      "epoch": 0.35340684195646027,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004999416923023316,
      "loss": 5.8617,
      "mean_token_accuracy": 0.18689277470111848,
      "num_tokens": 995721.0,
      "step": 1250
    },
    {
      "entropy": 5.8941511631011965,
      "epoch": 0.3548204693242861,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004999393272232409,
      "loss": 5.9628,
      "mean_token_accuracy": 0.1740591511130333,
      "num_tokens": 999763.0,
      "step": 1255
    },
    {
      "entropy": 6.133557987213135,
      "epoch": 0.356234096692112,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004999369151332218,
      "loss": 6.0322,
      "mean_token_accuracy": 0.17545345425605774,
      "num_tokens": 1003610.0,
      "step": 1260
    },
    {
      "entropy": 5.98744478225708,
      "epoch": 0.3576477240599378,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004999344560327784,
      "loss": 5.9228,
      "mean_token_accuracy": 0.17964100390672683,
      "num_tokens": 1007699.0,
      "step": 1265
    },
    {
      "entropy": 5.86641001701355,
      "epoch": 0.35906135142776363,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004999319499224247,
      "loss": 5.904,
      "mean_token_accuracy": 0.1844222739338875,
      "num_tokens": 1011573.0,
      "step": 1270
    },
    {
      "entropy": 5.9154126167297365,
      "epoch": 0.3604749787955895,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999293968026843,
      "loss": 6.0189,
      "mean_token_accuracy": 0.17944376170635223,
      "num_tokens": 1015569.0,
      "step": 1275
    },
    {
      "entropy": 6.00258355140686,
      "epoch": 0.36188860616341534,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004999267966740909,
      "loss": 5.8752,
      "mean_token_accuracy": 0.18310932219028472,
      "num_tokens": 1019451.0,
      "step": 1280
    },
    {
      "entropy": 5.903419113159179,
      "epoch": 0.3633022335312412,
      "grad_norm": 1.4375,
      "learning_rate": 0.000499924149537188,
      "loss": 5.8026,
      "mean_token_accuracy": 0.18908512145280837,
      "num_tokens": 1023084.0,
      "step": 1285
    },
    {
      "entropy": 5.935181188583374,
      "epoch": 0.364715860899067,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999214553925287,
      "loss": 5.9164,
      "mean_token_accuracy": 0.18504925072193146,
      "num_tokens": 1026941.0,
      "step": 1290
    },
    {
      "entropy": 5.918446159362793,
      "epoch": 0.36612948826689284,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999187142406761,
      "loss": 5.8357,
      "mean_token_accuracy": 0.1875179424881935,
      "num_tokens": 1030872.0,
      "step": 1295
    },
    {
      "entropy": 5.95609302520752,
      "epoch": 0.3675431156347187,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000499915926082203,
      "loss": 5.9616,
      "mean_token_accuracy": 0.17633716613054276,
      "num_tokens": 1035024.0,
      "step": 1300
    },
    {
      "entropy": 5.878836631774902,
      "epoch": 0.36895674300254455,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004999130909176923,
      "loss": 5.8229,
      "mean_token_accuracy": 0.17987947165966034,
      "num_tokens": 1038988.0,
      "step": 1305
    },
    {
      "entropy": 5.842767524719238,
      "epoch": 0.37037037037037035,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004999102087477362,
      "loss": 5.7784,
      "mean_token_accuracy": 0.19531663954257966,
      "num_tokens": 1043068.0,
      "step": 1310
    },
    {
      "entropy": 5.79612102508545,
      "epoch": 0.3717839977381962,
      "grad_norm": 1.375,
      "learning_rate": 0.0004999072795729376,
      "loss": 5.8488,
      "mean_token_accuracy": 0.18711667507886887,
      "num_tokens": 1046931.0,
      "step": 1315
    },
    {
      "entropy": 5.752354717254638,
      "epoch": 0.37319762510602206,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004999043033939081,
      "loss": 5.792,
      "mean_token_accuracy": 0.18978319317102432,
      "num_tokens": 1050718.0,
      "step": 1320
    },
    {
      "entropy": 5.9706896305084225,
      "epoch": 0.3746112524738479,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004999012802112701,
      "loss": 5.8829,
      "mean_token_accuracy": 0.17118766903877258,
      "num_tokens": 1054796.0,
      "step": 1325
    },
    {
      "entropy": 5.830494022369384,
      "epoch": 0.3760248798416737,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004998982100256552,
      "loss": 5.8218,
      "mean_token_accuracy": 0.1836886391043663,
      "num_tokens": 1058657.0,
      "step": 1330
    },
    {
      "entropy": 5.859378147125244,
      "epoch": 0.37743850720949956,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004998950928377052,
      "loss": 5.9446,
      "mean_token_accuracy": 0.18499152809381486,
      "num_tokens": 1062532.0,
      "step": 1335
    },
    {
      "entropy": 5.9526323795318605,
      "epoch": 0.3788521345773254,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004998919286480713,
      "loss": 5.7847,
      "mean_token_accuracy": 0.18415039628744126,
      "num_tokens": 1066274.0,
      "step": 1340
    },
    {
      "entropy": 5.845388412475586,
      "epoch": 0.38026576194515127,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004998887174574153,
      "loss": 5.898,
      "mean_token_accuracy": 0.1804192841053009,
      "num_tokens": 1070179.0,
      "step": 1345
    },
    {
      "entropy": 5.898585224151612,
      "epoch": 0.3816793893129771,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004998854592664079,
      "loss": 5.8693,
      "mean_token_accuracy": 0.1799743041396141,
      "num_tokens": 1073892.0,
      "step": 1350
    },
    {
      "entropy": 5.964764881134033,
      "epoch": 0.3830930166808029,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004998821540757301,
      "loss": 5.9282,
      "mean_token_accuracy": 0.18046344667673112,
      "num_tokens": 1078399.0,
      "step": 1355
    },
    {
      "entropy": 5.830195188522339,
      "epoch": 0.3845066440486288,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004998788018860728,
      "loss": 5.8309,
      "mean_token_accuracy": 0.18262680023908615,
      "num_tokens": 1082371.0,
      "step": 1360
    },
    {
      "entropy": 6.0038042068481445,
      "epoch": 0.38592027141645463,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004998754026981364,
      "loss": 5.9601,
      "mean_token_accuracy": 0.1667673259973526,
      "num_tokens": 1086474.0,
      "step": 1365
    },
    {
      "entropy": 5.914479732513428,
      "epoch": 0.3873338987842805,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004998719565126316,
      "loss": 5.8866,
      "mean_token_accuracy": 0.18313463181257247,
      "num_tokens": 1090519.0,
      "step": 1370
    },
    {
      "entropy": 5.778615236282349,
      "epoch": 0.3887475261521063,
      "grad_norm": 1.25,
      "learning_rate": 0.0004998684633302783,
      "loss": 5.8137,
      "mean_token_accuracy": 0.17858949154615403,
      "num_tokens": 1094450.0,
      "step": 1375
    },
    {
      "entropy": 6.106761884689331,
      "epoch": 0.39016115351993214,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004998649231518068,
      "loss": 5.8724,
      "mean_token_accuracy": 0.1757311686873436,
      "num_tokens": 1098249.0,
      "step": 1380
    },
    {
      "entropy": 5.754768514633179,
      "epoch": 0.391574780887758,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004998613359779568,
      "loss": 5.8116,
      "mean_token_accuracy": 0.18666931986808777,
      "num_tokens": 1101951.0,
      "step": 1385
    },
    {
      "entropy": 5.7896898746490475,
      "epoch": 0.39298840825558384,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004998577018094782,
      "loss": 5.8577,
      "mean_token_accuracy": 0.18320763409137725,
      "num_tokens": 1106097.0,
      "step": 1390
    },
    {
      "entropy": 5.87284574508667,
      "epoch": 0.3944020356234097,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004998540206471304,
      "loss": 5.7767,
      "mean_token_accuracy": 0.19519320279359817,
      "num_tokens": 1110031.0,
      "step": 1395
    },
    {
      "entropy": 5.86262526512146,
      "epoch": 0.3958156629912355,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004998502924916826,
      "loss": 5.9447,
      "mean_token_accuracy": 0.1763763576745987,
      "num_tokens": 1114146.0,
      "step": 1400
    },
    {
      "entropy": 5.977203845977783,
      "epoch": 0.39722929035906135,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004998465173439142,
      "loss": 5.9148,
      "mean_token_accuracy": 0.18122531175613404,
      "num_tokens": 1118161.0,
      "step": 1405
    },
    {
      "entropy": 5.8424163341522215,
      "epoch": 0.3986429177268872,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004998426952046142,
      "loss": 5.699,
      "mean_token_accuracy": 0.2005566254258156,
      "num_tokens": 1122014.0,
      "step": 1410
    },
    {
      "entropy": 5.684796667098999,
      "epoch": 0.40005654509471306,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004998388260745811,
      "loss": 5.8875,
      "mean_token_accuracy": 0.18261732757091523,
      "num_tokens": 1125929.0,
      "step": 1415
    },
    {
      "entropy": 5.9797446727752686,
      "epoch": 0.40147017246253885,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004998349099546238,
      "loss": 5.8277,
      "mean_token_accuracy": 0.17909095138311387,
      "num_tokens": 1129781.0,
      "step": 1420
    },
    {
      "entropy": 5.744542932510376,
      "epoch": 0.4028837998303647,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004998309468455606,
      "loss": 5.7346,
      "mean_token_accuracy": 0.19376294761896135,
      "num_tokens": 1133591.0,
      "step": 1425
    },
    {
      "entropy": 5.857560825347901,
      "epoch": 0.40429742719819056,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004998269367482199,
      "loss": 5.877,
      "mean_token_accuracy": 0.184601990878582,
      "num_tokens": 1137406.0,
      "step": 1430
    },
    {
      "entropy": 5.766568803787232,
      "epoch": 0.4057110545660164,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004998228796634396,
      "loss": 5.8087,
      "mean_token_accuracy": 0.18684434592723848,
      "num_tokens": 1141393.0,
      "step": 1435
    },
    {
      "entropy": 5.924833011627197,
      "epoch": 0.4071246819338422,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004998187755920677,
      "loss": 5.9094,
      "mean_token_accuracy": 0.18260146528482438,
      "num_tokens": 1145155.0,
      "step": 1440
    },
    {
      "entropy": 5.778390026092529,
      "epoch": 0.40853830930166807,
      "grad_norm": 1.375,
      "learning_rate": 0.0004998146245349621,
      "loss": 5.8873,
      "mean_token_accuracy": 0.18451330810785294,
      "num_tokens": 1149104.0,
      "step": 1445
    },
    {
      "entropy": 5.859285831451416,
      "epoch": 0.4099519366694939,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004998104264929901,
      "loss": 5.8317,
      "mean_token_accuracy": 0.1783102497458458,
      "num_tokens": 1153022.0,
      "step": 1450
    },
    {
      "entropy": 5.8625123500823975,
      "epoch": 0.4113655640373198,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004998061814670291,
      "loss": 5.8618,
      "mean_token_accuracy": 0.18217997699975969,
      "num_tokens": 1156965.0,
      "step": 1455
    },
    {
      "entropy": 5.861889934539795,
      "epoch": 0.41277919140514563,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004998018894579664,
      "loss": 5.7965,
      "mean_token_accuracy": 0.19408051073551177,
      "num_tokens": 1160893.0,
      "step": 1460
    },
    {
      "entropy": 5.9454748153686525,
      "epoch": 0.4141928187729714,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004997975504666989,
      "loss": 5.9433,
      "mean_token_accuracy": 0.17259618788957595,
      "num_tokens": 1165037.0,
      "step": 1465
    },
    {
      "entropy": 5.8799817085266115,
      "epoch": 0.4156064461407973,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004997931644941335,
      "loss": 5.8166,
      "mean_token_accuracy": 0.17416656017303467,
      "num_tokens": 1168911.0,
      "step": 1470
    },
    {
      "entropy": 5.8028968334197994,
      "epoch": 0.41702007350862313,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004997887315411867,
      "loss": 5.815,
      "mean_token_accuracy": 0.19042931646108627,
      "num_tokens": 1172861.0,
      "step": 1475
    },
    {
      "entropy": 5.802371263504028,
      "epoch": 0.418433700876449,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004997842516087852,
      "loss": 5.852,
      "mean_token_accuracy": 0.18393263071775437,
      "num_tokens": 1177361.0,
      "step": 1480
    },
    {
      "entropy": 5.927545642852783,
      "epoch": 0.4198473282442748,
      "grad_norm": 1.375,
      "learning_rate": 0.0004997797246978651,
      "loss": 5.8468,
      "mean_token_accuracy": 0.17632220238447188,
      "num_tokens": 1181549.0,
      "step": 1485
    },
    {
      "entropy": 5.839781713485718,
      "epoch": 0.42126095561210064,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004997751508093724,
      "loss": 5.8349,
      "mean_token_accuracy": 0.1818211019039154,
      "num_tokens": 1185685.0,
      "step": 1490
    },
    {
      "entropy": 5.837100982666016,
      "epoch": 0.4226745829799265,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004997705299442632,
      "loss": 5.9818,
      "mean_token_accuracy": 0.17524709403514863,
      "num_tokens": 1189557.0,
      "step": 1495
    },
    {
      "entropy": 5.887031126022339,
      "epoch": 0.42408821034775235,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004997658621035033,
      "loss": 5.8772,
      "mean_token_accuracy": 0.1873631939291954,
      "num_tokens": 1193969.0,
      "step": 1500
    },
    {
      "entropy": 5.887158012390136,
      "epoch": 0.42550183771557815,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004997611472880681,
      "loss": 5.8533,
      "mean_token_accuracy": 0.19286071956157685,
      "num_tokens": 1197843.0,
      "step": 1505
    },
    {
      "entropy": 5.8976044178009035,
      "epoch": 0.426915465083404,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000499756385498943,
      "loss": 5.8637,
      "mean_token_accuracy": 0.17578741312026977,
      "num_tokens": 1201938.0,
      "step": 1510
    },
    {
      "entropy": 5.803966426849366,
      "epoch": 0.42832909245122985,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004997515767371231,
      "loss": 5.6951,
      "mean_token_accuracy": 0.1796621337532997,
      "num_tokens": 1205757.0,
      "step": 1515
    },
    {
      "entropy": 5.830338954925537,
      "epoch": 0.4297427198190557,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004997467210036135,
      "loss": 5.8731,
      "mean_token_accuracy": 0.18170178532600403,
      "num_tokens": 1209779.0,
      "step": 1520
    },
    {
      "entropy": 5.859526443481445,
      "epoch": 0.43115634718688156,
      "grad_norm": 1.359375,
      "learning_rate": 0.000499741818299429,
      "loss": 5.8099,
      "mean_token_accuracy": 0.18744910210371019,
      "num_tokens": 1213598.0,
      "step": 1525
    },
    {
      "entropy": 5.872071599960327,
      "epoch": 0.43256997455470736,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004997368686255943,
      "loss": 5.7304,
      "mean_token_accuracy": 0.18553417176008224,
      "num_tokens": 1217452.0,
      "step": 1530
    },
    {
      "entropy": 5.825813913345337,
      "epoch": 0.4339836019225332,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004997318719831438,
      "loss": 5.8642,
      "mean_token_accuracy": 0.1814598798751831,
      "num_tokens": 1221354.0,
      "step": 1535
    },
    {
      "entropy": 5.778457069396973,
      "epoch": 0.43539722929035907,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004997268283731216,
      "loss": 5.7543,
      "mean_token_accuracy": 0.1856408089399338,
      "num_tokens": 1225217.0,
      "step": 1540
    },
    {
      "entropy": 5.878414440155029,
      "epoch": 0.4368108566581849,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004997217377965821,
      "loss": 5.7805,
      "mean_token_accuracy": 0.17437221556901933,
      "num_tokens": 1229016.0,
      "step": 1545
    },
    {
      "entropy": 5.771785020828247,
      "epoch": 0.4382244840260107,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000499716600254589,
      "loss": 5.8818,
      "mean_token_accuracy": 0.1883689671754837,
      "num_tokens": 1232825.0,
      "step": 1550
    },
    {
      "entropy": 5.897240686416626,
      "epoch": 0.4396381113938366,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000499711415748216,
      "loss": 5.799,
      "mean_token_accuracy": 0.18249992430210113,
      "num_tokens": 1237032.0,
      "step": 1555
    },
    {
      "entropy": 5.602204322814941,
      "epoch": 0.4410517387616624,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004997061842785468,
      "loss": 5.6803,
      "mean_token_accuracy": 0.19400090873241424,
      "num_tokens": 1241043.0,
      "step": 1560
    },
    {
      "entropy": 5.717684745788574,
      "epoch": 0.4424653661294883,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004997009058466745,
      "loss": 5.8311,
      "mean_token_accuracy": 0.18614000529050828,
      "num_tokens": 1245175.0,
      "step": 1565
    },
    {
      "entropy": 5.799765300750733,
      "epoch": 0.44387899349731413,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004996955804537024,
      "loss": 5.7576,
      "mean_token_accuracy": 0.18300676196813584,
      "num_tokens": 1249540.0,
      "step": 1570
    },
    {
      "entropy": 5.823600387573242,
      "epoch": 0.44529262086513993,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004996902081007435,
      "loss": 5.73,
      "mean_token_accuracy": 0.18506772816181183,
      "num_tokens": 1253532.0,
      "step": 1575
    },
    {
      "entropy": 5.833622980117798,
      "epoch": 0.4467062482329658,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004996847887889205,
      "loss": 5.7169,
      "mean_token_accuracy": 0.18705481737852098,
      "num_tokens": 1257492.0,
      "step": 1580
    },
    {
      "entropy": 5.734306573867798,
      "epoch": 0.44811987560079164,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004996793225193662,
      "loss": 5.8279,
      "mean_token_accuracy": 0.18313267529010774,
      "num_tokens": 1261388.0,
      "step": 1585
    },
    {
      "entropy": 5.834798336029053,
      "epoch": 0.4495335029686175,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004996738092932227,
      "loss": 5.7397,
      "mean_token_accuracy": 0.18718260675668716,
      "num_tokens": 1265417.0,
      "step": 1590
    },
    {
      "entropy": 5.671156358718872,
      "epoch": 0.4509471303364433,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004996682491116425,
      "loss": 5.7071,
      "mean_token_accuracy": 0.19657584577798842,
      "num_tokens": 1269454.0,
      "step": 1595
    },
    {
      "entropy": 5.7325883388519285,
      "epoch": 0.45236075770426915,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004996626419757875,
      "loss": 5.8176,
      "mean_token_accuracy": 0.18559218645095826,
      "num_tokens": 1273677.0,
      "step": 1600
    },
    {
      "entropy": 5.821763038635254,
      "epoch": 0.453774385072095,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004996569878868296,
      "loss": 5.6485,
      "mean_token_accuracy": 0.19062581062316894,
      "num_tokens": 1277683.0,
      "step": 1605
    },
    {
      "entropy": 5.6672616481781,
      "epoch": 0.45518801243992085,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004996512868459505,
      "loss": 5.7728,
      "mean_token_accuracy": 0.18369143903255464,
      "num_tokens": 1281945.0,
      "step": 1610
    },
    {
      "entropy": 5.595178031921387,
      "epoch": 0.45660163980774665,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004996455388543416,
      "loss": 5.7011,
      "mean_token_accuracy": 0.1911506325006485,
      "num_tokens": 1285709.0,
      "step": 1615
    },
    {
      "entropy": 5.877404975891113,
      "epoch": 0.4580152671755725,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004996397439132042,
      "loss": 5.8661,
      "mean_token_accuracy": 0.17192684412002562,
      "num_tokens": 1289772.0,
      "step": 1620
    },
    {
      "entropy": 5.772934436798096,
      "epoch": 0.45942889454339836,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004996339020237496,
      "loss": 5.7621,
      "mean_token_accuracy": 0.18746172338724137,
      "num_tokens": 1293671.0,
      "step": 1625
    },
    {
      "entropy": 5.825155878067017,
      "epoch": 0.4608425219112242,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004996280131871985,
      "loss": 5.6843,
      "mean_token_accuracy": 0.1883016973733902,
      "num_tokens": 1297334.0,
      "step": 1630
    },
    {
      "entropy": 5.71227011680603,
      "epoch": 0.46225614927905007,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004996220774047816,
      "loss": 5.7632,
      "mean_token_accuracy": 0.18459136486053468,
      "num_tokens": 1301219.0,
      "step": 1635
    },
    {
      "entropy": 5.700411558151245,
      "epoch": 0.46366977664687586,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004996160946777398,
      "loss": 5.7423,
      "mean_token_accuracy": 0.18890873938798905,
      "num_tokens": 1305166.0,
      "step": 1640
    },
    {
      "entropy": 5.676112222671509,
      "epoch": 0.4650834040147017,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004996100650073229,
      "loss": 5.6738,
      "mean_token_accuracy": 0.19222360402345656,
      "num_tokens": 1308958.0,
      "step": 1645
    },
    {
      "entropy": 5.767787170410156,
      "epoch": 0.46649703138252757,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004996039883947915,
      "loss": 5.909,
      "mean_token_accuracy": 0.17916598170995712,
      "num_tokens": 1313289.0,
      "step": 1650
    },
    {
      "entropy": 5.912643051147461,
      "epoch": 0.4679106587503534,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004995978648414154,
      "loss": 5.7442,
      "mean_token_accuracy": 0.18683811128139496,
      "num_tokens": 1317366.0,
      "step": 1655
    },
    {
      "entropy": 5.729746007919312,
      "epoch": 0.4693242861181792,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004995916943484745,
      "loss": 5.7547,
      "mean_token_accuracy": 0.1879502385854721,
      "num_tokens": 1321362.0,
      "step": 1660
    },
    {
      "entropy": 5.774411725997925,
      "epoch": 0.4707379134860051,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004995854769172583,
      "loss": 5.7037,
      "mean_token_accuracy": 0.18776724189519883,
      "num_tokens": 1325147.0,
      "step": 1665
    },
    {
      "entropy": 5.843069744110108,
      "epoch": 0.47215154085383093,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004995792125490662,
      "loss": 5.778,
      "mean_token_accuracy": 0.17874198108911515,
      "num_tokens": 1329072.0,
      "step": 1670
    },
    {
      "entropy": 5.639906167984009,
      "epoch": 0.4735651682216568,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004995729012452074,
      "loss": 5.6564,
      "mean_token_accuracy": 0.20082349181175232,
      "num_tokens": 1332947.0,
      "step": 1675
    },
    {
      "entropy": 5.622266244888306,
      "epoch": 0.47497879558948264,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004995665430070009,
      "loss": 5.6718,
      "mean_token_accuracy": 0.1948164239525795,
      "num_tokens": 1336903.0,
      "step": 1680
    },
    {
      "entropy": 5.792095184326172,
      "epoch": 0.47639242295730844,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004995601378357757,
      "loss": 5.7422,
      "mean_token_accuracy": 0.18691984713077545,
      "num_tokens": 1340989.0,
      "step": 1685
    },
    {
      "entropy": 5.864161443710327,
      "epoch": 0.4778060503251343,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004995536857328702,
      "loss": 5.8213,
      "mean_token_accuracy": 0.17689985930919647,
      "num_tokens": 1345251.0,
      "step": 1690
    },
    {
      "entropy": 5.904580068588257,
      "epoch": 0.47921967769296014,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004995471866996332,
      "loss": 5.9124,
      "mean_token_accuracy": 0.17539610639214515,
      "num_tokens": 1349167.0,
      "step": 1695
    },
    {
      "entropy": 5.778077411651611,
      "epoch": 0.480633305060786,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004995406407374226,
      "loss": 5.6184,
      "mean_token_accuracy": 0.2023831859230995,
      "num_tokens": 1353309.0,
      "step": 1700
    },
    {
      "entropy": 5.721492433547974,
      "epoch": 0.4820469324286118,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004995340478476067,
      "loss": 5.7534,
      "mean_token_accuracy": 0.19217092245817186,
      "num_tokens": 1357129.0,
      "step": 1705
    },
    {
      "entropy": 5.7220958232879635,
      "epoch": 0.48346055979643765,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004995274080315631,
      "loss": 5.692,
      "mean_token_accuracy": 0.18690054416656493,
      "num_tokens": 1360824.0,
      "step": 1710
    },
    {
      "entropy": 5.69368953704834,
      "epoch": 0.4848741871642635,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004995207212906798,
      "loss": 5.7049,
      "mean_token_accuracy": 0.2015200823545456,
      "num_tokens": 1364688.0,
      "step": 1715
    },
    {
      "entropy": 5.706001472473145,
      "epoch": 0.48628781453208936,
      "grad_norm": 1.25,
      "learning_rate": 0.000499513987626354,
      "loss": 5.6655,
      "mean_token_accuracy": 0.19725179374217988,
      "num_tokens": 1368889.0,
      "step": 1720
    },
    {
      "entropy": 5.746529531478882,
      "epoch": 0.48770144189991516,
      "grad_norm": 1.375,
      "learning_rate": 0.0004995072070399933,
      "loss": 5.7719,
      "mean_token_accuracy": 0.20015778094530107,
      "num_tokens": 1372912.0,
      "step": 1725
    },
    {
      "entropy": 5.6686793804168705,
      "epoch": 0.489115069267741,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004995003795330146,
      "loss": 5.6134,
      "mean_token_accuracy": 0.1967315584421158,
      "num_tokens": 1376566.0,
      "step": 1730
    },
    {
      "entropy": 5.498853397369385,
      "epoch": 0.49052869663556686,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004994935051068447,
      "loss": 5.6377,
      "mean_token_accuracy": 0.19570326954126357,
      "num_tokens": 1380710.0,
      "step": 1735
    },
    {
      "entropy": 5.740759801864624,
      "epoch": 0.4919423240033927,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004994865837629206,
      "loss": 5.7169,
      "mean_token_accuracy": 0.19236577451229095,
      "num_tokens": 1384485.0,
      "step": 1740
    },
    {
      "entropy": 5.751363229751587,
      "epoch": 0.49335595137121857,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004994796155026887,
      "loss": 5.6802,
      "mean_token_accuracy": 0.19381756484508514,
      "num_tokens": 1388297.0,
      "step": 1745
    },
    {
      "entropy": 5.605444765090942,
      "epoch": 0.49476957873904437,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004994726003276053,
      "loss": 5.7062,
      "mean_token_accuracy": 0.19098659604787827,
      "num_tokens": 1392390.0,
      "step": 1750
    },
    {
      "entropy": 5.733315515518188,
      "epoch": 0.4961832061068702,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004994655382391365,
      "loss": 5.7065,
      "mean_token_accuracy": 0.18861762136220933,
      "num_tokens": 1396561.0,
      "step": 1755
    },
    {
      "entropy": 5.754003095626831,
      "epoch": 0.4975968334746961,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004994584292387582,
      "loss": 5.7473,
      "mean_token_accuracy": 0.19067753851413727,
      "num_tokens": 1400556.0,
      "step": 1760
    },
    {
      "entropy": 5.826990175247192,
      "epoch": 0.49901046084252193,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004994512733279562,
      "loss": 5.7787,
      "mean_token_accuracy": 0.18018803298473357,
      "num_tokens": 1404503.0,
      "step": 1765
    },
    {
      "entropy": 5.709981107711792,
      "epoch": 0.5004240882103478,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004994440705082261,
      "loss": 5.7347,
      "mean_token_accuracy": 0.1765083149075508,
      "num_tokens": 1408504.0,
      "step": 1770
    },
    {
      "entropy": 5.694883728027344,
      "epoch": 0.5018377155781736,
      "grad_norm": 1.375,
      "learning_rate": 0.0004994368207810731,
      "loss": 5.5512,
      "mean_token_accuracy": 0.20656106472015381,
      "num_tokens": 1412429.0,
      "step": 1775
    },
    {
      "entropy": 5.594587898254394,
      "epoch": 0.5032513429459994,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004994295241480126,
      "loss": 5.6607,
      "mean_token_accuracy": 0.19336072951555253,
      "num_tokens": 1416466.0,
      "step": 1780
    },
    {
      "entropy": 5.69144377708435,
      "epoch": 0.5046649703138253,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004994221806105692,
      "loss": 5.6484,
      "mean_token_accuracy": 0.189523321390152,
      "num_tokens": 1420226.0,
      "step": 1785
    },
    {
      "entropy": 5.595638227462769,
      "epoch": 0.5060785976816511,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000499414790170278,
      "loss": 5.6555,
      "mean_token_accuracy": 0.18988582044839858,
      "num_tokens": 1423987.0,
      "step": 1790
    },
    {
      "entropy": 5.80430588722229,
      "epoch": 0.507492225049477,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004994073528286833,
      "loss": 5.69,
      "mean_token_accuracy": 0.18927541226148606,
      "num_tokens": 1428264.0,
      "step": 1795
    },
    {
      "entropy": 5.692843675613403,
      "epoch": 0.5089058524173028,
      "grad_norm": 1.375,
      "learning_rate": 0.0004993998685873395,
      "loss": 5.5987,
      "mean_token_accuracy": 0.1881292626261711,
      "num_tokens": 1432350.0,
      "step": 1800
    },
    {
      "entropy": 5.507918214797973,
      "epoch": 0.5103194797851286,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004993923374478109,
      "loss": 5.5961,
      "mean_token_accuracy": 0.20146704465150833,
      "num_tokens": 1436298.0,
      "step": 1805
    },
    {
      "entropy": 5.571535301208496,
      "epoch": 0.5117331071529545,
      "grad_norm": 1.375,
      "learning_rate": 0.0004993847594116713,
      "loss": 5.5513,
      "mean_token_accuracy": 0.20241557210683822,
      "num_tokens": 1440206.0,
      "step": 1810
    },
    {
      "entropy": 5.7816972732543945,
      "epoch": 0.5131467345207803,
      "grad_norm": 1.375,
      "learning_rate": 0.0004993771344805046,
      "loss": 5.7427,
      "mean_token_accuracy": 0.18699136078357698,
      "num_tokens": 1444566.0,
      "step": 1815
    },
    {
      "entropy": 5.65059723854065,
      "epoch": 0.5145603618886062,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004993694626559043,
      "loss": 5.6951,
      "mean_token_accuracy": 0.19032856076955795,
      "num_tokens": 1448792.0,
      "step": 1820
    },
    {
      "entropy": 5.783305358886719,
      "epoch": 0.515973989256432,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004993617439394737,
      "loss": 5.7512,
      "mean_token_accuracy": 0.18141021430492402,
      "num_tokens": 1452456.0,
      "step": 1825
    },
    {
      "entropy": 5.619701290130616,
      "epoch": 0.5173876166242578,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004993539783328261,
      "loss": 5.6665,
      "mean_token_accuracy": 0.19362071603536607,
      "num_tokens": 1456415.0,
      "step": 1830
    },
    {
      "entropy": 5.583071756362915,
      "epoch": 0.5188012439920837,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004993461658375845,
      "loss": 5.5992,
      "mean_token_accuracy": 0.20504723042249678,
      "num_tokens": 1460390.0,
      "step": 1835
    },
    {
      "entropy": 5.701898860931396,
      "epoch": 0.5202148713599095,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004993383064553814,
      "loss": 5.742,
      "mean_token_accuracy": 0.19340634495019912,
      "num_tokens": 1464363.0,
      "step": 1840
    },
    {
      "entropy": 5.671453237533569,
      "epoch": 0.5216284987277354,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004993304001878595,
      "loss": 5.6109,
      "mean_token_accuracy": 0.20281730890274047,
      "num_tokens": 1468372.0,
      "step": 1845
    },
    {
      "entropy": 5.493208885192871,
      "epoch": 0.5230421260955612,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004993224470366712,
      "loss": 5.7438,
      "mean_token_accuracy": 0.18834396749734877,
      "num_tokens": 1472272.0,
      "step": 1850
    },
    {
      "entropy": 5.684529161453247,
      "epoch": 0.524455753463387,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004993144470034787,
      "loss": 5.506,
      "mean_token_accuracy": 0.2056361585855484,
      "num_tokens": 1476195.0,
      "step": 1855
    },
    {
      "entropy": 5.637450647354126,
      "epoch": 0.5258693808312129,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004993064000899538,
      "loss": 5.6958,
      "mean_token_accuracy": 0.18898120522499084,
      "num_tokens": 1480364.0,
      "step": 1860
    },
    {
      "entropy": 5.476922035217285,
      "epoch": 0.5272830081990387,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004992983062977784,
      "loss": 5.6845,
      "mean_token_accuracy": 0.19763999432325363,
      "num_tokens": 1484523.0,
      "step": 1865
    },
    {
      "entropy": 5.7142651081085205,
      "epoch": 0.5286966355668645,
      "grad_norm": 1.5,
      "learning_rate": 0.0004992901656286439,
      "loss": 5.7271,
      "mean_token_accuracy": 0.18795137852430344,
      "num_tokens": 1488456.0,
      "step": 1870
    },
    {
      "entropy": 5.598298358917236,
      "epoch": 0.5301102629346904,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000499281978084252,
      "loss": 5.5556,
      "mean_token_accuracy": 0.20567425787448884,
      "num_tokens": 1492599.0,
      "step": 1875
    },
    {
      "entropy": 5.616948032379151,
      "epoch": 0.5315238903025162,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004992737436663134,
      "loss": 5.5724,
      "mean_token_accuracy": 0.21480470895767212,
      "num_tokens": 1496373.0,
      "step": 1880
    },
    {
      "entropy": 5.541384172439575,
      "epoch": 0.5329375176703421,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004992654623765493,
      "loss": 5.6192,
      "mean_token_accuracy": 0.19836001247167587,
      "num_tokens": 1500288.0,
      "step": 1885
    },
    {
      "entropy": 5.764281368255615,
      "epoch": 0.5343511450381679,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004992571342166904,
      "loss": 5.7485,
      "mean_token_accuracy": 0.18413073122501372,
      "num_tokens": 1504169.0,
      "step": 1890
    },
    {
      "entropy": 5.751831531524658,
      "epoch": 0.5357647724059937,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004992487591884772,
      "loss": 5.6864,
      "mean_token_accuracy": 0.19226602613925933,
      "num_tokens": 1508292.0,
      "step": 1895
    },
    {
      "entropy": 5.658072376251221,
      "epoch": 0.5371783997738196,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00049924033729366,
      "loss": 5.7019,
      "mean_token_accuracy": 0.19271002262830733,
      "num_tokens": 1512366.0,
      "step": 1900
    },
    {
      "entropy": 5.639523458480835,
      "epoch": 0.5385920271416454,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000499231868533999,
      "loss": 5.6872,
      "mean_token_accuracy": 0.19010829776525498,
      "num_tokens": 1516627.0,
      "step": 1905
    },
    {
      "entropy": 5.601494407653808,
      "epoch": 0.5400056545094714,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004992233529112642,
      "loss": 5.5366,
      "mean_token_accuracy": 0.20517586618661882,
      "num_tokens": 1520541.0,
      "step": 1910
    },
    {
      "entropy": 5.734471940994263,
      "epoch": 0.5414192818772972,
      "grad_norm": 1.375,
      "learning_rate": 0.0004992147904272349,
      "loss": 5.6739,
      "mean_token_accuracy": 0.1813078209757805,
      "num_tokens": 1524505.0,
      "step": 1915
    },
    {
      "entropy": 5.605487871170044,
      "epoch": 0.542832909245123,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004992061810837012,
      "loss": 5.6159,
      "mean_token_accuracy": 0.19128397554159166,
      "num_tokens": 1528658.0,
      "step": 1920
    },
    {
      "entropy": 5.487127208709717,
      "epoch": 0.5442465366129489,
      "grad_norm": 1.359375,
      "learning_rate": 0.000499197524882462,
      "loss": 5.563,
      "mean_token_accuracy": 0.20296506136655806,
      "num_tokens": 1532620.0,
      "step": 1925
    },
    {
      "entropy": 5.600309658050537,
      "epoch": 0.5456601639807747,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004991888218253265,
      "loss": 5.6726,
      "mean_token_accuracy": 0.17892235070466994,
      "num_tokens": 1536636.0,
      "step": 1930
    },
    {
      "entropy": 5.678558778762818,
      "epoch": 0.5470737913486005,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004991800719141136,
      "loss": 5.6827,
      "mean_token_accuracy": 0.18799914866685868,
      "num_tokens": 1540821.0,
      "step": 1935
    },
    {
      "entropy": 5.50041823387146,
      "epoch": 0.5484874187164264,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004991712751506518,
      "loss": 5.5987,
      "mean_token_accuracy": 0.19327885210514067,
      "num_tokens": 1544700.0,
      "step": 1940
    },
    {
      "entropy": 5.620293855667114,
      "epoch": 0.5499010460842522,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004991624315367798,
      "loss": 5.6134,
      "mean_token_accuracy": 0.19923721998929977,
      "num_tokens": 1548770.0,
      "step": 1945
    },
    {
      "entropy": 5.722182416915894,
      "epoch": 0.5513146734520781,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004991535410743459,
      "loss": 5.6994,
      "mean_token_accuracy": 0.18269776701927185,
      "num_tokens": 1553015.0,
      "step": 1950
    },
    {
      "entropy": 5.609702014923096,
      "epoch": 0.5527283008199039,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004991446037652079,
      "loss": 5.674,
      "mean_token_accuracy": 0.18456978499889373,
      "num_tokens": 1556849.0,
      "step": 1955
    },
    {
      "entropy": 5.5437146663665775,
      "epoch": 0.5541419281877297,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004991356196112339,
      "loss": 5.6453,
      "mean_token_accuracy": 0.19666575640439987,
      "num_tokens": 1560407.0,
      "step": 1960
    },
    {
      "entropy": 5.6902313232421875,
      "epoch": 0.5555555555555556,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004991265886143014,
      "loss": 5.5509,
      "mean_token_accuracy": 0.20052429139614106,
      "num_tokens": 1564473.0,
      "step": 1965
    },
    {
      "entropy": 5.525343418121338,
      "epoch": 0.5569691829233814,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004991175107762978,
      "loss": 5.7307,
      "mean_token_accuracy": 0.19222700595855713,
      "num_tokens": 1568141.0,
      "step": 1970
    },
    {
      "entropy": 5.702168750762939,
      "epoch": 0.5583828102912073,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004991083860991204,
      "loss": 5.7182,
      "mean_token_accuracy": 0.1973864495754242,
      "num_tokens": 1572130.0,
      "step": 1975
    },
    {
      "entropy": 5.691175413131714,
      "epoch": 0.5597964376590331,
      "grad_norm": 1.4375,
      "learning_rate": 0.000499099214584676,
      "loss": 5.6929,
      "mean_token_accuracy": 0.19352537095546724,
      "num_tokens": 1576243.0,
      "step": 1980
    },
    {
      "entropy": 5.743717002868652,
      "epoch": 0.5612100650268589,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004990899962348817,
      "loss": 5.6288,
      "mean_token_accuracy": 0.19254789650440216,
      "num_tokens": 1579916.0,
      "step": 1985
    },
    {
      "entropy": 5.515009832382202,
      "epoch": 0.5626236923946848,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004990807310516638,
      "loss": 5.4585,
      "mean_token_accuracy": 0.20692042708396913,
      "num_tokens": 1584075.0,
      "step": 1990
    },
    {
      "entropy": 5.662541103363037,
      "epoch": 0.5640373197625106,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004990714190369588,
      "loss": 5.6659,
      "mean_token_accuracy": 0.1924101397395134,
      "num_tokens": 1588026.0,
      "step": 1995
    },
    {
      "entropy": 5.66917872428894,
      "epoch": 0.5654509471303364,
      "grad_norm": 1.3125,
      "learning_rate": 0.000499062060192713,
      "loss": 5.6495,
      "mean_token_accuracy": 0.1943264052271843,
      "num_tokens": 1592420.0,
      "step": 2000
    },
    {
      "epoch": 0.5654509471303364,
      "eval_entropy": 5.499324915850103,
      "eval_loss": 5.823847770690918,
      "eval_mean_token_accuracy": 0.19281912291413245,
      "eval_num_tokens": 1592420.0,
      "eval_runtime": 6.4866,
      "eval_samples_per_second": 1739.58,
      "eval_steps_per_second": 217.525,
      "step": 2000
    },
    {
      "entropy": 5.519136238098144,
      "epoch": 0.5668645744981623,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000499052654520882,
      "loss": 5.5261,
      "mean_token_accuracy": 0.20804719775915145,
      "num_tokens": 1596409.0,
      "step": 2005
    },
    {
      "entropy": 5.650597095489502,
      "epoch": 0.5682782018659881,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004990432020234317,
      "loss": 5.6336,
      "mean_token_accuracy": 0.19576438665390014,
      "num_tokens": 1600549.0,
      "step": 2010
    },
    {
      "entropy": 5.564015674591064,
      "epoch": 0.569691829233814,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004990337027023378,
      "loss": 5.5238,
      "mean_token_accuracy": 0.1996455729007721,
      "num_tokens": 1604145.0,
      "step": 2015
    },
    {
      "entropy": 5.581352424621582,
      "epoch": 0.5711054566016398,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004990241565595853,
      "loss": 5.6523,
      "mean_token_accuracy": 0.18710000663995743,
      "num_tokens": 1608078.0,
      "step": 2020
    },
    {
      "entropy": 5.547347831726074,
      "epoch": 0.5725190839694656,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004990145635971694,
      "loss": 5.5746,
      "mean_token_accuracy": 0.19167159497737885,
      "num_tokens": 1612038.0,
      "step": 2025
    },
    {
      "entropy": 5.5548622608184814,
      "epoch": 0.5739327113372915,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004990049238170949,
      "loss": 5.6632,
      "mean_token_accuracy": 0.18320369720458984,
      "num_tokens": 1616059.0,
      "step": 2030
    },
    {
      "entropy": 5.510371541976928,
      "epoch": 0.5753463387051173,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004989952372213767,
      "loss": 5.4918,
      "mean_token_accuracy": 0.21322744935750962,
      "num_tokens": 1619653.0,
      "step": 2035
    },
    {
      "entropy": 5.580099773406983,
      "epoch": 0.5767599660729432,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004989855038120389,
      "loss": 5.5736,
      "mean_token_accuracy": 0.19756310880184175,
      "num_tokens": 1623820.0,
      "step": 2040
    },
    {
      "entropy": 5.677244377136231,
      "epoch": 0.578173593440769,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000498975723591116,
      "loss": 5.7295,
      "mean_token_accuracy": 0.19284142553806305,
      "num_tokens": 1627606.0,
      "step": 2045
    },
    {
      "entropy": 5.780484628677368,
      "epoch": 0.5795872208085948,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004989658965606519,
      "loss": 5.6816,
      "mean_token_accuracy": 0.18871309161186217,
      "num_tokens": 1631581.0,
      "step": 2050
    },
    {
      "entropy": 5.728699779510498,
      "epoch": 0.5810008481764207,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004989560227227003,
      "loss": 5.644,
      "mean_token_accuracy": 0.1916726589202881,
      "num_tokens": 1635670.0,
      "step": 2055
    },
    {
      "entropy": 5.658165550231933,
      "epoch": 0.5824144755442465,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004989461020793248,
      "loss": 5.6331,
      "mean_token_accuracy": 0.19252020567655564,
      "num_tokens": 1639860.0,
      "step": 2060
    },
    {
      "entropy": 5.598748302459716,
      "epoch": 0.5838281029120724,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004989361346325989,
      "loss": 5.6358,
      "mean_token_accuracy": 0.19433485567569733,
      "num_tokens": 1644119.0,
      "step": 2065
    },
    {
      "entropy": 5.515763998031616,
      "epoch": 0.5852417302798982,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004989261203846056,
      "loss": 5.5825,
      "mean_token_accuracy": 0.19371080547571182,
      "num_tokens": 1648106.0,
      "step": 2070
    },
    {
      "entropy": 5.616602373123169,
      "epoch": 0.586655357647724,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004989160593374379,
      "loss": 5.6641,
      "mean_token_accuracy": 0.1891575813293457,
      "num_tokens": 1652430.0,
      "step": 2075
    },
    {
      "entropy": 5.664251041412354,
      "epoch": 0.5880689850155499,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004989059514931985,
      "loss": 5.6261,
      "mean_token_accuracy": 0.1945906773209572,
      "num_tokens": 1656775.0,
      "step": 2080
    },
    {
      "entropy": 5.666239261627197,
      "epoch": 0.5894826123833757,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004988957968539997,
      "loss": 5.7317,
      "mean_token_accuracy": 0.19427189528942107,
      "num_tokens": 1661158.0,
      "step": 2085
    },
    {
      "entropy": 5.61161093711853,
      "epoch": 0.5908962397512015,
      "grad_norm": 1.421875,
      "learning_rate": 0.000498885595421964,
      "loss": 5.561,
      "mean_token_accuracy": 0.1957390457391739,
      "num_tokens": 1664839.0,
      "step": 2090
    },
    {
      "entropy": 5.4742598056793215,
      "epoch": 0.5923098671190274,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004988753471992233,
      "loss": 5.4318,
      "mean_token_accuracy": 0.20599592179059983,
      "num_tokens": 1668493.0,
      "step": 2095
    },
    {
      "entropy": 5.54174165725708,
      "epoch": 0.5937234944868532,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004988650521879195,
      "loss": 5.6488,
      "mean_token_accuracy": 0.1846747562289238,
      "num_tokens": 1672376.0,
      "step": 2100
    },
    {
      "entropy": 5.673175811767578,
      "epoch": 0.5951371218546792,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000498854710390204,
      "loss": 5.5404,
      "mean_token_accuracy": 0.1986124485731125,
      "num_tokens": 1676315.0,
      "step": 2105
    },
    {
      "entropy": 5.482773733139038,
      "epoch": 0.596550749222505,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004988443218082384,
      "loss": 5.6602,
      "mean_token_accuracy": 0.19914872348308563,
      "num_tokens": 1680195.0,
      "step": 2110
    },
    {
      "entropy": 5.455413579940796,
      "epoch": 0.5979643765903307,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004988338864441937,
      "loss": 5.4779,
      "mean_token_accuracy": 0.19919375330209732,
      "num_tokens": 1684194.0,
      "step": 2115
    },
    {
      "entropy": 5.688107585906982,
      "epoch": 0.5993780039581567,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004988234043002508,
      "loss": 5.5401,
      "mean_token_accuracy": 0.19828267991542817,
      "num_tokens": 1688303.0,
      "step": 2120
    },
    {
      "entropy": 5.557444143295288,
      "epoch": 0.6007916313259825,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004988128753786006,
      "loss": 5.7322,
      "mean_token_accuracy": 0.1830365002155304,
      "num_tokens": 1692806.0,
      "step": 2125
    },
    {
      "entropy": 5.669254350662231,
      "epoch": 0.6022052586938084,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004988022996814435,
      "loss": 5.6519,
      "mean_token_accuracy": 0.19104730039834977,
      "num_tokens": 1697028.0,
      "step": 2130
    },
    {
      "entropy": 5.5850043296813965,
      "epoch": 0.6036188860616342,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004987916772109897,
      "loss": 5.6018,
      "mean_token_accuracy": 0.1972200408577919,
      "num_tokens": 1701117.0,
      "step": 2135
    },
    {
      "entropy": 5.506943130493164,
      "epoch": 0.60503251342946,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004987810079694591,
      "loss": 5.624,
      "mean_token_accuracy": 0.20141157060861586,
      "num_tokens": 1705157.0,
      "step": 2140
    },
    {
      "entropy": 5.590616226196289,
      "epoch": 0.6064461407972859,
      "grad_norm": 1.375,
      "learning_rate": 0.0004987702919590819,
      "loss": 5.6385,
      "mean_token_accuracy": 0.20066839158535005,
      "num_tokens": 1709476.0,
      "step": 2145
    },
    {
      "entropy": 5.7273688316345215,
      "epoch": 0.6078597681651117,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004987595291820972,
      "loss": 5.6672,
      "mean_token_accuracy": 0.18960647583007811,
      "num_tokens": 1713700.0,
      "step": 2150
    },
    {
      "entropy": 5.628032875061035,
      "epoch": 0.6092733955329375,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004987487196407547,
      "loss": 5.539,
      "mean_token_accuracy": 0.20090654045343398,
      "num_tokens": 1717628.0,
      "step": 2155
    },
    {
      "entropy": 5.513024950027466,
      "epoch": 0.6106870229007634,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004987378633373134,
      "loss": 5.4426,
      "mean_token_accuracy": 0.20594781041145324,
      "num_tokens": 1721618.0,
      "step": 2160
    },
    {
      "entropy": 5.536725044250488,
      "epoch": 0.6121006502685892,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004987269602740423,
      "loss": 5.4352,
      "mean_token_accuracy": 0.21327007859945296,
      "num_tokens": 1725527.0,
      "step": 2165
    },
    {
      "entropy": 5.4519782066345215,
      "epoch": 0.6135142776364151,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004987160104532198,
      "loss": 5.523,
      "mean_token_accuracy": 0.2006515771150589,
      "num_tokens": 1729562.0,
      "step": 2170
    },
    {
      "entropy": 5.55487961769104,
      "epoch": 0.6149279050042409,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004987050138771347,
      "loss": 5.5945,
      "mean_token_accuracy": 0.196965529024601,
      "num_tokens": 1733491.0,
      "step": 2175
    },
    {
      "entropy": 5.517650032043457,
      "epoch": 0.6163415323720667,
      "grad_norm": 1.34375,
      "learning_rate": 0.000498693970548085,
      "loss": 5.4475,
      "mean_token_accuracy": 0.20709712356328963,
      "num_tokens": 1737309.0,
      "step": 2180
    },
    {
      "entropy": 5.535513687133789,
      "epoch": 0.6177551597398926,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004986828804683787,
      "loss": 5.6915,
      "mean_token_accuracy": 0.18559163808822632,
      "num_tokens": 1741514.0,
      "step": 2185
    },
    {
      "entropy": 5.583703565597534,
      "epoch": 0.6191687871077184,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004986717436403337,
      "loss": 5.5444,
      "mean_token_accuracy": 0.20028999596834182,
      "num_tokens": 1745249.0,
      "step": 2190
    },
    {
      "entropy": 5.3787237167358395,
      "epoch": 0.6205824144755443,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986605600662773,
      "loss": 5.5534,
      "mean_token_accuracy": 0.202446810901165,
      "num_tokens": 1749318.0,
      "step": 2195
    },
    {
      "entropy": 5.5711894035339355,
      "epoch": 0.6219960418433701,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004986493297485469,
      "loss": 5.6395,
      "mean_token_accuracy": 0.19081166982650757,
      "num_tokens": 1753419.0,
      "step": 2200
    },
    {
      "entropy": 5.748907613754272,
      "epoch": 0.6234096692111959,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004986380526894896,
      "loss": 5.5461,
      "mean_token_accuracy": 0.2043151706457138,
      "num_tokens": 1757337.0,
      "step": 2205
    },
    {
      "entropy": 5.498734807968139,
      "epoch": 0.6248232965790218,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004986267288914623,
      "loss": 5.5262,
      "mean_token_accuracy": 0.19927159696817398,
      "num_tokens": 1761335.0,
      "step": 2210
    },
    {
      "entropy": 5.368144416809082,
      "epoch": 0.6262369239468476,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004986153583568315,
      "loss": 5.5009,
      "mean_token_accuracy": 0.21026842445135116,
      "num_tokens": 1765370.0,
      "step": 2215
    },
    {
      "entropy": 5.737651777267456,
      "epoch": 0.6276505513146734,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004986039410879737,
      "loss": 5.6564,
      "mean_token_accuracy": 0.1879636362195015,
      "num_tokens": 1769272.0,
      "step": 2220
    },
    {
      "entropy": 5.5789541721344,
      "epoch": 0.6290641786824993,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004985924770872748,
      "loss": 5.538,
      "mean_token_accuracy": 0.20575214624404908,
      "num_tokens": 1773329.0,
      "step": 2225
    },
    {
      "entropy": 5.461138582229614,
      "epoch": 0.6304778060503251,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004985809663571309,
      "loss": 5.4912,
      "mean_token_accuracy": 0.21328796297311783,
      "num_tokens": 1777223.0,
      "step": 2230
    },
    {
      "entropy": 5.552439975738525,
      "epoch": 0.631891433418151,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004985694088999476,
      "loss": 5.5993,
      "mean_token_accuracy": 0.1935050055384636,
      "num_tokens": 1781359.0,
      "step": 2235
    },
    {
      "entropy": 5.4960675716400145,
      "epoch": 0.6333050607859768,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004985578047181404,
      "loss": 5.5384,
      "mean_token_accuracy": 0.1997242197394371,
      "num_tokens": 1785605.0,
      "step": 2240
    },
    {
      "entropy": 5.547979068756104,
      "epoch": 0.6347186881538026,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004985461538141344,
      "loss": 5.5762,
      "mean_token_accuracy": 0.19534142762422563,
      "num_tokens": 1789187.0,
      "step": 2245
    },
    {
      "entropy": 5.586332702636719,
      "epoch": 0.6361323155216285,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004985344561903646,
      "loss": 5.6442,
      "mean_token_accuracy": 0.18766020089387894,
      "num_tokens": 1792971.0,
      "step": 2250
    },
    {
      "entropy": 5.566289854049683,
      "epoch": 0.6375459428894543,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004985227118492758,
      "loss": 5.6188,
      "mean_token_accuracy": 0.2003902390599251,
      "num_tokens": 1796938.0,
      "step": 2255
    },
    {
      "entropy": 5.509633111953735,
      "epoch": 0.6389595702572802,
      "grad_norm": 1.375,
      "learning_rate": 0.0004985109207933224,
      "loss": 5.6196,
      "mean_token_accuracy": 0.19459052234888077,
      "num_tokens": 1801017.0,
      "step": 2260
    },
    {
      "entropy": 5.6678627014160154,
      "epoch": 0.640373197625106,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004984990830249687,
      "loss": 5.6308,
      "mean_token_accuracy": 0.18731064796447755,
      "num_tokens": 1804819.0,
      "step": 2265
    },
    {
      "entropy": 5.608151149749756,
      "epoch": 0.6417868249929318,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004984871985466888,
      "loss": 5.3741,
      "mean_token_accuracy": 0.21539203375577926,
      "num_tokens": 1808570.0,
      "step": 2270
    },
    {
      "entropy": 5.439461183547974,
      "epoch": 0.6432004523607577,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004984752673609663,
      "loss": 5.657,
      "mean_token_accuracy": 0.19133957028388976,
      "num_tokens": 1812762.0,
      "step": 2275
    },
    {
      "entropy": 5.415857172012329,
      "epoch": 0.6446140797285835,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004984632894702949,
      "loss": 5.5041,
      "mean_token_accuracy": 0.19951552748680115,
      "num_tokens": 1816836.0,
      "step": 2280
    },
    {
      "entropy": 5.485863256454468,
      "epoch": 0.6460277070964093,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004984512648771779,
      "loss": 5.4458,
      "mean_token_accuracy": 0.2191944897174835,
      "num_tokens": 1820663.0,
      "step": 2285
    },
    {
      "entropy": 5.403131818771362,
      "epoch": 0.6474413344642352,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004984391935841283,
      "loss": 5.6004,
      "mean_token_accuracy": 0.1996430054306984,
      "num_tokens": 1825018.0,
      "step": 2290
    },
    {
      "entropy": 5.495661401748658,
      "epoch": 0.648854961832061,
      "grad_norm": 1.515625,
      "learning_rate": 0.000498427075593669,
      "loss": 5.513,
      "mean_token_accuracy": 0.1977506771683693,
      "num_tokens": 1829068.0,
      "step": 2295
    },
    {
      "entropy": 5.555873918533325,
      "epoch": 0.650268589199887,
      "grad_norm": 1.375,
      "learning_rate": 0.0004984149109083325,
      "loss": 5.5065,
      "mean_token_accuracy": 0.2082097426056862,
      "num_tokens": 1832971.0,
      "step": 2300
    },
    {
      "entropy": 5.449091529846191,
      "epoch": 0.6516822165677127,
      "grad_norm": 1.375,
      "learning_rate": 0.000498402699530661,
      "loss": 5.5092,
      "mean_token_accuracy": 0.20229289680719376,
      "num_tokens": 1837021.0,
      "step": 2305
    },
    {
      "entropy": 5.43753342628479,
      "epoch": 0.6530958439355385,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000498390441463207,
      "loss": 5.4355,
      "mean_token_accuracy": 0.213383686542511,
      "num_tokens": 1841161.0,
      "step": 2310
    },
    {
      "entropy": 5.483521795272827,
      "epoch": 0.6545094713033645,
      "grad_norm": 1.3125,
      "learning_rate": 0.000498378136708532,
      "loss": 5.5358,
      "mean_token_accuracy": 0.20597505569458008,
      "num_tokens": 1845219.0,
      "step": 2315
    },
    {
      "entropy": 5.548485374450683,
      "epoch": 0.6559230986711903,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004983657852692079,
      "loss": 5.5688,
      "mean_token_accuracy": 0.19546880424022675,
      "num_tokens": 1849101.0,
      "step": 2320
    },
    {
      "entropy": 5.624959945678711,
      "epoch": 0.6573367260390162,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983533871478159,
      "loss": 5.6539,
      "mean_token_accuracy": 0.1901628702878952,
      "num_tokens": 1853160.0,
      "step": 2325
    },
    {
      "entropy": 5.425306749343872,
      "epoch": 0.658750353406842,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000498340942346947,
      "loss": 5.4682,
      "mean_token_accuracy": 0.20828127562999726,
      "num_tokens": 1857002.0,
      "step": 2330
    },
    {
      "entropy": 5.424382829666138,
      "epoch": 0.6601639807746678,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004983284508692025,
      "loss": 5.4225,
      "mean_token_accuracy": 0.20704644918441772,
      "num_tokens": 1860393.0,
      "step": 2335
    },
    {
      "entropy": 5.483639097213745,
      "epoch": 0.6615776081424937,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004983159127171926,
      "loss": 5.5574,
      "mean_token_accuracy": 0.19761840552091597,
      "num_tokens": 1864240.0,
      "step": 2340
    },
    {
      "entropy": 5.416661596298217,
      "epoch": 0.6629912355103195,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004983033278935379,
      "loss": 5.4823,
      "mean_token_accuracy": 0.20184509605169296,
      "num_tokens": 1868158.0,
      "step": 2345
    },
    {
      "entropy": 5.361017656326294,
      "epoch": 0.6644048628781454,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004982906964008686,
      "loss": 5.4491,
      "mean_token_accuracy": 0.2116091802716255,
      "num_tokens": 1872136.0,
      "step": 2350
    },
    {
      "entropy": 5.572614479064941,
      "epoch": 0.6658184902459712,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004982780182418245,
      "loss": 5.5577,
      "mean_token_accuracy": 0.19661566764116287,
      "num_tokens": 1876278.0,
      "step": 2355
    },
    {
      "entropy": 5.607740068435669,
      "epoch": 0.667232117613797,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004982652934190553,
      "loss": 5.5853,
      "mean_token_accuracy": 0.19677325785160066,
      "num_tokens": 1880477.0,
      "step": 2360
    },
    {
      "entropy": 5.323018741607666,
      "epoch": 0.6686457449816229,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004982525219352204,
      "loss": 5.3642,
      "mean_token_accuracy": 0.21515843123197556,
      "num_tokens": 1884636.0,
      "step": 2365
    },
    {
      "entropy": 5.469184637069702,
      "epoch": 0.6700593723494487,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004982397037929891,
      "loss": 5.5606,
      "mean_token_accuracy": 0.20818958729505538,
      "num_tokens": 1888473.0,
      "step": 2370
    },
    {
      "entropy": 5.518539524078369,
      "epoch": 0.6714729997172745,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004982268389950399,
      "loss": 5.4924,
      "mean_token_accuracy": 0.1975516512989998,
      "num_tokens": 1892235.0,
      "step": 2375
    },
    {
      "entropy": 5.512858057022095,
      "epoch": 0.6728866270851004,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000498213927544062,
      "loss": 5.5959,
      "mean_token_accuracy": 0.20426901131868364,
      "num_tokens": 1896294.0,
      "step": 2380
    },
    {
      "entropy": 5.645758581161499,
      "epoch": 0.6743002544529262,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004982009694427535,
      "loss": 5.5961,
      "mean_token_accuracy": 0.196840038895607,
      "num_tokens": 1900517.0,
      "step": 2385
    },
    {
      "entropy": 5.562209510803223,
      "epoch": 0.6757138818207521,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004981879646938225,
      "loss": 5.4792,
      "mean_token_accuracy": 0.21025276482105254,
      "num_tokens": 1904385.0,
      "step": 2390
    },
    {
      "entropy": 5.491833019256592,
      "epoch": 0.6771275091885779,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004981749132999872,
      "loss": 5.6736,
      "mean_token_accuracy": 0.19446842223405839,
      "num_tokens": 1908378.0,
      "step": 2395
    },
    {
      "entropy": 5.65158748626709,
      "epoch": 0.6785411365564037,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004981618152639749,
      "loss": 5.624,
      "mean_token_accuracy": 0.1936953619122505,
      "num_tokens": 1912287.0,
      "step": 2400
    },
    {
      "entropy": 5.613665294647217,
      "epoch": 0.6799547639242296,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004981486705885234,
      "loss": 5.4354,
      "mean_token_accuracy": 0.20573356747627258,
      "num_tokens": 1916489.0,
      "step": 2405
    },
    {
      "entropy": 5.502536630630493,
      "epoch": 0.6813683912920554,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004981354792763796,
      "loss": 5.4324,
      "mean_token_accuracy": 0.20649812370538712,
      "num_tokens": 1920443.0,
      "step": 2410
    },
    {
      "entropy": 5.481741237640381,
      "epoch": 0.6827820186598813,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004981222413303004,
      "loss": 5.5184,
      "mean_token_accuracy": 0.20677600800991058,
      "num_tokens": 1924483.0,
      "step": 2415
    },
    {
      "entropy": 5.492375707626342,
      "epoch": 0.6841956460277071,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004981089567530525,
      "loss": 5.5449,
      "mean_token_accuracy": 0.20652807503938675,
      "num_tokens": 1928307.0,
      "step": 2420
    },
    {
      "entropy": 5.591255378723145,
      "epoch": 0.6856092733955329,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004980956255474124,
      "loss": 5.522,
      "mean_token_accuracy": 0.2046024441719055,
      "num_tokens": 1932279.0,
      "step": 2425
    },
    {
      "entropy": 5.524838542938232,
      "epoch": 0.6870229007633588,
      "grad_norm": 1.359375,
      "learning_rate": 0.000498082247716166,
      "loss": 5.4614,
      "mean_token_accuracy": 0.20246441215276717,
      "num_tokens": 1936420.0,
      "step": 2430
    },
    {
      "entropy": 5.501322221755982,
      "epoch": 0.6884365281311846,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004980688232621095,
      "loss": 5.4592,
      "mean_token_accuracy": 0.20154910832643508,
      "num_tokens": 1940252.0,
      "step": 2435
    },
    {
      "entropy": 5.546559858322143,
      "epoch": 0.6898501554990104,
      "grad_norm": 1.375,
      "learning_rate": 0.0004980553521880482,
      "loss": 5.527,
      "mean_token_accuracy": 0.19189683198928834,
      "num_tokens": 1944346.0,
      "step": 2440
    },
    {
      "entropy": 5.53267879486084,
      "epoch": 0.6912637828668363,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004980418344967978,
      "loss": 5.473,
      "mean_token_accuracy": 0.20286434441804885,
      "num_tokens": 1948330.0,
      "step": 2445
    },
    {
      "entropy": 5.405476427078247,
      "epoch": 0.6926774102346621,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000498028270191183,
      "loss": 5.4914,
      "mean_token_accuracy": 0.20566392093896865,
      "num_tokens": 1952489.0,
      "step": 2450
    },
    {
      "entropy": 5.5635826110839846,
      "epoch": 0.694091037602488,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004980146592740391,
      "loss": 5.475,
      "mean_token_accuracy": 0.20497844517230987,
      "num_tokens": 1956850.0,
      "step": 2455
    },
    {
      "entropy": 5.649666452407837,
      "epoch": 0.6955046649703138,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004980010017482104,
      "loss": 5.549,
      "mean_token_accuracy": 0.19773884862661362,
      "num_tokens": 1960901.0,
      "step": 2460
    },
    {
      "entropy": 5.498390483856201,
      "epoch": 0.6969182923381396,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004979872976165513,
      "loss": 5.546,
      "mean_token_accuracy": 0.20099450945854186,
      "num_tokens": 1964730.0,
      "step": 2465
    },
    {
      "entropy": 5.774654197692871,
      "epoch": 0.6983319197059655,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000497973546881926,
      "loss": 5.5816,
      "mean_token_accuracy": 0.1982674315571785,
      "num_tokens": 1968489.0,
      "step": 2470
    },
    {
      "entropy": 5.673654413223266,
      "epoch": 0.6997455470737913,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004979597495472081,
      "loss": 5.682,
      "mean_token_accuracy": 0.18073946982622147,
      "num_tokens": 1972536.0,
      "step": 2475
    },
    {
      "entropy": 5.651551675796509,
      "epoch": 0.7011591744416172,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004979459056152814,
      "loss": 5.6503,
      "mean_token_accuracy": 0.18998856246471404,
      "num_tokens": 1976410.0,
      "step": 2480
    },
    {
      "entropy": 5.6095246315002445,
      "epoch": 0.702572801809443,
      "grad_norm": 1.421875,
      "learning_rate": 0.000497932015089039,
      "loss": 5.4748,
      "mean_token_accuracy": 0.2007334977388382,
      "num_tokens": 1980294.0,
      "step": 2485
    },
    {
      "entropy": 5.490608882904053,
      "epoch": 0.7039864291772688,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004979180779713841,
      "loss": 5.4416,
      "mean_token_accuracy": 0.20976191759109497,
      "num_tokens": 1984096.0,
      "step": 2490
    },
    {
      "entropy": 5.545828485488892,
      "epoch": 0.7054000565450947,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004979040942652294,
      "loss": 5.4237,
      "mean_token_accuracy": 0.20762723237276076,
      "num_tokens": 1987714.0,
      "step": 2495
    },
    {
      "entropy": 5.487170839309693,
      "epoch": 0.7068136839129205,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004978900639734974,
      "loss": 5.5378,
      "mean_token_accuracy": 0.1992262378334999,
      "num_tokens": 1991695.0,
      "step": 2500
    },
    {
      "entropy": 5.654662799835205,
      "epoch": 0.7082273112807463,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004978759870991203,
      "loss": 5.6352,
      "mean_token_accuracy": 0.1820220798254013,
      "num_tokens": 1995735.0,
      "step": 2505
    },
    {
      "entropy": 5.492784404754639,
      "epoch": 0.7096409386485723,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004978618636450401,
      "loss": 5.4504,
      "mean_token_accuracy": 0.21076122373342515,
      "num_tokens": 1999642.0,
      "step": 2510
    },
    {
      "entropy": 5.40449481010437,
      "epoch": 0.711054566016398,
      "grad_norm": 1.5,
      "learning_rate": 0.0004978476936142085,
      "loss": 5.5149,
      "mean_token_accuracy": 0.20163996070623397,
      "num_tokens": 2003374.0,
      "step": 2515
    },
    {
      "entropy": 5.44085955619812,
      "epoch": 0.712468193384224,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000497833477009587,
      "loss": 5.3847,
      "mean_token_accuracy": 0.21002986133098603,
      "num_tokens": 2007286.0,
      "step": 2520
    },
    {
      "entropy": 5.440703868865967,
      "epoch": 0.7138818207520498,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004978192138341468,
      "loss": 5.417,
      "mean_token_accuracy": 0.19510221779346465,
      "num_tokens": 2011245.0,
      "step": 2525
    },
    {
      "entropy": 5.36399245262146,
      "epoch": 0.7152954481198756,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004978049040908688,
      "loss": 5.4004,
      "mean_token_accuracy": 0.21078222543001174,
      "num_tokens": 2015535.0,
      "step": 2530
    },
    {
      "entropy": 5.599104022979736,
      "epoch": 0.7167090754877015,
      "grad_norm": 1.375,
      "learning_rate": 0.0004977905477827437,
      "loss": 5.568,
      "mean_token_accuracy": 0.19067409485578538,
      "num_tokens": 2019531.0,
      "step": 2535
    },
    {
      "entropy": 5.504277086257934,
      "epoch": 0.7181227028555273,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004977761449127716,
      "loss": 5.464,
      "mean_token_accuracy": 0.19502160847187042,
      "num_tokens": 2023665.0,
      "step": 2540
    },
    {
      "entropy": 5.451600646972656,
      "epoch": 0.7195363302233532,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000497761695483963,
      "loss": 5.4159,
      "mean_token_accuracy": 0.2057444855570793,
      "num_tokens": 2027718.0,
      "step": 2545
    },
    {
      "entropy": 5.519393682479858,
      "epoch": 0.720949957591179,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004977471994993375,
      "loss": 5.5453,
      "mean_token_accuracy": 0.1898050606250763,
      "num_tokens": 2031752.0,
      "step": 2550
    },
    {
      "entropy": 5.537784576416016,
      "epoch": 0.7223635849590048,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004977326569619247,
      "loss": 5.5663,
      "mean_token_accuracy": 0.2046726256608963,
      "num_tokens": 2035678.0,
      "step": 2555
    },
    {
      "entropy": 5.5787208557128904,
      "epoch": 0.7237772123268307,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004977180678747639,
      "loss": 5.4916,
      "mean_token_accuracy": 0.1908958524465561,
      "num_tokens": 2039659.0,
      "step": 2560
    },
    {
      "entropy": 5.4686425685882565,
      "epoch": 0.7251908396946565,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004977034322409043,
      "loss": 5.3982,
      "mean_token_accuracy": 0.21743353605270385,
      "num_tokens": 2043451.0,
      "step": 2565
    },
    {
      "entropy": 5.432757949829101,
      "epoch": 0.7266044670624824,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004976887500634044,
      "loss": 5.4393,
      "mean_token_accuracy": 0.20736369043588637,
      "num_tokens": 2047388.0,
      "step": 2570
    },
    {
      "entropy": 5.428971529006958,
      "epoch": 0.7280180944303082,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004976740213453329,
      "loss": 5.4046,
      "mean_token_accuracy": 0.20552811473608018,
      "num_tokens": 2051153.0,
      "step": 2575
    },
    {
      "entropy": 5.363815879821777,
      "epoch": 0.729431721798134,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004976592460897678,
      "loss": 5.3929,
      "mean_token_accuracy": 0.21189287304878235,
      "num_tokens": 2055150.0,
      "step": 2580
    },
    {
      "entropy": 5.440291070938111,
      "epoch": 0.7308453491659599,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004976444242997973,
      "loss": 5.3387,
      "mean_token_accuracy": 0.20777003467082977,
      "num_tokens": 2059065.0,
      "step": 2585
    },
    {
      "entropy": 5.4319572925567625,
      "epoch": 0.7322589765337857,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004976295559785187,
      "loss": 5.4752,
      "mean_token_accuracy": 0.19871820211410524,
      "num_tokens": 2063137.0,
      "step": 2590
    },
    {
      "entropy": 5.506852388381958,
      "epoch": 0.7336726039016115,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004976146411290398,
      "loss": 5.6009,
      "mean_token_accuracy": 0.19600762277841569,
      "num_tokens": 2067112.0,
      "step": 2595
    },
    {
      "entropy": 5.53849458694458,
      "epoch": 0.7350862312694374,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004975996797544774,
      "loss": 5.555,
      "mean_token_accuracy": 0.20375431180000306,
      "num_tokens": 2071336.0,
      "step": 2600
    },
    {
      "entropy": 5.500842618942261,
      "epoch": 0.7364998586372632,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004975846718579585,
      "loss": 5.3818,
      "mean_token_accuracy": 0.20658586025238038,
      "num_tokens": 2075072.0,
      "step": 2605
    },
    {
      "entropy": 5.413701963424683,
      "epoch": 0.7379134860050891,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004975696174426196,
      "loss": 5.4469,
      "mean_token_accuracy": 0.20455588698387145,
      "num_tokens": 2079177.0,
      "step": 2610
    },
    {
      "entropy": 5.6096607685089115,
      "epoch": 0.7393271133729149,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004975545165116069,
      "loss": 5.5943,
      "mean_token_accuracy": 0.19681017547845842,
      "num_tokens": 2083010.0,
      "step": 2615
    },
    {
      "entropy": 5.4185761451721195,
      "epoch": 0.7407407407407407,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004975393690680765,
      "loss": 5.4683,
      "mean_token_accuracy": 0.20828914493322373,
      "num_tokens": 2087225.0,
      "step": 2620
    },
    {
      "entropy": 5.488922500610352,
      "epoch": 0.7421543681085666,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004975241751151942,
      "loss": 5.4228,
      "mean_token_accuracy": 0.19531789422035217,
      "num_tokens": 2091053.0,
      "step": 2625
    },
    {
      "entropy": 5.570777654647827,
      "epoch": 0.7435679954763924,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004975089346561354,
      "loss": 5.5811,
      "mean_token_accuracy": 0.19731894731521607,
      "num_tokens": 2095152.0,
      "step": 2630
    },
    {
      "entropy": 5.477892971038818,
      "epoch": 0.7449816228442183,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974936476940851,
      "loss": 5.4442,
      "mean_token_accuracy": 0.20835249572992326,
      "num_tokens": 2099282.0,
      "step": 2635
    },
    {
      "entropy": 5.4942488193511965,
      "epoch": 0.7463952502120441,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004974783142322383,
      "loss": 5.5205,
      "mean_token_accuracy": 0.19406803846359252,
      "num_tokens": 2103358.0,
      "step": 2640
    },
    {
      "entropy": 5.524212217330932,
      "epoch": 0.7478088775798699,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004974629342737995,
      "loss": 5.5178,
      "mean_token_accuracy": 0.2022445544600487,
      "num_tokens": 2107180.0,
      "step": 2645
    },
    {
      "entropy": 5.381645727157593,
      "epoch": 0.7492225049476958,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004974475078219833,
      "loss": 5.429,
      "mean_token_accuracy": 0.20344440937042235,
      "num_tokens": 2111002.0,
      "step": 2650
    },
    {
      "entropy": 5.47999210357666,
      "epoch": 0.7506361323155216,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004974320348800134,
      "loss": 5.3441,
      "mean_token_accuracy": 0.19606085270643234,
      "num_tokens": 2114925.0,
      "step": 2655
    },
    {
      "entropy": 5.530167818069458,
      "epoch": 0.7520497596833474,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004974165154511237,
      "loss": 5.4202,
      "mean_token_accuracy": 0.20337026417255402,
      "num_tokens": 2119039.0,
      "step": 2660
    },
    {
      "entropy": 5.519847202301025,
      "epoch": 0.7534633870511733,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004974009495385578,
      "loss": 5.4474,
      "mean_token_accuracy": 0.20683934539556503,
      "num_tokens": 2123215.0,
      "step": 2665
    },
    {
      "entropy": 5.484922122955322,
      "epoch": 0.7548770144189991,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004973853371455686,
      "loss": 5.4241,
      "mean_token_accuracy": 0.19822932183742523,
      "num_tokens": 2127013.0,
      "step": 2670
    },
    {
      "entropy": 5.4893275737762455,
      "epoch": 0.756290641786825,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004973696782754191,
      "loss": 5.533,
      "mean_token_accuracy": 0.19578728973865508,
      "num_tokens": 2131310.0,
      "step": 2675
    },
    {
      "entropy": 5.470749950408935,
      "epoch": 0.7577042691546508,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004973539729313819,
      "loss": 5.4614,
      "mean_token_accuracy": 0.20932352244853974,
      "num_tokens": 2135203.0,
      "step": 2680
    },
    {
      "entropy": 5.344170427322387,
      "epoch": 0.7591178965224766,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004973382211167394,
      "loss": 5.4245,
      "mean_token_accuracy": 0.2103756472468376,
      "num_tokens": 2139212.0,
      "step": 2685
    },
    {
      "entropy": 5.444655799865723,
      "epoch": 0.7605315238903025,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004973224228347837,
      "loss": 5.3908,
      "mean_token_accuracy": 0.20362839549779893,
      "num_tokens": 2143215.0,
      "step": 2690
    },
    {
      "entropy": 5.442157936096192,
      "epoch": 0.7619451512581283,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004973065780888164,
      "loss": 5.4118,
      "mean_token_accuracy": 0.2113314688205719,
      "num_tokens": 2146984.0,
      "step": 2695
    },
    {
      "entropy": 5.49960618019104,
      "epoch": 0.7633587786259542,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004972906868821489,
      "loss": 5.4596,
      "mean_token_accuracy": 0.21158111989498138,
      "num_tokens": 2151107.0,
      "step": 2700
    },
    {
      "entropy": 5.425730562210083,
      "epoch": 0.76477240599378,
      "grad_norm": 1.5,
      "learning_rate": 0.0004972747492181024,
      "loss": 5.4754,
      "mean_token_accuracy": 0.19603095650672914,
      "num_tokens": 2155036.0,
      "step": 2705
    },
    {
      "entropy": 5.462151288986206,
      "epoch": 0.7661860333616058,
      "grad_norm": 1.40625,
      "learning_rate": 0.000497258765100008,
      "loss": 5.4598,
      "mean_token_accuracy": 0.19829021096229554,
      "num_tokens": 2159043.0,
      "step": 2710
    },
    {
      "entropy": 5.47222375869751,
      "epoch": 0.7675996607294318,
      "grad_norm": 1.375,
      "learning_rate": 0.000497242734531206,
      "loss": 5.4575,
      "mean_token_accuracy": 0.19492304623126983,
      "num_tokens": 2163089.0,
      "step": 2715
    },
    {
      "entropy": 5.415404176712036,
      "epoch": 0.7690132880972576,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004972266575150468,
      "loss": 5.3975,
      "mean_token_accuracy": 0.200945146381855,
      "num_tokens": 2167089.0,
      "step": 2720
    },
    {
      "entropy": 5.451752805709839,
      "epoch": 0.7704269154650834,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004972105340548903,
      "loss": 5.4645,
      "mean_token_accuracy": 0.20999046713113784,
      "num_tokens": 2170899.0,
      "step": 2725
    },
    {
      "entropy": 5.509784698486328,
      "epoch": 0.7718405428329093,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971943641541065,
      "loss": 5.4759,
      "mean_token_accuracy": 0.1974717140197754,
      "num_tokens": 2174820.0,
      "step": 2730
    },
    {
      "entropy": 5.537976455688477,
      "epoch": 0.7732541702007351,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004971781478160743,
      "loss": 5.4336,
      "mean_token_accuracy": 0.21598929315805435,
      "num_tokens": 2179240.0,
      "step": 2735
    },
    {
      "entropy": 5.4725525856018065,
      "epoch": 0.774667797568561,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004971618850441833,
      "loss": 5.5562,
      "mean_token_accuracy": 0.20348277688026428,
      "num_tokens": 2183302.0,
      "step": 2740
    },
    {
      "entropy": 5.600423669815063,
      "epoch": 0.7760814249363868,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004971455758418322,
      "loss": 5.502,
      "mean_token_accuracy": 0.187813638150692,
      "num_tokens": 2187397.0,
      "step": 2745
    },
    {
      "entropy": 5.394171714782715,
      "epoch": 0.7774950523042126,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004971292202124293,
      "loss": 5.4385,
      "mean_token_accuracy": 0.20303665846586227,
      "num_tokens": 2191475.0,
      "step": 2750
    },
    {
      "entropy": 5.392899370193481,
      "epoch": 0.7789086796720385,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004971128181593931,
      "loss": 5.5618,
      "mean_token_accuracy": 0.20110445618629455,
      "num_tokens": 2195952.0,
      "step": 2755
    },
    {
      "entropy": 5.555618238449097,
      "epoch": 0.7803223070398643,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004970963696861512,
      "loss": 5.4539,
      "mean_token_accuracy": 0.1939147248864174,
      "num_tokens": 2200555.0,
      "step": 2760
    },
    {
      "entropy": 5.540684795379638,
      "epoch": 0.7817359344076902,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004970798747961417,
      "loss": 5.4977,
      "mean_token_accuracy": 0.19708138406276704,
      "num_tokens": 2204477.0,
      "step": 2765
    },
    {
      "entropy": 5.503489255905151,
      "epoch": 0.783149561775516,
      "grad_norm": 1.625,
      "learning_rate": 0.0004970633334928116,
      "loss": 5.5269,
      "mean_token_accuracy": 0.1929205074906349,
      "num_tokens": 2208252.0,
      "step": 2770
    },
    {
      "entropy": 5.565472221374511,
      "epoch": 0.7845631891433418,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004970467457796179,
      "loss": 5.4232,
      "mean_token_accuracy": 0.1986708164215088,
      "num_tokens": 2211933.0,
      "step": 2775
    },
    {
      "entropy": 5.3583128452301025,
      "epoch": 0.7859768165111677,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004970301116600274,
      "loss": 5.4395,
      "mean_token_accuracy": 0.20127756148576736,
      "num_tokens": 2215966.0,
      "step": 2780
    },
    {
      "entropy": 5.3603196144104,
      "epoch": 0.7873904438789935,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004970134311375168,
      "loss": 5.4652,
      "mean_token_accuracy": 0.19634118378162385,
      "num_tokens": 2219976.0,
      "step": 2785
    },
    {
      "entropy": 5.591530513763428,
      "epoch": 0.7888040712468194,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004969967042155717,
      "loss": 5.5805,
      "mean_token_accuracy": 0.19613414853811265,
      "num_tokens": 2224236.0,
      "step": 2790
    },
    {
      "entropy": 5.541605424880982,
      "epoch": 0.7902176986146452,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004969799308976883,
      "loss": 5.3779,
      "mean_token_accuracy": 0.2137235939502716,
      "num_tokens": 2228137.0,
      "step": 2795
    },
    {
      "entropy": 5.225785923004151,
      "epoch": 0.791631325982471,
      "grad_norm": 1.46875,
      "learning_rate": 0.000496963111187372,
      "loss": 5.2868,
      "mean_token_accuracy": 0.20965032577514647,
      "num_tokens": 2232186.0,
      "step": 2800
    },
    {
      "entropy": 5.361642456054687,
      "epoch": 0.7930449533502969,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004969462450881382,
      "loss": 5.4373,
      "mean_token_accuracy": 0.20115301758050919,
      "num_tokens": 2236183.0,
      "step": 2805
    },
    {
      "entropy": 5.471748876571655,
      "epoch": 0.7944585807181227,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004969293326035115,
      "loss": 5.394,
      "mean_token_accuracy": 0.20902985483407974,
      "num_tokens": 2240087.0,
      "step": 2810
    },
    {
      "entropy": 5.421245145797729,
      "epoch": 0.7958722080859485,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004969123737370267,
      "loss": 5.4525,
      "mean_token_accuracy": 0.21186718791723252,
      "num_tokens": 2243905.0,
      "step": 2815
    },
    {
      "entropy": 5.429970359802246,
      "epoch": 0.7972858354537744,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004968953684922281,
      "loss": 5.4214,
      "mean_token_accuracy": 0.20257548689842225,
      "num_tokens": 2248190.0,
      "step": 2820
    },
    {
      "entropy": 5.560571479797363,
      "epoch": 0.7986994628216002,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004968783168726697,
      "loss": 5.5068,
      "mean_token_accuracy": 0.19931748509407043,
      "num_tokens": 2252375.0,
      "step": 2825
    },
    {
      "entropy": 5.377751159667969,
      "epoch": 0.8001130901894261,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004968612188819149,
      "loss": 5.3476,
      "mean_token_accuracy": 0.21681973487138748,
      "num_tokens": 2256152.0,
      "step": 2830
    },
    {
      "entropy": 5.319108629226685,
      "epoch": 0.8015267175572519,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004968440745235375,
      "loss": 5.4944,
      "mean_token_accuracy": 0.19941969364881515,
      "num_tokens": 2260023.0,
      "step": 2835
    },
    {
      "entropy": 5.391258811950683,
      "epoch": 0.8029403449250777,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004968268838011203,
      "loss": 5.3394,
      "mean_token_accuracy": 0.2063663497567177,
      "num_tokens": 2264116.0,
      "step": 2840
    },
    {
      "entropy": 5.403754138946534,
      "epoch": 0.8043539722929036,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004968096467182561,
      "loss": 5.4178,
      "mean_token_accuracy": 0.21182752698659896,
      "num_tokens": 2268113.0,
      "step": 2845
    },
    {
      "entropy": 5.342642307281494,
      "epoch": 0.8057675996607294,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004967923632785473,
      "loss": 5.3794,
      "mean_token_accuracy": 0.2177852764725685,
      "num_tokens": 2271781.0,
      "step": 2850
    },
    {
      "entropy": 5.419791269302368,
      "epoch": 0.8071812270285553,
      "grad_norm": 1.375,
      "learning_rate": 0.0004967750334856063,
      "loss": 5.3684,
      "mean_token_accuracy": 0.2125452294945717,
      "num_tokens": 2275662.0,
      "step": 2855
    },
    {
      "entropy": 5.459193181991577,
      "epoch": 0.8085948543963811,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004967576573430546,
      "loss": 5.4551,
      "mean_token_accuracy": 0.20523227453231813,
      "num_tokens": 2279753.0,
      "step": 2860
    },
    {
      "entropy": 5.426430130004883,
      "epoch": 0.8100084817642069,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004967402348545237,
      "loss": 5.4358,
      "mean_token_accuracy": 0.2110658660531044,
      "num_tokens": 2283684.0,
      "step": 2865
    },
    {
      "entropy": 5.566097354888916,
      "epoch": 0.8114221091320328,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004967227660236551,
      "loss": 5.5038,
      "mean_token_accuracy": 0.189190636575222,
      "num_tokens": 2287894.0,
      "step": 2870
    },
    {
      "entropy": 5.424854040145874,
      "epoch": 0.8128357364998586,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004967052508540995,
      "loss": 5.3023,
      "mean_token_accuracy": 0.2083859771490097,
      "num_tokens": 2291833.0,
      "step": 2875
    },
    {
      "entropy": 5.304775667190552,
      "epoch": 0.8142493638676844,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004966876893495174,
      "loss": 5.3373,
      "mean_token_accuracy": 0.2097040370106697,
      "num_tokens": 2295800.0,
      "step": 2880
    },
    {
      "entropy": 5.382280778884888,
      "epoch": 0.8156629912355103,
      "grad_norm": 1.484375,
      "learning_rate": 0.000496670081513579,
      "loss": 5.4374,
      "mean_token_accuracy": 0.1975208267569542,
      "num_tokens": 2299704.0,
      "step": 2885
    },
    {
      "entropy": 5.510290908813476,
      "epoch": 0.8170766186033361,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004966524273499644,
      "loss": 5.5089,
      "mean_token_accuracy": 0.19660034626722336,
      "num_tokens": 2303646.0,
      "step": 2890
    },
    {
      "entropy": 5.411760520935059,
      "epoch": 0.818490245971162,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004966347268623632,
      "loss": 5.3326,
      "mean_token_accuracy": 0.2152049422264099,
      "num_tokens": 2307587.0,
      "step": 2895
    },
    {
      "entropy": 5.396860408782959,
      "epoch": 0.8199038733389878,
      "grad_norm": 1.5,
      "learning_rate": 0.0004966169800544746,
      "loss": 5.4141,
      "mean_token_accuracy": 0.20583823025226594,
      "num_tokens": 2311372.0,
      "step": 2900
    },
    {
      "entropy": 5.4382915019989015,
      "epoch": 0.8213175007068136,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004965991869300076,
      "loss": 5.5091,
      "mean_token_accuracy": 0.19858545660972596,
      "num_tokens": 2315804.0,
      "step": 2905
    },
    {
      "entropy": 5.426002740859985,
      "epoch": 0.8227311280746396,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004965813474926809,
      "loss": 5.4399,
      "mean_token_accuracy": 0.19376754462718965,
      "num_tokens": 2320095.0,
      "step": 2910
    },
    {
      "entropy": 5.447252321243286,
      "epoch": 0.8241447554424653,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004965634617462227,
      "loss": 5.5984,
      "mean_token_accuracy": 0.1918902039527893,
      "num_tokens": 2324324.0,
      "step": 2915
    },
    {
      "entropy": 5.569087553024292,
      "epoch": 0.8255583828102913,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004965455296943712,
      "loss": 5.5009,
      "mean_token_accuracy": 0.20151453763246535,
      "num_tokens": 2328419.0,
      "step": 2920
    },
    {
      "entropy": 5.494377708435058,
      "epoch": 0.8269720101781171,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004965275513408739,
      "loss": 5.4344,
      "mean_token_accuracy": 0.20523380637168884,
      "num_tokens": 2332338.0,
      "step": 2925
    },
    {
      "entropy": 5.233751964569092,
      "epoch": 0.8283856375459429,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004965095266894884,
      "loss": 5.3563,
      "mean_token_accuracy": 0.20502769947052002,
      "num_tokens": 2336139.0,
      "step": 2930
    },
    {
      "entropy": 5.43345012664795,
      "epoch": 0.8297992649137688,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004964914557439815,
      "loss": 5.5184,
      "mean_token_accuracy": 0.19930208325386048,
      "num_tokens": 2340222.0,
      "step": 2935
    },
    {
      "entropy": 5.627100038528442,
      "epoch": 0.8312128922815946,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004964733385081299,
      "loss": 5.5587,
      "mean_token_accuracy": 0.18537889569997787,
      "num_tokens": 2343895.0,
      "step": 2940
    },
    {
      "entropy": 5.414453506469727,
      "epoch": 0.8326265196494204,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004964551749857203,
      "loss": 5.2786,
      "mean_token_accuracy": 0.21607314348220824,
      "num_tokens": 2347924.0,
      "step": 2945
    },
    {
      "entropy": 5.469025421142578,
      "epoch": 0.8340401470172463,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004964369651805484,
      "loss": 5.568,
      "mean_token_accuracy": 0.191763836145401,
      "num_tokens": 2352261.0,
      "step": 2950
    },
    {
      "entropy": 5.4523872375488285,
      "epoch": 0.8354537743850721,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004964187090964202,
      "loss": 5.4866,
      "mean_token_accuracy": 0.20151374191045762,
      "num_tokens": 2356127.0,
      "step": 2955
    },
    {
      "entropy": 5.386123514175415,
      "epoch": 0.836867401752898,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004964004067371509,
      "loss": 5.3526,
      "mean_token_accuracy": 0.21623794734477997,
      "num_tokens": 2360318.0,
      "step": 2960
    },
    {
      "entropy": 5.479382753372192,
      "epoch": 0.8382810291207238,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004963820581065659,
      "loss": 5.4538,
      "mean_token_accuracy": 0.19994483292102813,
      "num_tokens": 2364421.0,
      "step": 2965
    },
    {
      "entropy": 5.529749536514283,
      "epoch": 0.8396946564885496,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004963636632084996,
      "loss": 5.4253,
      "mean_token_accuracy": 0.2110798642039299,
      "num_tokens": 2368360.0,
      "step": 2970
    },
    {
      "entropy": 5.412200593948365,
      "epoch": 0.8411082838563755,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004963452220467966,
      "loss": 5.3276,
      "mean_token_accuracy": 0.2055237829685211,
      "num_tokens": 2372233.0,
      "step": 2975
    },
    {
      "entropy": 5.427595901489258,
      "epoch": 0.8425219112242013,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004963267346253108,
      "loss": 5.3143,
      "mean_token_accuracy": 0.21237905770540239,
      "num_tokens": 2376175.0,
      "step": 2980
    },
    {
      "entropy": 5.3335240364074705,
      "epoch": 0.8439355385920272,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004963082009479063,
      "loss": 5.4244,
      "mean_token_accuracy": 0.20374088287353515,
      "num_tokens": 2380052.0,
      "step": 2985
    },
    {
      "entropy": 5.425597667694092,
      "epoch": 0.845349165959853,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004962896210184563,
      "loss": 5.5381,
      "mean_token_accuracy": 0.1938463866710663,
      "num_tokens": 2384273.0,
      "step": 2990
    },
    {
      "entropy": 5.479777717590332,
      "epoch": 0.8467627933276788,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004962709948408439,
      "loss": 5.4082,
      "mean_token_accuracy": 0.20560099333524703,
      "num_tokens": 2388113.0,
      "step": 2995
    },
    {
      "entropy": 5.367301034927368,
      "epoch": 0.8481764206955047,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004962523224189618,
      "loss": 5.5472,
      "mean_token_accuracy": 0.1989892229437828,
      "num_tokens": 2392345.0,
      "step": 3000
    },
    {
      "epoch": 0.8481764206955047,
      "eval_entropy": 5.264917620693344,
      "eval_loss": 5.590817928314209,
      "eval_mean_token_accuracy": 0.20367044986646113,
      "eval_num_tokens": 2392345.0,
      "eval_runtime": 6.7142,
      "eval_samples_per_second": 1680.625,
      "eval_steps_per_second": 210.153,
      "step": 3000
    },
    {
      "entropy": 5.466040372848511,
      "epoch": 0.8495900480633305,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004962336037567126,
      "loss": 5.3321,
      "mean_token_accuracy": 0.21223577558994294,
      "num_tokens": 2396304.0,
      "step": 3005
    },
    {
      "entropy": 5.293815565109253,
      "epoch": 0.8510036754311563,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004962148388580083,
      "loss": 5.3477,
      "mean_token_accuracy": 0.20822013765573502,
      "num_tokens": 2400137.0,
      "step": 3010
    },
    {
      "entropy": 5.370921182632446,
      "epoch": 0.8524173027989822,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004961960277267705,
      "loss": 5.401,
      "mean_token_accuracy": 0.20404562950134278,
      "num_tokens": 2404047.0,
      "step": 3015
    },
    {
      "entropy": 5.294777250289917,
      "epoch": 0.853830930166808,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004961771703669308,
      "loss": 5.4032,
      "mean_token_accuracy": 0.2028578281402588,
      "num_tokens": 2408212.0,
      "step": 3020
    },
    {
      "entropy": 5.4464034080505375,
      "epoch": 0.8552445575346339,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004961582667824303,
      "loss": 5.4155,
      "mean_token_accuracy": 0.2065512552857399,
      "num_tokens": 2412073.0,
      "step": 3025
    },
    {
      "entropy": 5.523953914642334,
      "epoch": 0.8566581849024597,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004961393169772196,
      "loss": 5.4915,
      "mean_token_accuracy": 0.20319596529006959,
      "num_tokens": 2416432.0,
      "step": 3030
    },
    {
      "entropy": 5.422474098205567,
      "epoch": 0.8580718122702855,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004961203209552591,
      "loss": 5.4431,
      "mean_token_accuracy": 0.20577302873134612,
      "num_tokens": 2420331.0,
      "step": 3035
    },
    {
      "entropy": 5.391149473190308,
      "epoch": 0.8594854396381114,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004961012787205188,
      "loss": 5.4071,
      "mean_token_accuracy": 0.1985209420323372,
      "num_tokens": 2424082.0,
      "step": 3040
    },
    {
      "entropy": 5.283867502212525,
      "epoch": 0.8608990670059372,
      "grad_norm": 1.5,
      "learning_rate": 0.0004960821902769786,
      "loss": 5.3261,
      "mean_token_accuracy": 0.21183746606111525,
      "num_tokens": 2427702.0,
      "step": 3045
    },
    {
      "entropy": 5.409480476379395,
      "epoch": 0.8623126943737631,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004960630556286278,
      "loss": 5.3831,
      "mean_token_accuracy": 0.20391480773687362,
      "num_tokens": 2431589.0,
      "step": 3050
    },
    {
      "entropy": 5.390744924545288,
      "epoch": 0.8637263217415889,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004960438747794654,
      "loss": 5.4412,
      "mean_token_accuracy": 0.20160709470510482,
      "num_tokens": 2435741.0,
      "step": 3055
    },
    {
      "entropy": 5.379309225082397,
      "epoch": 0.8651399491094147,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004960246477334998,
      "loss": 5.5135,
      "mean_token_accuracy": 0.21191916167736052,
      "num_tokens": 2439529.0,
      "step": 3060
    },
    {
      "entropy": 5.378274202346802,
      "epoch": 0.8665535764772406,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004960053744947499,
      "loss": 5.288,
      "mean_token_accuracy": 0.21649111956357955,
      "num_tokens": 2443511.0,
      "step": 3065
    },
    {
      "entropy": 5.4915688037872314,
      "epoch": 0.8679672038450664,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004959860550672433,
      "loss": 5.5025,
      "mean_token_accuracy": 0.2066110298037529,
      "num_tokens": 2447369.0,
      "step": 3070
    },
    {
      "entropy": 5.438264989852906,
      "epoch": 0.8693808312128923,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004959666894550177,
      "loss": 5.531,
      "mean_token_accuracy": 0.20990222245454787,
      "num_tokens": 2451458.0,
      "step": 3075
    },
    {
      "entropy": 5.395460748672486,
      "epoch": 0.8707944585807181,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004959472776621205,
      "loss": 5.3413,
      "mean_token_accuracy": 0.20516567826271057,
      "num_tokens": 2455681.0,
      "step": 3080
    },
    {
      "entropy": 5.368040418624878,
      "epoch": 0.8722080859485439,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004959278196926085,
      "loss": 5.3232,
      "mean_token_accuracy": 0.22193797379732133,
      "num_tokens": 2459887.0,
      "step": 3085
    },
    {
      "entropy": 5.292930221557617,
      "epoch": 0.8736217133163698,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004959083155505483,
      "loss": 5.3074,
      "mean_token_accuracy": 0.22182201892137526,
      "num_tokens": 2463837.0,
      "step": 3090
    },
    {
      "entropy": 5.449946117401123,
      "epoch": 0.8750353406841956,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004958887652400163,
      "loss": 5.4493,
      "mean_token_accuracy": 0.19624976366758345,
      "num_tokens": 2467714.0,
      "step": 3095
    },
    {
      "entropy": 5.3686493873596195,
      "epoch": 0.8764489680520214,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004958691687650983,
      "loss": 5.4122,
      "mean_token_accuracy": 0.2004915341734886,
      "num_tokens": 2472016.0,
      "step": 3100
    },
    {
      "entropy": 5.447443628311158,
      "epoch": 0.8778625954198473,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004958495261298898,
      "loss": 5.3433,
      "mean_token_accuracy": 0.21415164917707444,
      "num_tokens": 2475898.0,
      "step": 3105
    },
    {
      "entropy": 5.443156099319458,
      "epoch": 0.8792762227876731,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004958298373384961,
      "loss": 5.4545,
      "mean_token_accuracy": 0.2098313644528389,
      "num_tokens": 2480144.0,
      "step": 3110
    },
    {
      "entropy": 5.308551979064942,
      "epoch": 0.880689850155499,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004958101023950319,
      "loss": 5.3791,
      "mean_token_accuracy": 0.2096642553806305,
      "num_tokens": 2484127.0,
      "step": 3115
    },
    {
      "entropy": 5.341952514648438,
      "epoch": 0.8821034775233249,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004957903213036217,
      "loss": 5.4112,
      "mean_token_accuracy": 0.21089784950017929,
      "num_tokens": 2488107.0,
      "step": 3120
    },
    {
      "entropy": 5.266310405731201,
      "epoch": 0.8835171048911507,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004957704940683998,
      "loss": 5.2934,
      "mean_token_accuracy": 0.2120772823691368,
      "num_tokens": 2491981.0,
      "step": 3125
    },
    {
      "entropy": 5.336669969558716,
      "epoch": 0.8849307322589766,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004957506206935097,
      "loss": 5.2739,
      "mean_token_accuracy": 0.22724443972110747,
      "num_tokens": 2496240.0,
      "step": 3130
    },
    {
      "entropy": 5.347661018371582,
      "epoch": 0.8863443596268024,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004957307011831051,
      "loss": 5.2222,
      "mean_token_accuracy": 0.21300769001245498,
      "num_tokens": 2500175.0,
      "step": 3135
    },
    {
      "entropy": 5.4087928295135494,
      "epoch": 0.8877579869946283,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004957107355413489,
      "loss": 5.3914,
      "mean_token_accuracy": 0.2105944722890854,
      "num_tokens": 2503950.0,
      "step": 3140
    },
    {
      "entropy": 5.3735452651977536,
      "epoch": 0.8891716143624541,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004956907237724137,
      "loss": 5.421,
      "mean_token_accuracy": 0.20434045046567917,
      "num_tokens": 2507690.0,
      "step": 3145
    },
    {
      "entropy": 5.373710632324219,
      "epoch": 0.8905852417302799,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000495670665880482,
      "loss": 5.4131,
      "mean_token_accuracy": 0.19849524945020675,
      "num_tokens": 2511645.0,
      "step": 3150
    },
    {
      "entropy": 5.465575742721557,
      "epoch": 0.8919988690981058,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004956505618697456,
      "loss": 5.5169,
      "mean_token_accuracy": 0.19546956121921538,
      "num_tokens": 2515465.0,
      "step": 3155
    },
    {
      "entropy": 5.514111185073853,
      "epoch": 0.8934124964659316,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004956304117444064,
      "loss": 5.4052,
      "mean_token_accuracy": 0.20660127103328704,
      "num_tokens": 2519428.0,
      "step": 3160
    },
    {
      "entropy": 5.380162906646729,
      "epoch": 0.8948261238337574,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004956102155086754,
      "loss": 5.3122,
      "mean_token_accuracy": 0.20972901433706284,
      "num_tokens": 2523346.0,
      "step": 3165
    },
    {
      "entropy": 5.413879680633545,
      "epoch": 0.8962397512015833,
      "grad_norm": 1.375,
      "learning_rate": 0.0004955899731667737,
      "loss": 5.4258,
      "mean_token_accuracy": 0.20761295258998871,
      "num_tokens": 2527299.0,
      "step": 3170
    },
    {
      "entropy": 5.492557239532471,
      "epoch": 0.8976533785694091,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004955696847229315,
      "loss": 5.4444,
      "mean_token_accuracy": 0.2052128329873085,
      "num_tokens": 2531492.0,
      "step": 3175
    },
    {
      "entropy": 5.350836133956909,
      "epoch": 0.899067005937235,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004955493501813892,
      "loss": 5.3134,
      "mean_token_accuracy": 0.20852610021829604,
      "num_tokens": 2535641.0,
      "step": 3180
    },
    {
      "entropy": 5.369951057434082,
      "epoch": 0.9004806333050608,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004955289695463967,
      "loss": 5.4727,
      "mean_token_accuracy": 0.21253684163093567,
      "num_tokens": 2539674.0,
      "step": 3185
    },
    {
      "entropy": 5.37187089920044,
      "epoch": 0.9018942606728866,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004955085428222131,
      "loss": 5.3947,
      "mean_token_accuracy": 0.20510520339012145,
      "num_tokens": 2543574.0,
      "step": 3190
    },
    {
      "entropy": 5.388985538482666,
      "epoch": 0.9033078880407125,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004954880700131077,
      "loss": 5.4543,
      "mean_token_accuracy": 0.20005883425474166,
      "num_tokens": 2547381.0,
      "step": 3195
    },
    {
      "entropy": 5.431127548217773,
      "epoch": 0.9047215154085383,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004954675511233592,
      "loss": 5.3738,
      "mean_token_accuracy": 0.20665386617183684,
      "num_tokens": 2551545.0,
      "step": 3200
    },
    {
      "entropy": 5.346138429641724,
      "epoch": 0.9061351427763642,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004954469861572557,
      "loss": 5.3804,
      "mean_token_accuracy": 0.20939448177814485,
      "num_tokens": 2555336.0,
      "step": 3205
    },
    {
      "entropy": 5.348866081237793,
      "epoch": 0.90754877014419,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004954263751190954,
      "loss": 5.3487,
      "mean_token_accuracy": 0.2036546289920807,
      "num_tokens": 2559301.0,
      "step": 3210
    },
    {
      "entropy": 5.544688940048218,
      "epoch": 0.9089623975120158,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004954057180131857,
      "loss": 5.3933,
      "mean_token_accuracy": 0.19913605898618697,
      "num_tokens": 2563185.0,
      "step": 3215
    },
    {
      "entropy": 5.490880727767944,
      "epoch": 0.9103760248798417,
      "grad_norm": 1.375,
      "learning_rate": 0.0004953850148438439,
      "loss": 5.4419,
      "mean_token_accuracy": 0.2030777409672737,
      "num_tokens": 2567110.0,
      "step": 3220
    },
    {
      "entropy": 5.265858268737793,
      "epoch": 0.9117896522476675,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004953642656153968,
      "loss": 5.3543,
      "mean_token_accuracy": 0.20732428282499313,
      "num_tokens": 2571350.0,
      "step": 3225
    },
    {
      "entropy": 5.432380151748657,
      "epoch": 0.9132032796154933,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000495343470332181,
      "loss": 5.3594,
      "mean_token_accuracy": 0.2041091129183769,
      "num_tokens": 2575139.0,
      "step": 3230
    },
    {
      "entropy": 5.417063283920288,
      "epoch": 0.9146169069833192,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004953226289985425,
      "loss": 5.3202,
      "mean_token_accuracy": 0.2021815821528435,
      "num_tokens": 2579124.0,
      "step": 3235
    },
    {
      "entropy": 5.469454669952393,
      "epoch": 0.916030534351145,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004953017416188367,
      "loss": 5.4209,
      "mean_token_accuracy": 0.2019435152411461,
      "num_tokens": 2583130.0,
      "step": 3240
    },
    {
      "entropy": 5.320521831512451,
      "epoch": 0.9174441617189709,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004952808081974294,
      "loss": 5.4288,
      "mean_token_accuracy": 0.1974789559841156,
      "num_tokens": 2587210.0,
      "step": 3245
    },
    {
      "entropy": 5.390252923965454,
      "epoch": 0.9188577890867967,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004952598287386954,
      "loss": 5.3111,
      "mean_token_accuracy": 0.20741842389106752,
      "num_tokens": 2591372.0,
      "step": 3250
    },
    {
      "entropy": 5.510976982116699,
      "epoch": 0.9202714164546225,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004952388032470191,
      "loss": 5.4366,
      "mean_token_accuracy": 0.20590081959962844,
      "num_tokens": 2595683.0,
      "step": 3255
    },
    {
      "entropy": 5.3200623989105225,
      "epoch": 0.9216850438224484,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004952177317267949,
      "loss": 5.2738,
      "mean_token_accuracy": 0.22108482271432878,
      "num_tokens": 2599472.0,
      "step": 3260
    },
    {
      "entropy": 5.315955686569214,
      "epoch": 0.9230986711902742,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004951966141824265,
      "loss": 5.2678,
      "mean_token_accuracy": 0.21187404096126555,
      "num_tokens": 2603509.0,
      "step": 3265
    },
    {
      "entropy": 5.31221375465393,
      "epoch": 0.9245122985581001,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004951754506183275,
      "loss": 5.2705,
      "mean_token_accuracy": 0.21165193170309066,
      "num_tokens": 2607191.0,
      "step": 3270
    },
    {
      "entropy": 5.396156311035156,
      "epoch": 0.9259259259259259,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004951542410389208,
      "loss": 5.3678,
      "mean_token_accuracy": 0.21061211675405503,
      "num_tokens": 2611698.0,
      "step": 3275
    },
    {
      "entropy": 5.394646692276001,
      "epoch": 0.9273395532937517,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004951329854486392,
      "loss": 5.365,
      "mean_token_accuracy": 0.2175261914730072,
      "num_tokens": 2615340.0,
      "step": 3280
    },
    {
      "entropy": 5.262735033035279,
      "epoch": 0.9287531806615776,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004951116838519248,
      "loss": 5.3495,
      "mean_token_accuracy": 0.21772853434085845,
      "num_tokens": 2619465.0,
      "step": 3285
    },
    {
      "entropy": 5.306789255142212,
      "epoch": 0.9301668080294034,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004950903362532296,
      "loss": 5.2069,
      "mean_token_accuracy": 0.21759452074766159,
      "num_tokens": 2623625.0,
      "step": 3290
    },
    {
      "entropy": 5.370097017288208,
      "epoch": 0.9315804353972293,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004950689426570151,
      "loss": 5.2423,
      "mean_token_accuracy": 0.22313392162322998,
      "num_tokens": 2627368.0,
      "step": 3295
    },
    {
      "entropy": 5.3717914581298825,
      "epoch": 0.9329940627650551,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004950475030677525,
      "loss": 5.3754,
      "mean_token_accuracy": 0.2174657925963402,
      "num_tokens": 2631280.0,
      "step": 3300
    },
    {
      "entropy": 5.217867136001587,
      "epoch": 0.9344076901328809,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004950260174899225,
      "loss": 5.3116,
      "mean_token_accuracy": 0.21754174381494523,
      "num_tokens": 2635420.0,
      "step": 3305
    },
    {
      "entropy": 5.3360615253448485,
      "epoch": 0.9358213175007069,
      "grad_norm": 1.375,
      "learning_rate": 0.0004950044859280154,
      "loss": 5.2812,
      "mean_token_accuracy": 0.21761411130428315,
      "num_tokens": 2639370.0,
      "step": 3310
    },
    {
      "entropy": 5.510275888442993,
      "epoch": 0.9372349448685326,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004949829083865312,
      "loss": 5.3481,
      "mean_token_accuracy": 0.20918049663305283,
      "num_tokens": 2643195.0,
      "step": 3315
    },
    {
      "entropy": 5.340539121627808,
      "epoch": 0.9386485722363584,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004949612848699795,
      "loss": 5.4611,
      "mean_token_accuracy": 0.19524372071027757,
      "num_tokens": 2647144.0,
      "step": 3320
    },
    {
      "entropy": 5.301290702819824,
      "epoch": 0.9400621996041844,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004949396153828794,
      "loss": 5.2641,
      "mean_token_accuracy": 0.20904955863952637,
      "num_tokens": 2651201.0,
      "step": 3325
    },
    {
      "entropy": 5.3793394565582275,
      "epoch": 0.9414758269720102,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004949178999297599,
      "loss": 5.3657,
      "mean_token_accuracy": 0.20661818683147432,
      "num_tokens": 2655281.0,
      "step": 3330
    },
    {
      "entropy": 5.312300682067871,
      "epoch": 0.9428894543398361,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004948961385151592,
      "loss": 5.1949,
      "mean_token_accuracy": 0.22709580808877944,
      "num_tokens": 2659226.0,
      "step": 3335
    },
    {
      "entropy": 5.29730715751648,
      "epoch": 0.9443030817076619,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004948743311436254,
      "loss": 5.293,
      "mean_token_accuracy": 0.20665379315614701,
      "num_tokens": 2663168.0,
      "step": 3340
    },
    {
      "entropy": 5.291025257110595,
      "epoch": 0.9457167090754877,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004948524778197159,
      "loss": 5.4316,
      "mean_token_accuracy": 0.19959478527307511,
      "num_tokens": 2667258.0,
      "step": 3345
    },
    {
      "entropy": 5.387850999832153,
      "epoch": 0.9471303364433136,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004948305785479982,
      "loss": 5.3246,
      "mean_token_accuracy": 0.2101334184408188,
      "num_tokens": 2671149.0,
      "step": 3350
    },
    {
      "entropy": 5.375804805755616,
      "epoch": 0.9485439638111394,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004948086333330488,
      "loss": 5.4035,
      "mean_token_accuracy": 0.20752620100975036,
      "num_tokens": 2675287.0,
      "step": 3355
    },
    {
      "entropy": 5.319958162307739,
      "epoch": 0.9499575911789653,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004947866421794545,
      "loss": 5.2974,
      "mean_token_accuracy": 0.20366615653038025,
      "num_tokens": 2679148.0,
      "step": 3360
    },
    {
      "entropy": 5.331606769561768,
      "epoch": 0.9513712185467911,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004947646050918109,
      "loss": 5.2879,
      "mean_token_accuracy": 0.21156328320503234,
      "num_tokens": 2683125.0,
      "step": 3365
    },
    {
      "entropy": 5.242379379272461,
      "epoch": 0.9527848459146169,
      "grad_norm": 1.421875,
      "learning_rate": 0.000494742522074724,
      "loss": 5.3068,
      "mean_token_accuracy": 0.21346987932920455,
      "num_tokens": 2687079.0,
      "step": 3370
    },
    {
      "entropy": 5.454464149475098,
      "epoch": 0.9541984732824428,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004947203931328087,
      "loss": 5.4766,
      "mean_token_accuracy": 0.1971137121319771,
      "num_tokens": 2691065.0,
      "step": 3375
    },
    {
      "entropy": 5.349691915512085,
      "epoch": 0.9556121006502686,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004946982182706901,
      "loss": 5.3014,
      "mean_token_accuracy": 0.21622299104928971,
      "num_tokens": 2694844.0,
      "step": 3380
    },
    {
      "entropy": 5.315452194213867,
      "epoch": 0.9570257280180944,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004946759974930023,
      "loss": 5.4002,
      "mean_token_accuracy": 0.21129725277423858,
      "num_tokens": 2698932.0,
      "step": 3385
    },
    {
      "entropy": 5.222040033340454,
      "epoch": 0.9584393553859203,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004946537308043895,
      "loss": 5.2096,
      "mean_token_accuracy": 0.21286771297454835,
      "num_tokens": 2702810.0,
      "step": 3390
    },
    {
      "entropy": 5.316049957275391,
      "epoch": 0.9598529827537461,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004946314182095053,
      "loss": 5.4626,
      "mean_token_accuracy": 0.1963511198759079,
      "num_tokens": 2706701.0,
      "step": 3395
    },
    {
      "entropy": 5.2958512783050535,
      "epoch": 0.961266610121572,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004946090597130128,
      "loss": 5.3797,
      "mean_token_accuracy": 0.2054529070854187,
      "num_tokens": 2710533.0,
      "step": 3400
    },
    {
      "entropy": 5.382183790206909,
      "epoch": 0.9626802374893978,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004945866553195848,
      "loss": 5.4001,
      "mean_token_accuracy": 0.20687114000320433,
      "num_tokens": 2714586.0,
      "step": 3405
    },
    {
      "entropy": 5.3121726512908936,
      "epoch": 0.9640938648572236,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004945642050339038,
      "loss": 5.4154,
      "mean_token_accuracy": 0.20771215111017227,
      "num_tokens": 2718357.0,
      "step": 3410
    },
    {
      "entropy": 5.5918089866638185,
      "epoch": 0.9655074922250495,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004945417088606617,
      "loss": 5.6496,
      "mean_token_accuracy": 0.1879683941602707,
      "num_tokens": 2722811.0,
      "step": 3415
    },
    {
      "entropy": 5.407415056228638,
      "epoch": 0.9669211195928753,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004945191668045601,
      "loss": 5.3271,
      "mean_token_accuracy": 0.21375867277383803,
      "num_tokens": 2726815.0,
      "step": 3420
    },
    {
      "entropy": 5.357066202163696,
      "epoch": 0.9683347469607012,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004944965788703101,
      "loss": 5.3922,
      "mean_token_accuracy": 0.20253707319498063,
      "num_tokens": 2730683.0,
      "step": 3425
    },
    {
      "entropy": 5.308847379684448,
      "epoch": 0.969748374328527,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004944739450626323,
      "loss": 5.2203,
      "mean_token_accuracy": 0.21702735275030136,
      "num_tokens": 2734950.0,
      "step": 3430
    },
    {
      "entropy": 5.230969333648682,
      "epoch": 0.9711620016963528,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004944512653862572,
      "loss": 5.2431,
      "mean_token_accuracy": 0.2184822827577591,
      "num_tokens": 2738894.0,
      "step": 3435
    },
    {
      "entropy": 5.30678424835205,
      "epoch": 0.9725756290641787,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004944285398459247,
      "loss": 5.3887,
      "mean_token_accuracy": 0.21177172064781188,
      "num_tokens": 2742848.0,
      "step": 3440
    },
    {
      "entropy": 5.21362042427063,
      "epoch": 0.9739892564320045,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004944057684463844,
      "loss": 5.345,
      "mean_token_accuracy": 0.21214967370033264,
      "num_tokens": 2746809.0,
      "step": 3445
    },
    {
      "entropy": 5.26718373298645,
      "epoch": 0.9754028837998303,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004943829511923952,
      "loss": 5.3199,
      "mean_token_accuracy": 0.21215527802705764,
      "num_tokens": 2750450.0,
      "step": 3450
    },
    {
      "entropy": 5.308568000793457,
      "epoch": 0.9768165111676562,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004943600880887258,
      "loss": 5.3691,
      "mean_token_accuracy": 0.20462850779294967,
      "num_tokens": 2754572.0,
      "step": 3455
    },
    {
      "entropy": 5.425100803375244,
      "epoch": 0.978230138535482,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004943371791401544,
      "loss": 5.4842,
      "mean_token_accuracy": 0.1998129740357399,
      "num_tokens": 2758614.0,
      "step": 3460
    },
    {
      "entropy": 5.456117296218872,
      "epoch": 0.9796437659033079,
      "grad_norm": 1.296875,
      "learning_rate": 0.000494314224351469,
      "loss": 5.3523,
      "mean_token_accuracy": 0.20461326837539673,
      "num_tokens": 2763031.0,
      "step": 3465
    },
    {
      "entropy": 5.412753391265869,
      "epoch": 0.9810573932711337,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004942912237274668,
      "loss": 5.5255,
      "mean_token_accuracy": 0.1959344446659088,
      "num_tokens": 2766953.0,
      "step": 3470
    },
    {
      "entropy": 5.379322052001953,
      "epoch": 0.9824710206389595,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000494268177272955,
      "loss": 5.3584,
      "mean_token_accuracy": 0.20648946315050126,
      "num_tokens": 2770822.0,
      "step": 3475
    },
    {
      "entropy": 5.38189845085144,
      "epoch": 0.9838846480067854,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00049424508499275,
      "loss": 5.3405,
      "mean_token_accuracy": 0.2107373371720314,
      "num_tokens": 2774599.0,
      "step": 3480
    },
    {
      "entropy": 5.2463294506073,
      "epoch": 0.9852982753746112,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000494221946891678,
      "loss": 5.3092,
      "mean_token_accuracy": 0.2126726135611534,
      "num_tokens": 2778508.0,
      "step": 3485
    },
    {
      "entropy": 5.371696043014526,
      "epoch": 0.9867119027424371,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004941987629745747,
      "loss": 5.432,
      "mean_token_accuracy": 0.20082706212997437,
      "num_tokens": 2782535.0,
      "step": 3490
    },
    {
      "entropy": 5.401870393753052,
      "epoch": 0.9881255301102629,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004941755332462856,
      "loss": 5.4365,
      "mean_token_accuracy": 0.19900313913822174,
      "num_tokens": 2786339.0,
      "step": 3495
    },
    {
      "entropy": 5.395771121978759,
      "epoch": 0.9895391574780887,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004941522577116652,
      "loss": 5.3906,
      "mean_token_accuracy": 0.21107237339019774,
      "num_tokens": 2790466.0,
      "step": 3500
    },
    {
      "entropy": 5.314546251296997,
      "epoch": 0.9909527848459146,
      "grad_norm": 1.375,
      "learning_rate": 0.0004941289363755782,
      "loss": 5.368,
      "mean_token_accuracy": 0.21377213150262833,
      "num_tokens": 2794267.0,
      "step": 3505
    },
    {
      "entropy": 5.3443827629089355,
      "epoch": 0.9923664122137404,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004941055692428987,
      "loss": 5.3681,
      "mean_token_accuracy": 0.20529019683599473,
      "num_tokens": 2798394.0,
      "step": 3510
    },
    {
      "entropy": 5.266571569442749,
      "epoch": 0.9937800395815664,
      "grad_norm": 1.5,
      "learning_rate": 0.00049408215631851,
      "loss": 5.3353,
      "mean_token_accuracy": 0.20843572318553924,
      "num_tokens": 2802429.0,
      "step": 3515
    },
    {
      "entropy": 5.37566990852356,
      "epoch": 0.9951936669493922,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004940586976073054,
      "loss": 5.3752,
      "mean_token_accuracy": 0.2042718470096588,
      "num_tokens": 2806303.0,
      "step": 3520
    },
    {
      "entropy": 5.349119377136231,
      "epoch": 0.996607294317218,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004940351931141875,
      "loss": 5.2832,
      "mean_token_accuracy": 0.22043754458427428,
      "num_tokens": 2810123.0,
      "step": 3525
    },
    {
      "entropy": 5.228527593612671,
      "epoch": 0.9980209216850439,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004940116428440688,
      "loss": 5.3466,
      "mean_token_accuracy": 0.20550204664468766,
      "num_tokens": 2813991.0,
      "step": 3530
    },
    {
      "entropy": 5.278249073028564,
      "epoch": 0.9994345490528697,
      "grad_norm": 1.375,
      "learning_rate": 0.0004939880468018712,
      "loss": 5.3245,
      "mean_token_accuracy": 0.21513051092624663,
      "num_tokens": 2818203.0,
      "step": 3535
    },
    {
      "entropy": 5.317620420455933,
      "epoch": 1.0008481764206956,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004939644049925259,
      "loss": 5.2118,
      "mean_token_accuracy": 0.2047821506857872,
      "num_tokens": 2822334.0,
      "step": 3540
    },
    {
      "entropy": 5.4087518692016605,
      "epoch": 1.0022618037885214,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000493940717420974,
      "loss": 5.3371,
      "mean_token_accuracy": 0.20358794927597046,
      "num_tokens": 2826277.0,
      "step": 3545
    },
    {
      "entropy": 5.272971773147583,
      "epoch": 1.0036754311563472,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000493916984092166,
      "loss": 5.1337,
      "mean_token_accuracy": 0.21596817970275878,
      "num_tokens": 2830093.0,
      "step": 3550
    },
    {
      "entropy": 5.206704711914062,
      "epoch": 1.005089058524173,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004938932050110621,
      "loss": 5.192,
      "mean_token_accuracy": 0.2177193596959114,
      "num_tokens": 2833961.0,
      "step": 3555
    },
    {
      "entropy": 5.258702182769776,
      "epoch": 1.0065026858919988,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000493869380182632,
      "loss": 5.0877,
      "mean_token_accuracy": 0.2187505453824997,
      "num_tokens": 2837923.0,
      "step": 3560
    },
    {
      "entropy": 5.173738670349121,
      "epoch": 1.0079163132598248,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004938455096118548,
      "loss": 4.9735,
      "mean_token_accuracy": 0.23251553624868393,
      "num_tokens": 2841820.0,
      "step": 3565
    },
    {
      "entropy": 5.150343608856201,
      "epoch": 1.0093299406276506,
      "grad_norm": 1.375,
      "learning_rate": 0.0004938215933037195,
      "loss": 5.1746,
      "mean_token_accuracy": 0.21782366931438446,
      "num_tokens": 2845901.0,
      "step": 3570
    },
    {
      "entropy": 5.283948040008545,
      "epoch": 1.0107435679954764,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004937976312632242,
      "loss": 5.1169,
      "mean_token_accuracy": 0.21362414062023163,
      "num_tokens": 2849841.0,
      "step": 3575
    },
    {
      "entropy": 5.270070743560791,
      "epoch": 1.0121571953633022,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000493773623495377,
      "loss": 4.9722,
      "mean_token_accuracy": 0.2279058039188385,
      "num_tokens": 2853601.0,
      "step": 3580
    },
    {
      "entropy": 5.2371851921081545,
      "epoch": 1.013570822731128,
      "grad_norm": 1.375,
      "learning_rate": 0.0004937495700051955,
      "loss": 5.1194,
      "mean_token_accuracy": 0.21858393251895905,
      "num_tokens": 2857967.0,
      "step": 3585
    },
    {
      "entropy": 5.205569744110107,
      "epoch": 1.014984450098954,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004937254707977065,
      "loss": 5.1483,
      "mean_token_accuracy": 0.21485333144664764,
      "num_tokens": 2862009.0,
      "step": 3590
    },
    {
      "entropy": 5.18958740234375,
      "epoch": 1.0163980774667798,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004937013258779465,
      "loss": 5.0857,
      "mean_token_accuracy": 0.21460189372301103,
      "num_tokens": 2866078.0,
      "step": 3595
    },
    {
      "entropy": 5.267029333114624,
      "epoch": 1.0178117048346056,
      "grad_norm": 1.546875,
      "learning_rate": 0.000493677135250962,
      "loss": 5.1461,
      "mean_token_accuracy": 0.20465624928474427,
      "num_tokens": 2870041.0,
      "step": 3600
    },
    {
      "entropy": 5.249391555786133,
      "epoch": 1.0192253322024314,
      "grad_norm": 1.5,
      "learning_rate": 0.0004936528989218084,
      "loss": 5.072,
      "mean_token_accuracy": 0.21645056456327438,
      "num_tokens": 2873901.0,
      "step": 3605
    },
    {
      "entropy": 5.1907267570495605,
      "epoch": 1.0206389595702572,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004936286168955511,
      "loss": 5.1136,
      "mean_token_accuracy": 0.21813808530569076,
      "num_tokens": 2877916.0,
      "step": 3610
    },
    {
      "entropy": 5.158541250228882,
      "epoch": 1.0220525869380832,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004936042891772647,
      "loss": 5.0817,
      "mean_token_accuracy": 0.21026738733053207,
      "num_tokens": 2882207.0,
      "step": 3615
    },
    {
      "entropy": 5.281835222244263,
      "epoch": 1.023466214305909,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004935799157720338,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21685025691986085,
      "num_tokens": 2886211.0,
      "step": 3620
    },
    {
      "entropy": 5.264526414871216,
      "epoch": 1.0248798416737348,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004935554966849522,
      "loss": 5.3132,
      "mean_token_accuracy": 0.2117941990494728,
      "num_tokens": 2890167.0,
      "step": 3625
    },
    {
      "entropy": 5.1619974136352536,
      "epoch": 1.0262934690415606,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004935310319211231,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22505988031625748,
      "num_tokens": 2894214.0,
      "step": 3630
    },
    {
      "entropy": 5.226195907592773,
      "epoch": 1.0277070964093864,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004935065214856599,
      "loss": 5.1723,
      "mean_token_accuracy": 0.22310581803321838,
      "num_tokens": 2898423.0,
      "step": 3635
    },
    {
      "entropy": 5.188407135009766,
      "epoch": 1.0291207237772124,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004934819653836848,
      "loss": 5.0733,
      "mean_token_accuracy": 0.22534722983837127,
      "num_tokens": 2902728.0,
      "step": 3640
    },
    {
      "entropy": 5.132740020751953,
      "epoch": 1.0305343511450382,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004934573636203299,
      "loss": 5.1118,
      "mean_token_accuracy": 0.2192051351070404,
      "num_tokens": 2906687.0,
      "step": 3645
    },
    {
      "entropy": 5.220104312896728,
      "epoch": 1.031947978512864,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004934327162007371,
      "loss": 5.1194,
      "mean_token_accuracy": 0.21550490707159042,
      "num_tokens": 2910708.0,
      "step": 3650
    },
    {
      "entropy": 5.215138483047485,
      "epoch": 1.0333616058806898,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004934080231300573,
      "loss": 5.1651,
      "mean_token_accuracy": 0.21759526580572128,
      "num_tokens": 2914416.0,
      "step": 3655
    },
    {
      "entropy": 5.278332233428955,
      "epoch": 1.0347752332485156,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004933832844134511,
      "loss": 5.1765,
      "mean_token_accuracy": 0.22026587575674056,
      "num_tokens": 2918738.0,
      "step": 3660
    },
    {
      "entropy": 5.259198236465454,
      "epoch": 1.0361888606163416,
      "grad_norm": 1.375,
      "learning_rate": 0.0004933585000560891,
      "loss": 5.1457,
      "mean_token_accuracy": 0.21962920278310777,
      "num_tokens": 2922575.0,
      "step": 3665
    },
    {
      "entropy": 5.286904191970825,
      "epoch": 1.0376024879841674,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004933336700631508,
      "loss": 5.0971,
      "mean_token_accuracy": 0.22029909789562224,
      "num_tokens": 2926707.0,
      "step": 3670
    },
    {
      "entropy": 5.190115880966187,
      "epoch": 1.0390161153519932,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004933087944398256,
      "loss": 5.1447,
      "mean_token_accuracy": 0.22542219310998918,
      "num_tokens": 2930785.0,
      "step": 3675
    },
    {
      "entropy": 5.010173511505127,
      "epoch": 1.040429742719819,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004932838731913124,
      "loss": 4.9454,
      "mean_token_accuracy": 0.2361238792538643,
      "num_tokens": 2934893.0,
      "step": 3680
    },
    {
      "entropy": 5.144097805023193,
      "epoch": 1.0418433700876448,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004932589063228194,
      "loss": 5.1796,
      "mean_token_accuracy": 0.20868771225214006,
      "num_tokens": 2939034.0,
      "step": 3685
    },
    {
      "entropy": 5.239131784439087,
      "epoch": 1.0432569974554706,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004932338938395647,
      "loss": 5.0644,
      "mean_token_accuracy": 0.22203764021396638,
      "num_tokens": 2943089.0,
      "step": 3690
    },
    {
      "entropy": 5.205106735229492,
      "epoch": 1.0446706248232966,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004932088357467757,
      "loss": 5.1724,
      "mean_token_accuracy": 0.21926317363977432,
      "num_tokens": 2947128.0,
      "step": 3695
    },
    {
      "entropy": 5.227344417572022,
      "epoch": 1.0460842521911224,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004931837320496894,
      "loss": 5.2061,
      "mean_token_accuracy": 0.2127397507429123,
      "num_tokens": 2951295.0,
      "step": 3700
    },
    {
      "entropy": 5.240218257904052,
      "epoch": 1.0474978795589482,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004931585827535524,
      "loss": 5.2099,
      "mean_token_accuracy": 0.20746835619211196,
      "num_tokens": 2955522.0,
      "step": 3705
    },
    {
      "entropy": 5.284507751464844,
      "epoch": 1.048911506926774,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004931333878636205,
      "loss": 5.1074,
      "mean_token_accuracy": 0.21079284995794295,
      "num_tokens": 2959410.0,
      "step": 3710
    },
    {
      "entropy": 5.182925081253051,
      "epoch": 1.0503251342945998,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004931081473851595,
      "loss": 5.1011,
      "mean_token_accuracy": 0.21961848735809325,
      "num_tokens": 2963411.0,
      "step": 3715
    },
    {
      "entropy": 5.1357635974884035,
      "epoch": 1.0517387616624259,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004930828613234446,
      "loss": 5.1234,
      "mean_token_accuracy": 0.21083565503358842,
      "num_tokens": 2967623.0,
      "step": 3720
    },
    {
      "entropy": 5.2366697788238525,
      "epoch": 1.0531523890302517,
      "grad_norm": 1.40625,
      "learning_rate": 0.00049305752968376,
      "loss": 5.0573,
      "mean_token_accuracy": 0.22657375037670135,
      "num_tokens": 2971708.0,
      "step": 3725
    },
    {
      "entropy": 5.128624773025512,
      "epoch": 1.0545660163980775,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004930321524714002,
      "loss": 5.0949,
      "mean_token_accuracy": 0.22488353103399278,
      "num_tokens": 2975491.0,
      "step": 3730
    },
    {
      "entropy": 5.220402336120605,
      "epoch": 1.0559796437659033,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004930067296916689,
      "loss": 5.1806,
      "mean_token_accuracy": 0.2148841291666031,
      "num_tokens": 2979368.0,
      "step": 3735
    },
    {
      "entropy": 5.17947735786438,
      "epoch": 1.057393271133729,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004929812613498791,
      "loss": 5.1157,
      "mean_token_accuracy": 0.22063000053167342,
      "num_tokens": 2983939.0,
      "step": 3740
    },
    {
      "entropy": 5.137882232666016,
      "epoch": 1.058806898501555,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004929557474513537,
      "loss": 5.0549,
      "mean_token_accuracy": 0.2171742931008339,
      "num_tokens": 2987895.0,
      "step": 3745
    },
    {
      "entropy": 5.177118396759033,
      "epoch": 1.0602205258693809,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004929301880014249,
      "loss": 5.1365,
      "mean_token_accuracy": 0.21334730982780456,
      "num_tokens": 2992017.0,
      "step": 3750
    },
    {
      "entropy": 5.217937231063843,
      "epoch": 1.0616341532372067,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004929045830054344,
      "loss": 5.1038,
      "mean_token_accuracy": 0.20568303912878036,
      "num_tokens": 2995957.0,
      "step": 3755
    },
    {
      "entropy": 5.117181968688965,
      "epoch": 1.0630477806050325,
      "grad_norm": 1.5,
      "learning_rate": 0.0004928789324687335,
      "loss": 4.9179,
      "mean_token_accuracy": 0.2357058584690094,
      "num_tokens": 2999816.0,
      "step": 3760
    },
    {
      "entropy": 5.129719686508179,
      "epoch": 1.0644614079728583,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004928532363966831,
      "loss": 5.0958,
      "mean_token_accuracy": 0.22046199142932893,
      "num_tokens": 3003801.0,
      "step": 3765
    },
    {
      "entropy": 5.2426612854003904,
      "epoch": 1.0658750353406843,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004928274947946534,
      "loss": 5.1781,
      "mean_token_accuracy": 0.20456161350011826,
      "num_tokens": 3007710.0,
      "step": 3770
    },
    {
      "entropy": 5.277953386306763,
      "epoch": 1.06728866270851,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004928017076680244,
      "loss": 5.1457,
      "mean_token_accuracy": 0.21807983964681626,
      "num_tokens": 3011573.0,
      "step": 3775
    },
    {
      "entropy": 5.169133996963501,
      "epoch": 1.0687022900763359,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004927758750221852,
      "loss": 5.1423,
      "mean_token_accuracy": 0.22268428951501845,
      "num_tokens": 3015832.0,
      "step": 3780
    },
    {
      "entropy": 5.183489561080933,
      "epoch": 1.0701159174441617,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004927499968625349,
      "loss": 5.06,
      "mean_token_accuracy": 0.22604098916053772,
      "num_tokens": 3019878.0,
      "step": 3785
    },
    {
      "entropy": 5.092603254318237,
      "epoch": 1.0715295448119875,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004927240731944818,
      "loss": 4.965,
      "mean_token_accuracy": 0.23898600190877914,
      "num_tokens": 3023725.0,
      "step": 3790
    },
    {
      "entropy": 5.276487064361572,
      "epoch": 1.0729431721798135,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004926981040234437,
      "loss": 5.2201,
      "mean_token_accuracy": 0.2106180176138878,
      "num_tokens": 3028055.0,
      "step": 3795
    },
    {
      "entropy": 5.121365165710449,
      "epoch": 1.0743567995476393,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004926720893548482,
      "loss": 5.0798,
      "mean_token_accuracy": 0.22679545879364013,
      "num_tokens": 3032214.0,
      "step": 3800
    },
    {
      "entropy": 5.160611343383789,
      "epoch": 1.075770426915465,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004926460291941319,
      "loss": 5.0611,
      "mean_token_accuracy": 0.21699931621551513,
      "num_tokens": 3036507.0,
      "step": 3805
    },
    {
      "entropy": 5.162946701049805,
      "epoch": 1.077184054283291,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004926199235467414,
      "loss": 5.2197,
      "mean_token_accuracy": 0.21580809652805327,
      "num_tokens": 3040720.0,
      "step": 3810
    },
    {
      "entropy": 5.217563152313232,
      "epoch": 1.0785976816511167,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004925937724181326,
      "loss": 5.1239,
      "mean_token_accuracy": 0.21833766549825667,
      "num_tokens": 3044826.0,
      "step": 3815
    },
    {
      "entropy": 5.15655665397644,
      "epoch": 1.0800113090189427,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004925675758137711,
      "loss": 5.1502,
      "mean_token_accuracy": 0.22103135883808137,
      "num_tokens": 3048840.0,
      "step": 3820
    },
    {
      "entropy": 5.111991596221924,
      "epoch": 1.0814249363867685,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004925413337391315,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21856762021780013,
      "num_tokens": 3052716.0,
      "step": 3825
    },
    {
      "entropy": 5.169492101669311,
      "epoch": 1.0828385637545943,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004925150461996984,
      "loss": 5.099,
      "mean_token_accuracy": 0.2193046286702156,
      "num_tokens": 3056662.0,
      "step": 3830
    },
    {
      "entropy": 5.2572277069091795,
      "epoch": 1.08425219112242,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004924887132009656,
      "loss": 5.1132,
      "mean_token_accuracy": 0.21358698308467866,
      "num_tokens": 3060638.0,
      "step": 3835
    },
    {
      "entropy": 5.1547552108764645,
      "epoch": 1.085665818490246,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004924623347484369,
      "loss": 5.1267,
      "mean_token_accuracy": 0.2179493010044098,
      "num_tokens": 3064438.0,
      "step": 3840
    },
    {
      "entropy": 5.1431396961212155,
      "epoch": 1.087079445858072,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004924359108476248,
      "loss": 4.9942,
      "mean_token_accuracy": 0.2318810537457466,
      "num_tokens": 3068099.0,
      "step": 3845
    },
    {
      "entropy": 5.060890436172485,
      "epoch": 1.0884930732258977,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004924094415040519,
      "loss": 4.9353,
      "mean_token_accuracy": 0.226554836332798,
      "num_tokens": 3072039.0,
      "step": 3850
    },
    {
      "entropy": 5.062123107910156,
      "epoch": 1.0899067005937235,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004923829267232503,
      "loss": 5.1234,
      "mean_token_accuracy": 0.22134972512722015,
      "num_tokens": 3076110.0,
      "step": 3855
    },
    {
      "entropy": 5.249336242675781,
      "epoch": 1.0913203279615493,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004923563665107612,
      "loss": 5.0262,
      "mean_token_accuracy": 0.22480145394802092,
      "num_tokens": 3080514.0,
      "step": 3860
    },
    {
      "entropy": 5.200931787490845,
      "epoch": 1.0927339553293751,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004923297608721355,
      "loss": 5.1069,
      "mean_token_accuracy": 0.21789889186620712,
      "num_tokens": 3084537.0,
      "step": 3865
    },
    {
      "entropy": 5.098612928390503,
      "epoch": 1.094147582697201,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004923031098129338,
      "loss": 5.1373,
      "mean_token_accuracy": 0.22479343861341478,
      "num_tokens": 3088107.0,
      "step": 3870
    },
    {
      "entropy": 5.085871076583862,
      "epoch": 1.095561210065027,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004922764133387259,
      "loss": 5.1018,
      "mean_token_accuracy": 0.2229100853204727,
      "num_tokens": 3091790.0,
      "step": 3875
    },
    {
      "entropy": 5.149891138076782,
      "epoch": 1.0969748374328527,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004922496714550912,
      "loss": 4.9944,
      "mean_token_accuracy": 0.22427933812141418,
      "num_tokens": 3095757.0,
      "step": 3880
    },
    {
      "entropy": 5.217122030258179,
      "epoch": 1.0983884648006785,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004922228841676186,
      "loss": 5.1482,
      "mean_token_accuracy": 0.21787536442279815,
      "num_tokens": 3100172.0,
      "step": 3885
    },
    {
      "entropy": 5.16765193939209,
      "epoch": 1.0998020921685043,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004921960514819065,
      "loss": 5.2507,
      "mean_token_accuracy": 0.20131237655878068,
      "num_tokens": 3103857.0,
      "step": 3890
    },
    {
      "entropy": 5.189775514602661,
      "epoch": 1.1012157195363301,
      "grad_norm": 1.5,
      "learning_rate": 0.0004921691734035627,
      "loss": 5.1922,
      "mean_token_accuracy": 0.21736215502023698,
      "num_tokens": 3107817.0,
      "step": 3895
    },
    {
      "entropy": 5.232204532623291,
      "epoch": 1.1026293469041561,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004921422499382046,
      "loss": 5.083,
      "mean_token_accuracy": 0.2302867665886879,
      "num_tokens": 3111655.0,
      "step": 3900
    },
    {
      "entropy": 5.321524429321289,
      "epoch": 1.104042974271982,
      "grad_norm": 1.4140625,
      "learning_rate": 0.000492115281091459,
      "loss": 5.2113,
      "mean_token_accuracy": 0.2267227962613106,
      "num_tokens": 3115959.0,
      "step": 3905
    },
    {
      "entropy": 5.100022649765014,
      "epoch": 1.1054566016398077,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004920882668689623,
      "loss": 4.9499,
      "mean_token_accuracy": 0.23122830390930177,
      "num_tokens": 3119930.0,
      "step": 3910
    },
    {
      "entropy": 4.9987283706665036,
      "epoch": 1.1068702290076335,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004920612072763603,
      "loss": 5.1033,
      "mean_token_accuracy": 0.22130937427282332,
      "num_tokens": 3123806.0,
      "step": 3915
    },
    {
      "entropy": 5.186697578430175,
      "epoch": 1.1082838563754593,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004920341023193082,
      "loss": 5.0954,
      "mean_token_accuracy": 0.22833006232976913,
      "num_tokens": 3127942.0,
      "step": 3920
    },
    {
      "entropy": 5.148104429244995,
      "epoch": 1.1096974837432854,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004920069520034708,
      "loss": 5.1382,
      "mean_token_accuracy": 0.2229016363620758,
      "num_tokens": 3132014.0,
      "step": 3925
    },
    {
      "entropy": 5.220484828948974,
      "epoch": 1.1111111111111112,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004919797563345224,
      "loss": 5.1709,
      "mean_token_accuracy": 0.2237967699766159,
      "num_tokens": 3135854.0,
      "step": 3930
    },
    {
      "entropy": 5.262665748596191,
      "epoch": 1.112524738478937,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004919525153181466,
      "loss": 5.193,
      "mean_token_accuracy": 0.22024215012788773,
      "num_tokens": 3139804.0,
      "step": 3935
    },
    {
      "entropy": 5.2919939994812015,
      "epoch": 1.1139383658467628,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004919252289600367,
      "loss": 5.2713,
      "mean_token_accuracy": 0.21416167318820953,
      "num_tokens": 3143473.0,
      "step": 3940
    },
    {
      "entropy": 5.207576608657837,
      "epoch": 1.1153519932145886,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004918978972658954,
      "loss": 5.11,
      "mean_token_accuracy": 0.21900631934404374,
      "num_tokens": 3147626.0,
      "step": 3945
    },
    {
      "entropy": 5.203172016143799,
      "epoch": 1.1167656205824146,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004918705202414348,
      "loss": 5.1733,
      "mean_token_accuracy": 0.20733029246330262,
      "num_tokens": 3151758.0,
      "step": 3950
    },
    {
      "entropy": 5.146705293655396,
      "epoch": 1.1181792479502404,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004918430978923768,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2245933935046196,
      "num_tokens": 3155490.0,
      "step": 3955
    },
    {
      "entropy": 5.283050489425659,
      "epoch": 1.1195928753180662,
      "grad_norm": 1.546875,
      "learning_rate": 0.000491815630224452,
      "loss": 5.2531,
      "mean_token_accuracy": 0.20974818170070647,
      "num_tokens": 3159385.0,
      "step": 3960
    },
    {
      "entropy": 5.300577306747437,
      "epoch": 1.121006502685892,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004917881172434012,
      "loss": 5.2486,
      "mean_token_accuracy": 0.2130150318145752,
      "num_tokens": 3163198.0,
      "step": 3965
    },
    {
      "entropy": 5.199048089981079,
      "epoch": 1.1224201300537178,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004917605589549746,
      "loss": 5.1292,
      "mean_token_accuracy": 0.21350466161966325,
      "num_tokens": 3167140.0,
      "step": 3970
    },
    {
      "entropy": 5.065745544433594,
      "epoch": 1.1238337574215436,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004917329553649315,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2348677083849907,
      "num_tokens": 3170782.0,
      "step": 3975
    },
    {
      "entropy": 5.21138367652893,
      "epoch": 1.1252473847893696,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004917053064790411,
      "loss": 5.0412,
      "mean_token_accuracy": 0.22398332953453065,
      "num_tokens": 3174678.0,
      "step": 3980
    },
    {
      "entropy": 5.261939668655396,
      "epoch": 1.1266610121571954,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004916776123030815,
      "loss": 5.1506,
      "mean_token_accuracy": 0.22117561101913452,
      "num_tokens": 3178569.0,
      "step": 3985
    },
    {
      "entropy": 5.1857462406158445,
      "epoch": 1.1280746395250212,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004916498728428411,
      "loss": 5.0283,
      "mean_token_accuracy": 0.22664499431848525,
      "num_tokens": 3182328.0,
      "step": 3990
    },
    {
      "entropy": 5.095700263977051,
      "epoch": 1.129488266892847,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004916220881041168,
      "loss": 5.0908,
      "mean_token_accuracy": 0.22369365096092225,
      "num_tokens": 3186477.0,
      "step": 3995
    },
    {
      "entropy": 5.13436369895935,
      "epoch": 1.1309018942606728,
      "grad_norm": 1.625,
      "learning_rate": 0.0004915942580927156,
      "loss": 5.1501,
      "mean_token_accuracy": 0.21428398340940474,
      "num_tokens": 3190517.0,
      "step": 4000
    },
    {
      "epoch": 1.1309018942606728,
      "eval_entropy": 5.076207057877181,
      "eval_loss": 5.468887805938721,
      "eval_mean_token_accuracy": 0.21311278204538567,
      "eval_num_tokens": 3190517.0,
      "eval_runtime": 6.5516,
      "eval_samples_per_second": 1722.34,
      "eval_steps_per_second": 215.369,
      "step": 4000
    },
    {
      "entropy": 5.2192095756530765,
      "epoch": 1.1323155216284988,
      "grad_norm": 1.5,
      "learning_rate": 0.000491566382814454,
      "loss": 5.0677,
      "mean_token_accuracy": 0.2309686318039894,
      "num_tokens": 3194434.0,
      "step": 4005
    },
    {
      "entropy": 5.170869970321656,
      "epoch": 1.1337291489963246,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004915384622751575,
      "loss": 5.0941,
      "mean_token_accuracy": 0.22967276126146316,
      "num_tokens": 3198365.0,
      "step": 4010
    },
    {
      "entropy": 5.160197019577026,
      "epoch": 1.1351427763641504,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004915104964806615,
      "loss": 5.0998,
      "mean_token_accuracy": 0.22692762166261674,
      "num_tokens": 3202272.0,
      "step": 4015
    },
    {
      "entropy": 5.1931580066680905,
      "epoch": 1.1365564037319762,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004914824854368105,
      "loss": 5.2275,
      "mean_token_accuracy": 0.21837261766195298,
      "num_tokens": 3206151.0,
      "step": 4020
    },
    {
      "entropy": 5.273421287536621,
      "epoch": 1.137970031099802,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004914544291494588,
      "loss": 5.1172,
      "mean_token_accuracy": 0.21438456922769547,
      "num_tokens": 3210038.0,
      "step": 4025
    },
    {
      "entropy": 5.2248083591461185,
      "epoch": 1.139383658467628,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004914263276244699,
      "loss": 5.1129,
      "mean_token_accuracy": 0.21759224832057952,
      "num_tokens": 3213883.0,
      "step": 4030
    },
    {
      "entropy": 5.110714769363403,
      "epoch": 1.1407972858354538,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004913981808677168,
      "loss": 5.084,
      "mean_token_accuracy": 0.22000448256731034,
      "num_tokens": 3217861.0,
      "step": 4035
    },
    {
      "entropy": 5.055943632125855,
      "epoch": 1.1422109132032796,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004913699888850822,
      "loss": 4.9511,
      "mean_token_accuracy": 0.24013029634952546,
      "num_tokens": 3221492.0,
      "step": 4040
    },
    {
      "entropy": 5.211551856994629,
      "epoch": 1.1436245405711054,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004913417516824578,
      "loss": 5.1485,
      "mean_token_accuracy": 0.2048600807785988,
      "num_tokens": 3225504.0,
      "step": 4045
    },
    {
      "entropy": 5.247188329696655,
      "epoch": 1.1450381679389312,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004913134692657451,
      "loss": 5.2425,
      "mean_token_accuracy": 0.2094692826271057,
      "num_tokens": 3229722.0,
      "step": 4050
    },
    {
      "entropy": 5.166671943664551,
      "epoch": 1.1464517953067572,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000491285141640855,
      "loss": 5.1254,
      "mean_token_accuracy": 0.22285528630018234,
      "num_tokens": 3233865.0,
      "step": 4055
    },
    {
      "entropy": 5.11505708694458,
      "epoch": 1.147865422674583,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004912567688137078,
      "loss": 5.1077,
      "mean_token_accuracy": 0.20433495193719864,
      "num_tokens": 3237586.0,
      "step": 4060
    },
    {
      "entropy": 5.049159288406372,
      "epoch": 1.1492790500424088,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004912283507902333,
      "loss": 5.1406,
      "mean_token_accuracy": 0.2199555218219757,
      "num_tokens": 3241431.0,
      "step": 4065
    },
    {
      "entropy": 5.288000154495239,
      "epoch": 1.1506926774102346,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004911998875763704,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2233517050743103,
      "num_tokens": 3245427.0,
      "step": 4070
    },
    {
      "entropy": 5.183496952056885,
      "epoch": 1.1521063047780604,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004911713791780681,
      "loss": 5.066,
      "mean_token_accuracy": 0.21957475394010545,
      "num_tokens": 3249361.0,
      "step": 4075
    },
    {
      "entropy": 5.205584955215454,
      "epoch": 1.1535199321458864,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004911428256012841,
      "loss": 5.0734,
      "mean_token_accuracy": 0.2139248251914978,
      "num_tokens": 3253181.0,
      "step": 4080
    },
    {
      "entropy": 5.058682441711426,
      "epoch": 1.1549335595137122,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004911142268519864,
      "loss": 4.991,
      "mean_token_accuracy": 0.22817762494087218,
      "num_tokens": 3257036.0,
      "step": 4085
    },
    {
      "entropy": 5.115410852432251,
      "epoch": 1.156347186881538,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004910855829361515,
      "loss": 5.1131,
      "mean_token_accuracy": 0.21949969679117204,
      "num_tokens": 3261180.0,
      "step": 4090
    },
    {
      "entropy": 5.207958221435547,
      "epoch": 1.1577608142493638,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004910568938597661,
      "loss": 4.9676,
      "mean_token_accuracy": 0.2311878651380539,
      "num_tokens": 3265502.0,
      "step": 4095
    },
    {
      "entropy": 5.248392581939697,
      "epoch": 1.1591744416171896,
      "grad_norm": 1.5,
      "learning_rate": 0.0004910281596288259,
      "loss": 5.1845,
      "mean_token_accuracy": 0.21716278195381164,
      "num_tokens": 3269261.0,
      "step": 4100
    },
    {
      "entropy": 5.27809329032898,
      "epoch": 1.1605880689850157,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004909993802493362,
      "loss": 5.0827,
      "mean_token_accuracy": 0.22198781818151475,
      "num_tokens": 3273700.0,
      "step": 4105
    },
    {
      "entropy": 5.288869762420655,
      "epoch": 1.1620016963528415,
      "grad_norm": 1.5,
      "learning_rate": 0.0004909705557273118,
      "loss": 5.1408,
      "mean_token_accuracy": 0.21834564805030823,
      "num_tokens": 3277829.0,
      "step": 4110
    },
    {
      "entropy": 5.20812029838562,
      "epoch": 1.1634153237206672,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004909416860687767,
      "loss": 5.1661,
      "mean_token_accuracy": 0.22114219516515732,
      "num_tokens": 3281833.0,
      "step": 4115
    },
    {
      "entropy": 5.094548654556275,
      "epoch": 1.164828951088493,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004909127712797646,
      "loss": 5.0392,
      "mean_token_accuracy": 0.2222913384437561,
      "num_tokens": 3285539.0,
      "step": 4120
    },
    {
      "entropy": 5.265209579467774,
      "epoch": 1.1662425784563188,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004908838113663185,
      "loss": 5.1473,
      "mean_token_accuracy": 0.2113668367266655,
      "num_tokens": 3289199.0,
      "step": 4125
    },
    {
      "entropy": 5.2038978099822994,
      "epoch": 1.1676562058241449,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004908548063344907,
      "loss": 5.0573,
      "mean_token_accuracy": 0.22192037105560303,
      "num_tokens": 3292892.0,
      "step": 4130
    },
    {
      "entropy": 5.1661670207977295,
      "epoch": 1.1690698331919707,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004908257561903433,
      "loss": 5.1015,
      "mean_token_accuracy": 0.2180330917239189,
      "num_tokens": 3297059.0,
      "step": 4135
    },
    {
      "entropy": 5.074080228805542,
      "epoch": 1.1704834605597965,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004907966609399474,
      "loss": 4.9808,
      "mean_token_accuracy": 0.2403806209564209,
      "num_tokens": 3300918.0,
      "step": 4140
    },
    {
      "entropy": 5.082246494293213,
      "epoch": 1.1718970879276223,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004907675205893838,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2399939313530922,
      "num_tokens": 3305087.0,
      "step": 4145
    },
    {
      "entropy": 5.212430858612061,
      "epoch": 1.173310715295448,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004907383351447428,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20667629092931747,
      "num_tokens": 3309439.0,
      "step": 4150
    },
    {
      "entropy": 5.168219041824341,
      "epoch": 1.174724342663274,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004907091046121236,
      "loss": 5.1294,
      "mean_token_accuracy": 0.2139601305127144,
      "num_tokens": 3313423.0,
      "step": 4155
    },
    {
      "entropy": 5.162868547439575,
      "epoch": 1.1761379700310999,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004906798289976357,
      "loss": 5.0975,
      "mean_token_accuracy": 0.22220794707536698,
      "num_tokens": 3317300.0,
      "step": 4160
    },
    {
      "entropy": 5.227300357818604,
      "epoch": 1.1775515973989257,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004906505083073971,
      "loss": 5.1894,
      "mean_token_accuracy": 0.2162318244576454,
      "num_tokens": 3321572.0,
      "step": 4165
    },
    {
      "entropy": 5.233516693115234,
      "epoch": 1.1789652247667515,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004906211425475358,
      "loss": 5.0983,
      "mean_token_accuracy": 0.22186366766691207,
      "num_tokens": 3325535.0,
      "step": 4170
    },
    {
      "entropy": 5.226093006134033,
      "epoch": 1.1803788521345773,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004905917317241891,
      "loss": 5.1181,
      "mean_token_accuracy": 0.21822239756584166,
      "num_tokens": 3329260.0,
      "step": 4175
    },
    {
      "entropy": 5.203704023361206,
      "epoch": 1.181792479502403,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004905622758435038,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2233029767870903,
      "num_tokens": 3333492.0,
      "step": 4180
    },
    {
      "entropy": 5.228781843185425,
      "epoch": 1.183206106870229,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004905327749116356,
      "loss": 5.2062,
      "mean_token_accuracy": 0.21565472781658174,
      "num_tokens": 3337730.0,
      "step": 4185
    },
    {
      "entropy": 5.174372863769531,
      "epoch": 1.1846197342380549,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004905032289347504,
      "loss": 5.2121,
      "mean_token_accuracy": 0.21797430366277695,
      "num_tokens": 3341699.0,
      "step": 4190
    },
    {
      "entropy": 5.23379111289978,
      "epoch": 1.1860333616058807,
      "grad_norm": 1.5,
      "learning_rate": 0.0004904736379190229,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2083887994289398,
      "num_tokens": 3345730.0,
      "step": 4195
    },
    {
      "entropy": 5.186162614822388,
      "epoch": 1.1874469889737065,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004904440018706377,
      "loss": 5.0994,
      "mean_token_accuracy": 0.22016676664352416,
      "num_tokens": 3349534.0,
      "step": 4200
    },
    {
      "entropy": 5.1115487575531,
      "epoch": 1.1888606163415323,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004904143207957882,
      "loss": 5.0293,
      "mean_token_accuracy": 0.21696435809135436,
      "num_tokens": 3353279.0,
      "step": 4205
    },
    {
      "entropy": 5.024718713760376,
      "epoch": 1.1902742437093583,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004903845947006778,
      "loss": 5.1487,
      "mean_token_accuracy": 0.21782537251710893,
      "num_tokens": 3357326.0,
      "step": 4210
    },
    {
      "entropy": 5.14876503944397,
      "epoch": 1.191687871077184,
      "grad_norm": 1.453125,
      "learning_rate": 0.000490354823591519,
      "loss": 5.1105,
      "mean_token_accuracy": 0.22573825865983962,
      "num_tokens": 3361219.0,
      "step": 4215
    },
    {
      "entropy": 5.19541335105896,
      "epoch": 1.19310149844501,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004903250074745338,
      "loss": 5.0937,
      "mean_token_accuracy": 0.21079928427934647,
      "num_tokens": 3365136.0,
      "step": 4220
    },
    {
      "entropy": 5.113995361328125,
      "epoch": 1.1945151258128357,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004902951463559535,
      "loss": 5.0199,
      "mean_token_accuracy": 0.22354227304458618,
      "num_tokens": 3369020.0,
      "step": 4225
    },
    {
      "entropy": 5.173697090148925,
      "epoch": 1.1959287531806615,
      "grad_norm": 1.515625,
      "learning_rate": 0.000490265240242019,
      "loss": 5.2176,
      "mean_token_accuracy": 0.21079761087894439,
      "num_tokens": 3373176.0,
      "step": 4230
    },
    {
      "entropy": 5.07566499710083,
      "epoch": 1.1973423805484873,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004902352891389804,
      "loss": 5.1057,
      "mean_token_accuracy": 0.21314375400543212,
      "num_tokens": 3377033.0,
      "step": 4235
    },
    {
      "entropy": 5.269766187667846,
      "epoch": 1.1987560079163133,
      "grad_norm": 1.5,
      "learning_rate": 0.0004902052930530975,
      "loss": 5.1346,
      "mean_token_accuracy": 0.21162494868040085,
      "num_tokens": 3381011.0,
      "step": 4240
    },
    {
      "entropy": 5.163282871246338,
      "epoch": 1.2001696352841391,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004901752519906391,
      "loss": 5.0653,
      "mean_token_accuracy": 0.22274785488843918,
      "num_tokens": 3384960.0,
      "step": 4245
    },
    {
      "entropy": 4.956600713729858,
      "epoch": 1.201583262651965,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004901451659578836,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22998282164335251,
      "num_tokens": 3388888.0,
      "step": 4250
    },
    {
      "entropy": 5.043566560745239,
      "epoch": 1.2029968900197907,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004901150349611187,
      "loss": 5.0198,
      "mean_token_accuracy": 0.233669313788414,
      "num_tokens": 3392719.0,
      "step": 4255
    },
    {
      "entropy": 5.209820318222046,
      "epoch": 1.2044105173876165,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004900848590066419,
      "loss": 5.1593,
      "mean_token_accuracy": 0.2096559539437294,
      "num_tokens": 3396884.0,
      "step": 4260
    },
    {
      "entropy": 5.0651538372039795,
      "epoch": 1.2058241447554425,
      "grad_norm": 1.5,
      "learning_rate": 0.0004900546381007595,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21814290583133697,
      "num_tokens": 3400764.0,
      "step": 4265
    },
    {
      "entropy": 5.105000019073486,
      "epoch": 1.2072377721232683,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004900243722497877,
      "loss": 5.1274,
      "mean_token_accuracy": 0.21617622673511505,
      "num_tokens": 3405120.0,
      "step": 4270
    },
    {
      "entropy": 5.146544885635376,
      "epoch": 1.2086513994910941,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004899940614600517,
      "loss": 5.1362,
      "mean_token_accuracy": 0.22179561108350754,
      "num_tokens": 3409200.0,
      "step": 4275
    },
    {
      "entropy": 5.196063804626465,
      "epoch": 1.21006502685892,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004899637057378864,
      "loss": 5.1363,
      "mean_token_accuracy": 0.21705241799354552,
      "num_tokens": 3413360.0,
      "step": 4280
    },
    {
      "entropy": 5.130340671539306,
      "epoch": 1.2114786542267457,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004899333050896358,
      "loss": 5.2179,
      "mean_token_accuracy": 0.20770107656717302,
      "num_tokens": 3417445.0,
      "step": 4285
    },
    {
      "entropy": 5.1846911907196045,
      "epoch": 1.2128922815945717,
      "grad_norm": 1.5,
      "learning_rate": 0.0004899028595216536,
      "loss": 5.0929,
      "mean_token_accuracy": 0.21372957080602645,
      "num_tokens": 3421289.0,
      "step": 4290
    },
    {
      "entropy": 5.084757900238037,
      "epoch": 1.2143059089623975,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004898723690403027,
      "loss": 5.016,
      "mean_token_accuracy": 0.22967990785837172,
      "num_tokens": 3425211.0,
      "step": 4295
    },
    {
      "entropy": 5.124425601959229,
      "epoch": 1.2157195363302233,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004898418336519553,
      "loss": 5.1838,
      "mean_token_accuracy": 0.21873701810836793,
      "num_tokens": 3429101.0,
      "step": 4300
    },
    {
      "entropy": 5.172448253631591,
      "epoch": 1.2171331636980491,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004898112533629933,
      "loss": 5.1345,
      "mean_token_accuracy": 0.21863100528717042,
      "num_tokens": 3433334.0,
      "step": 4305
    },
    {
      "entropy": 5.107789945602417,
      "epoch": 1.218546791065875,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004897806281798076,
      "loss": 5.13,
      "mean_token_accuracy": 0.22332694232463837,
      "num_tokens": 3437044.0,
      "step": 4310
    },
    {
      "entropy": 5.226331567764282,
      "epoch": 1.219960418433701,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004897499581087989,
      "loss": 5.15,
      "mean_token_accuracy": 0.21030321419239045,
      "num_tokens": 3440979.0,
      "step": 4315
    },
    {
      "entropy": 5.241788053512574,
      "epoch": 1.2213740458015268,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000489719243156377,
      "loss": 5.1948,
      "mean_token_accuracy": 0.2124111086130142,
      "num_tokens": 3445285.0,
      "step": 4320
    },
    {
      "entropy": 5.29642882347107,
      "epoch": 1.2227876731693526,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004896884833289609,
      "loss": 5.1629,
      "mean_token_accuracy": 0.2191924661397934,
      "num_tokens": 3449460.0,
      "step": 4325
    },
    {
      "entropy": 5.318815135955811,
      "epoch": 1.2242013005371783,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004896576786329794,
      "loss": 4.9657,
      "mean_token_accuracy": 0.22279219180345536,
      "num_tokens": 3453193.0,
      "step": 4330
    },
    {
      "entropy": 5.162304639816284,
      "epoch": 1.2256149279050041,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004896268290748704,
      "loss": 5.1979,
      "mean_token_accuracy": 0.2151190981268883,
      "num_tokens": 3457099.0,
      "step": 4335
    },
    {
      "entropy": 5.142337989807129,
      "epoch": 1.2270285552728302,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004895959346610815,
      "loss": 5.1618,
      "mean_token_accuracy": 0.21533561050891875,
      "num_tokens": 3461370.0,
      "step": 4340
    },
    {
      "entropy": 5.346613645553589,
      "epoch": 1.228442182640656,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004895649953980691,
      "loss": 5.1758,
      "mean_token_accuracy": 0.20752868354320525,
      "num_tokens": 3465663.0,
      "step": 4345
    },
    {
      "entropy": 5.180907678604126,
      "epoch": 1.2298558100084818,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004895340112922997,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2154103234410286,
      "num_tokens": 3469304.0,
      "step": 4350
    },
    {
      "entropy": 5.0228759288787845,
      "epoch": 1.2312694373763076,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004895029823502485,
      "loss": 5.0159,
      "mean_token_accuracy": 0.23500039279460908,
      "num_tokens": 3473199.0,
      "step": 4355
    },
    {
      "entropy": 5.107254934310913,
      "epoch": 1.2326830647441334,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004894719085784005,
      "loss": 4.9805,
      "mean_token_accuracy": 0.2298651322722435,
      "num_tokens": 3476989.0,
      "step": 4360
    },
    {
      "entropy": 5.173661375045777,
      "epoch": 1.2340966921119594,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004894407899832497,
      "loss": 5.1225,
      "mean_token_accuracy": 0.21907873451709747,
      "num_tokens": 3481123.0,
      "step": 4365
    },
    {
      "entropy": 5.008401584625244,
      "epoch": 1.2355103194797852,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004894096265713,
      "loss": 5.0057,
      "mean_token_accuracy": 0.22622695714235305,
      "num_tokens": 3484963.0,
      "step": 4370
    },
    {
      "entropy": 4.934173917770385,
      "epoch": 1.236923946847611,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000489378418349064,
      "loss": 5.0084,
      "mean_token_accuracy": 0.23443454802036284,
      "num_tokens": 3488754.0,
      "step": 4375
    },
    {
      "entropy": 5.064777660369873,
      "epoch": 1.2383375742154368,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004893471653230644,
      "loss": 4.9755,
      "mean_token_accuracy": 0.23205506652593613,
      "num_tokens": 3493147.0,
      "step": 4380
    },
    {
      "entropy": 5.19450192451477,
      "epoch": 1.2397512015832626,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004893158674998328,
      "loss": 5.0939,
      "mean_token_accuracy": 0.22253225296735762,
      "num_tokens": 3497217.0,
      "step": 4385
    },
    {
      "entropy": 5.10273380279541,
      "epoch": 1.2411648289510886,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004892845248859101,
      "loss": 5.1101,
      "mean_token_accuracy": 0.22393780052661896,
      "num_tokens": 3500948.0,
      "step": 4390
    },
    {
      "entropy": 5.16806640625,
      "epoch": 1.2425784563189144,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004892531374878468,
      "loss": 5.1672,
      "mean_token_accuracy": 0.22861336171627045,
      "num_tokens": 3504997.0,
      "step": 4395
    },
    {
      "entropy": 5.138900423049927,
      "epoch": 1.2439920836867402,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004892217053122026,
      "loss": 5.0292,
      "mean_token_accuracy": 0.2175416350364685,
      "num_tokens": 3509198.0,
      "step": 4400
    },
    {
      "entropy": 5.179997158050537,
      "epoch": 1.245405711054566,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004891902283655468,
      "loss": 5.2484,
      "mean_token_accuracy": 0.2087549686431885,
      "num_tokens": 3513002.0,
      "step": 4405
    },
    {
      "entropy": 5.211358213424683,
      "epoch": 1.2468193384223918,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004891587066544577,
      "loss": 5.0695,
      "mean_token_accuracy": 0.219163279235363,
      "num_tokens": 3516941.0,
      "step": 4410
    },
    {
      "entropy": 5.149922132492065,
      "epoch": 1.2482329657902178,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004891271401855232,
      "loss": 5.0816,
      "mean_token_accuracy": 0.21181935966014862,
      "num_tokens": 3520788.0,
      "step": 4415
    },
    {
      "entropy": 5.1027655601501465,
      "epoch": 1.2496465931580436,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004890955289653407,
      "loss": 5.2198,
      "mean_token_accuracy": 0.2137544795870781,
      "num_tokens": 3524690.0,
      "step": 4420
    },
    {
      "entropy": 5.127509832382202,
      "epoch": 1.2510602205258694,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004890638730005164,
      "loss": 5.0799,
      "mean_token_accuracy": 0.23154484629631042,
      "num_tokens": 3528826.0,
      "step": 4425
    },
    {
      "entropy": 5.172293424606323,
      "epoch": 1.2524738478936952,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004890321722976665,
      "loss": 5.0708,
      "mean_token_accuracy": 0.2198789343237877,
      "num_tokens": 3532660.0,
      "step": 4430
    },
    {
      "entropy": 5.167311763763427,
      "epoch": 1.253887475261521,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000489000426863416,
      "loss": 5.1099,
      "mean_token_accuracy": 0.21733296066522598,
      "num_tokens": 3536635.0,
      "step": 4435
    },
    {
      "entropy": 5.111905336380005,
      "epoch": 1.255301102629347,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004889686367043997,
      "loss": 5.1002,
      "mean_token_accuracy": 0.22555067539215087,
      "num_tokens": 3540468.0,
      "step": 4440
    },
    {
      "entropy": 5.074908351898193,
      "epoch": 1.2567147299971728,
      "grad_norm": 1.5,
      "learning_rate": 0.0004889368018272614,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2276523619890213,
      "num_tokens": 3544748.0,
      "step": 4445
    },
    {
      "entropy": 5.157138061523438,
      "epoch": 1.2581283573649986,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004889049222386547,
      "loss": 5.068,
      "mean_token_accuracy": 0.2201649382710457,
      "num_tokens": 3548766.0,
      "step": 4450
    },
    {
      "entropy": 5.2251551151275635,
      "epoch": 1.2595419847328244,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004888729979452417,
      "loss": 5.0935,
      "mean_token_accuracy": 0.21711475104093553,
      "num_tokens": 3553137.0,
      "step": 4455
    },
    {
      "entropy": 5.181662368774414,
      "epoch": 1.2609556121006502,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004888410289536949,
      "loss": 5.156,
      "mean_token_accuracy": 0.212474425137043,
      "num_tokens": 3557029.0,
      "step": 4460
    },
    {
      "entropy": 5.0712597370147705,
      "epoch": 1.2623692394684762,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004888090152706953,
      "loss": 4.9949,
      "mean_token_accuracy": 0.23251367211341858,
      "num_tokens": 3560919.0,
      "step": 4465
    },
    {
      "entropy": 5.05785059928894,
      "epoch": 1.263782866836302,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004887769569029336,
      "loss": 5.0579,
      "mean_token_accuracy": 0.21858127862215043,
      "num_tokens": 3565096.0,
      "step": 4470
    },
    {
      "entropy": 5.112567615509033,
      "epoch": 1.2651964942041278,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048874485385711,
      "loss": 4.9734,
      "mean_token_accuracy": 0.2266402766108513,
      "num_tokens": 3568780.0,
      "step": 4475
    },
    {
      "entropy": 5.105242919921875,
      "epoch": 1.2666101215719536,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004887127061399336,
      "loss": 5.0693,
      "mean_token_accuracy": 0.22039632201194764,
      "num_tokens": 3572743.0,
      "step": 4480
    },
    {
      "entropy": 5.079741287231445,
      "epoch": 1.2680237489397794,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004886805137581232,
      "loss": 5.1355,
      "mean_token_accuracy": 0.22051983177661896,
      "num_tokens": 3576721.0,
      "step": 4485
    },
    {
      "entropy": 5.054748249053955,
      "epoch": 1.2694373763076054,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004886482767184068,
      "loss": 4.9108,
      "mean_token_accuracy": 0.23976295739412307,
      "num_tokens": 3580388.0,
      "step": 4490
    },
    {
      "entropy": 5.157027721405029,
      "epoch": 1.270851003675431,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004886159950275217,
      "loss": 5.1384,
      "mean_token_accuracy": 0.22766706645488738,
      "num_tokens": 3584249.0,
      "step": 4495
    },
    {
      "entropy": 5.150328350067139,
      "epoch": 1.272264631043257,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004885836686922147,
      "loss": 5.1027,
      "mean_token_accuracy": 0.2279328241944313,
      "num_tokens": 3588539.0,
      "step": 4500
    },
    {
      "entropy": 5.103603506088257,
      "epoch": 1.2736782584110828,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004885512977192416,
      "loss": 5.1589,
      "mean_token_accuracy": 0.2213436871767044,
      "num_tokens": 3592756.0,
      "step": 4505
    },
    {
      "entropy": 5.153210258483886,
      "epoch": 1.2750918857789086,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004885188821153678,
      "loss": 4.9926,
      "mean_token_accuracy": 0.22525349259376526,
      "num_tokens": 3596569.0,
      "step": 4510
    },
    {
      "entropy": 5.142386102676392,
      "epoch": 1.2765055131467344,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000488486421887368,
      "loss": 5.1745,
      "mean_token_accuracy": 0.2156160518527031,
      "num_tokens": 3600541.0,
      "step": 4515
    },
    {
      "entropy": 5.163817882537842,
      "epoch": 1.2779191405145602,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004884539170420263,
      "loss": 5.0991,
      "mean_token_accuracy": 0.22168633788824083,
      "num_tokens": 3604567.0,
      "step": 4520
    },
    {
      "entropy": 5.1744026184082035,
      "epoch": 1.2793327678823863,
      "grad_norm": 1.484375,
      "learning_rate": 0.000488421367586136,
      "loss": 5.1481,
      "mean_token_accuracy": 0.21416470408439636,
      "num_tokens": 3608528.0,
      "step": 4525
    },
    {
      "entropy": 5.19086480140686,
      "epoch": 1.280746395250212,
      "grad_norm": 1.625,
      "learning_rate": 0.0004883887735264993,
      "loss": 5.1753,
      "mean_token_accuracy": 0.21246986836194992,
      "num_tokens": 3612408.0,
      "step": 4530
    },
    {
      "entropy": 5.0976402282714846,
      "epoch": 1.2821600226180379,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004883561348699287,
      "loss": 5.0165,
      "mean_token_accuracy": 0.2313650667667389,
      "num_tokens": 3616176.0,
      "step": 4535
    },
    {
      "entropy": 5.098331642150879,
      "epoch": 1.2835736499858637,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004883234516232452,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2282039165496826,
      "num_tokens": 3620446.0,
      "step": 4540
    },
    {
      "entropy": 5.137396621704101,
      "epoch": 1.2849872773536894,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004882907237932796,
      "loss": 5.0599,
      "mean_token_accuracy": 0.22394090741872788,
      "num_tokens": 3624437.0,
      "step": 4545
    },
    {
      "entropy": 5.179127550125122,
      "epoch": 1.2864009047215155,
      "grad_norm": 1.5,
      "learning_rate": 0.0004882579513868716,
      "loss": 5.1149,
      "mean_token_accuracy": 0.22676489800214766,
      "num_tokens": 3628418.0,
      "step": 4550
    },
    {
      "entropy": 5.144613933563233,
      "epoch": 1.2878145320893413,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048822513441087046,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22211112678050995,
      "num_tokens": 3632237.0,
      "step": 4555
    },
    {
      "entropy": 5.086827611923217,
      "epoch": 1.289228159457167,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004881922728721348,
      "loss": 5.0959,
      "mean_token_accuracy": 0.21826667934656144,
      "num_tokens": 3636275.0,
      "step": 4560
    },
    {
      "entropy": 5.039886045455932,
      "epoch": 1.2906417868249929,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048815936677753244,
      "loss": 5.1328,
      "mean_token_accuracy": 0.2117163583636284,
      "num_tokens": 3640319.0,
      "step": 4565
    },
    {
      "entropy": 5.211140918731689,
      "epoch": 1.2920554141928187,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048812641613394053,
      "loss": 5.0855,
      "mean_token_accuracy": 0.21473707407712936,
      "num_tokens": 3644200.0,
      "step": 4570
    },
    {
      "entropy": 5.008463764190674,
      "epoch": 1.2934690415606447,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004880934209482457,
      "loss": 5.0104,
      "mean_token_accuracy": 0.22811658382415773,
      "num_tokens": 3648314.0,
      "step": 4575
    },
    {
      "entropy": 5.1141821384429935,
      "epoch": 1.2948826689284705,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048806038122734354,
      "loss": 5.18,
      "mean_token_accuracy": 0.2087305799126625,
      "num_tokens": 3652241.0,
      "step": 4580
    },
    {
      "entropy": 5.253288507461548,
      "epoch": 1.2962962962962963,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048802729697813916,
      "loss": 5.1646,
      "mean_token_accuracy": 0.2080194264650345,
      "num_tokens": 3656162.0,
      "step": 4585
    },
    {
      "entropy": 5.202285861968994,
      "epoch": 1.297709923664122,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004879941682075472,
      "loss": 5.0875,
      "mean_token_accuracy": 0.22350511997938155,
      "num_tokens": 3660168.0,
      "step": 4590
    },
    {
      "entropy": 5.218620491027832,
      "epoch": 1.2991235510319479,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004879609949224911,
      "loss": 5.1903,
      "mean_token_accuracy": 0.2105267435312271,
      "num_tokens": 3664299.0,
      "step": 4595
    },
    {
      "entropy": 5.269051837921142,
      "epoch": 1.300537178399774,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048792777712990397,
      "loss": 5.1422,
      "mean_token_accuracy": 0.21296022832393646,
      "num_tokens": 3668054.0,
      "step": 4600
    },
    {
      "entropy": 5.237417268753052,
      "epoch": 1.3019508057675997,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004878945148367281,
      "loss": 5.0813,
      "mean_token_accuracy": 0.2233643040060997,
      "num_tokens": 3672161.0,
      "step": 4605
    },
    {
      "entropy": 5.297582340240479,
      "epoch": 1.3033644331354255,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048786120804991526,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20347945541143417,
      "num_tokens": 3676040.0,
      "step": 4610
    },
    {
      "entropy": 5.266462564468384,
      "epoch": 1.3047780605032513,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048782785677642604,
      "loss": 5.1696,
      "mean_token_accuracy": 0.21635618954896926,
      "num_tokens": 3680251.0,
      "step": 4615
    },
    {
      "entropy": 5.175524950027466,
      "epoch": 1.306191687871077,
      "grad_norm": 1.546875,
      "learning_rate": 0.000487794461023231,
      "loss": 5.0305,
      "mean_token_accuracy": 0.22404103577136994,
      "num_tokens": 3684194.0,
      "step": 4620
    },
    {
      "entropy": 5.093195199966431,
      "epoch": 1.307605315238903,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00048776102079730937,
      "loss": 5.0949,
      "mean_token_accuracy": 0.21288907527923584,
      "num_tokens": 3688224.0,
      "step": 4625
    },
    {
      "entropy": 5.097847127914429,
      "epoch": 1.309018942606729,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048772753610565014,
      "loss": 5.0997,
      "mean_token_accuracy": 0.2269148662686348,
      "num_tokens": 3691956.0,
      "step": 4630
    },
    {
      "entropy": 5.128762197494507,
      "epoch": 1.3104325699745547,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004876940069552513,
      "loss": 5.0301,
      "mean_token_accuracy": 0.23561992049217223,
      "num_tokens": 3696183.0,
      "step": 4635
    },
    {
      "entropy": 5.191114139556885,
      "epoch": 1.3118461973423805,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048766043335312026,
      "loss": 5.0796,
      "mean_token_accuracy": 0.21269580274820327,
      "num_tokens": 3699886.0,
      "step": 4640
    },
    {
      "entropy": 5.150349426269531,
      "epoch": 1.3132598247102063,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004876268153062738,
      "loss": 5.1503,
      "mean_token_accuracy": 0.2194637417793274,
      "num_tokens": 3703885.0,
      "step": 4645
    },
    {
      "entropy": 5.1658453941345215,
      "epoch": 1.3146734520780323,
      "grad_norm": 1.53125,
      "learning_rate": 0.00048759315282173766,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21020454466342925,
      "num_tokens": 3707493.0,
      "step": 4650
    },
    {
      "entropy": 5.091509294509888,
      "epoch": 1.3160870794458581,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004875594459065473,
      "loss": 5.1257,
      "mean_token_accuracy": 0.21664399802684783,
      "num_tokens": 3711466.0,
      "step": 4655
    },
    {
      "entropy": 5.052444744110107,
      "epoch": 1.317500706813684,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048752569456774717,
      "loss": 5.0692,
      "mean_token_accuracy": 0.21860056817531587,
      "num_tokens": 3715270.0,
      "step": 4660
    },
    {
      "entropy": 5.206305599212646,
      "epoch": 1.3189143341815097,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048749189881239105,
      "loss": 5.1226,
      "mean_token_accuracy": 0.2205084189772606,
      "num_tokens": 3719202.0,
      "step": 4665
    },
    {
      "entropy": 5.005170726776123,
      "epoch": 1.3203279615493355,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048745805864754214,
      "loss": 4.9586,
      "mean_token_accuracy": 0.23939103931188582,
      "num_tokens": 3723197.0,
      "step": 4670
    },
    {
      "entropy": 5.074552249908447,
      "epoch": 1.3217415889171615,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004874241740802727,
      "loss": 5.0567,
      "mean_token_accuracy": 0.22182919532060624,
      "num_tokens": 3727528.0,
      "step": 4675
    },
    {
      "entropy": 5.096038007736206,
      "epoch": 1.3231552162849873,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004873902451176645,
      "loss": 4.9954,
      "mean_token_accuracy": 0.23075418323278427,
      "num_tokens": 3731410.0,
      "step": 4680
    },
    {
      "entropy": 5.165931987762451,
      "epoch": 1.3245688436528131,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004873562717668084,
      "loss": 5.2019,
      "mean_token_accuracy": 0.22121675908565522,
      "num_tokens": 3735694.0,
      "step": 4685
    },
    {
      "entropy": 5.098274230957031,
      "epoch": 1.325982471020639,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004873222540348048,
      "loss": 5.1502,
      "mean_token_accuracy": 0.21669952869415282,
      "num_tokens": 3739836.0,
      "step": 4690
    },
    {
      "entropy": 5.045234298706054,
      "epoch": 1.3273960983884647,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004872881919287629,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2206188827753067,
      "num_tokens": 3743748.0,
      "step": 4695
    },
    {
      "entropy": 5.172725963592529,
      "epoch": 1.3288097257562907,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004872540854558018,
      "loss": 5.0552,
      "mean_token_accuracy": 0.22823820263147354,
      "num_tokens": 3748034.0,
      "step": 4700
    },
    {
      "entropy": 5.19414496421814,
      "epoch": 1.3302233531241165,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004872199346230493,
      "loss": 5.0234,
      "mean_token_accuracy": 0.22397832274436952,
      "num_tokens": 3751887.0,
      "step": 4705
    },
    {
      "entropy": 5.229147863388062,
      "epoch": 1.3316369804919423,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048718573943764285,
      "loss": 5.2127,
      "mean_token_accuracy": 0.20542629659175873,
      "num_tokens": 3756305.0,
      "step": 4710
    },
    {
      "entropy": 5.234424591064453,
      "epoch": 1.3330506078597681,
      "grad_norm": 1.4375,
      "learning_rate": 0.00048715149990672887,
      "loss": 5.1302,
      "mean_token_accuracy": 0.21097299009561538,
      "num_tokens": 3760139.0,
      "step": 4715
    },
    {
      "entropy": 5.241721963882446,
      "epoch": 1.334464235227594,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004871172160374634,
      "loss": 5.067,
      "mean_token_accuracy": 0.22382923811674119,
      "num_tokens": 3763966.0,
      "step": 4720
    },
    {
      "entropy": 5.063567209243774,
      "epoch": 1.33587786259542,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048708288783701144,
      "loss": 5.0537,
      "mean_token_accuracy": 0.22904675006866454,
      "num_tokens": 3767902.0,
      "step": 4725
    },
    {
      "entropy": 5.065946817398071,
      "epoch": 1.3372914899632458,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004870485153125474,
      "loss": 4.9658,
      "mean_token_accuracy": 0.22691674530506134,
      "num_tokens": 3771760.0,
      "step": 4730
    },
    {
      "entropy": 5.133263206481933,
      "epoch": 1.3387051173310716,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048701409847125494,
      "loss": 5.0657,
      "mean_token_accuracy": 0.2123713821172714,
      "num_tokens": 3775543.0,
      "step": 4735
    },
    {
      "entropy": 5.046080636978149,
      "epoch": 1.3401187446988974,
      "grad_norm": 1.359375,
      "learning_rate": 0.00048697963732032695,
      "loss": 5.0245,
      "mean_token_accuracy": 0.22067976742982864,
      "num_tokens": 3779950.0,
      "step": 4740
    },
    {
      "entropy": 5.085481071472168,
      "epoch": 1.3415323720667232,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00048694513186696566,
      "loss": 5.0601,
      "mean_token_accuracy": 0.22289158701896666,
      "num_tokens": 3784012.0,
      "step": 4745
    },
    {
      "entropy": 5.112040328979492,
      "epoch": 1.3429459994345492,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004869105821183824,
      "loss": 5.156,
      "mean_token_accuracy": 0.21847089529037475,
      "num_tokens": 3787973.0,
      "step": 4750
    },
    {
      "entropy": 5.133311796188354,
      "epoch": 1.344359626802375,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048687598808179786,
      "loss": 5.0385,
      "mean_token_accuracy": 0.22601410746574402,
      "num_tokens": 3791576.0,
      "step": 4755
    },
    {
      "entropy": 5.264501619338989,
      "epoch": 1.3457732541702008,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048684134976444203,
      "loss": 5.1829,
      "mean_token_accuracy": 0.2132762148976326,
      "num_tokens": 3795495.0,
      "step": 4760
    },
    {
      "entropy": 5.141721534729004,
      "epoch": 1.3471868815380266,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004868066671735541,
      "loss": 5.1006,
      "mean_token_accuracy": 0.22058866173028946,
      "num_tokens": 3799477.0,
      "step": 4765
    },
    {
      "entropy": 5.099866247177124,
      "epoch": 1.3486005089058524,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004867719403163825,
      "loss": 5.1697,
      "mean_token_accuracy": 0.21214720755815505,
      "num_tokens": 3803305.0,
      "step": 4770
    },
    {
      "entropy": 5.0876692771911625,
      "epoch": 1.3500141362736784,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004867371692001849,
      "loss": 5.0475,
      "mean_token_accuracy": 0.22982362061738967,
      "num_tokens": 3807374.0,
      "step": 4775
    },
    {
      "entropy": 5.012311220169067,
      "epoch": 1.351427763641504,
      "grad_norm": 1.5,
      "learning_rate": 0.0004867023538322283,
      "loss": 5.0124,
      "mean_token_accuracy": 0.22722697854042054,
      "num_tokens": 3811103.0,
      "step": 4780
    },
    {
      "entropy": 5.043846559524536,
      "epoch": 1.35284139100933,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004866674942197889,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22743183076381684,
      "num_tokens": 3814947.0,
      "step": 4785
    },
    {
      "entropy": 5.193683624267578,
      "epoch": 1.3542550183771558,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048663259037015204,
      "loss": 5.1306,
      "mean_token_accuracy": 0.2282833054661751,
      "num_tokens": 3818945.0,
      "step": 4790
    },
    {
      "entropy": 5.085645008087158,
      "epoch": 1.3556686457449816,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048659764229061264,
      "loss": 4.9962,
      "mean_token_accuracy": 0.23509571254253386,
      "num_tokens": 3822852.0,
      "step": 4795
    },
    {
      "entropy": 5.172564506530762,
      "epoch": 1.3570822731128076,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048656264998847435,
      "loss": 5.1353,
      "mean_token_accuracy": 0.216292904317379,
      "num_tokens": 3826663.0,
      "step": 4800
    },
    {
      "entropy": 5.069854831695556,
      "epoch": 1.3584959004806332,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004865276134710505,
      "loss": 5.0229,
      "mean_token_accuracy": 0.23079361021518707,
      "num_tokens": 3830585.0,
      "step": 4805
    },
    {
      "entropy": 5.054951953887939,
      "epoch": 1.3599095278484592,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004864925327456635,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22875631749629974,
      "num_tokens": 3834432.0,
      "step": 4810
    },
    {
      "entropy": 5.029392957687378,
      "epoch": 1.361323155216285,
      "grad_norm": 1.40625,
      "learning_rate": 0.00048645740781964506,
      "loss": 5.0756,
      "mean_token_accuracy": 0.22895398885011672,
      "num_tokens": 3838416.0,
      "step": 4815
    },
    {
      "entropy": 5.265322828292847,
      "epoch": 1.3627367825841108,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048642223870033584,
      "loss": 5.2281,
      "mean_token_accuracy": 0.2046325609087944,
      "num_tokens": 3842360.0,
      "step": 4820
    },
    {
      "entropy": 5.290566778182983,
      "epoch": 1.3641504099519366,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004863870253950863,
      "loss": 5.1176,
      "mean_token_accuracy": 0.20970345735549928,
      "num_tokens": 3846496.0,
      "step": 4825
    },
    {
      "entropy": 5.1237067699432375,
      "epoch": 1.3655640373197624,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004863517679112555,
      "loss": 5.0671,
      "mean_token_accuracy": 0.2383072406053543,
      "num_tokens": 3850550.0,
      "step": 4830
    },
    {
      "entropy": 5.00981912612915,
      "epoch": 1.3669776646875884,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004863164662562122,
      "loss": 5.1151,
      "mean_token_accuracy": 0.2137703388929367,
      "num_tokens": 3854751.0,
      "step": 4835
    },
    {
      "entropy": 5.089965152740478,
      "epoch": 1.3683912920554142,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004862811204373342,
      "loss": 5.1043,
      "mean_token_accuracy": 0.21813636720180513,
      "num_tokens": 3858554.0,
      "step": 4840
    },
    {
      "entropy": 5.052470684051514,
      "epoch": 1.36980491942324,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004862457304620086,
      "loss": 4.9847,
      "mean_token_accuracy": 0.2297527864575386,
      "num_tokens": 3862312.0,
      "step": 4845
    },
    {
      "entropy": 5.108477210998535,
      "epoch": 1.3712185467910658,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048621029633763157,
      "loss": 5.0924,
      "mean_token_accuracy": 0.2256404548883438,
      "num_tokens": 3866595.0,
      "step": 4850
    },
    {
      "entropy": 5.103623342514038,
      "epoch": 1.3726321741588916,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048617481807160874,
      "loss": 5.0723,
      "mean_token_accuracy": 0.23027922511100768,
      "num_tokens": 3870863.0,
      "step": 4855
    },
    {
      "entropy": 5.144025278091431,
      "epoch": 1.3740458015267176,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004861392956713547,
      "loss": 5.1807,
      "mean_token_accuracy": 0.21710853576660155,
      "num_tokens": 3874797.0,
      "step": 4860
    },
    {
      "entropy": 5.08786792755127,
      "epoch": 1.3754594288945434,
      "grad_norm": 1.375,
      "learning_rate": 0.00048610372914429365,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2201899215579033,
      "num_tokens": 3878965.0,
      "step": 4865
    },
    {
      "entropy": 5.078768396377564,
      "epoch": 1.3768730562623692,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004860681184978586,
      "loss": 5.0784,
      "mean_token_accuracy": 0.22701516896486282,
      "num_tokens": 3883299.0,
      "step": 4870
    },
    {
      "entropy": 5.27328519821167,
      "epoch": 1.378286683630195,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048603246373949197,
      "loss": 5.1893,
      "mean_token_accuracy": 0.2095549076795578,
      "num_tokens": 3887616.0,
      "step": 4875
    },
    {
      "entropy": 5.224088525772094,
      "epoch": 1.3797003109980208,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048599676487664535,
      "loss": 5.1613,
      "mean_token_accuracy": 0.21445799469947815,
      "num_tokens": 3892043.0,
      "step": 4880
    },
    {
      "entropy": 5.049299240112305,
      "epoch": 1.3811139383658468,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004859610219167797,
      "loss": 5.0531,
      "mean_token_accuracy": 0.23092842251062393,
      "num_tokens": 3896385.0,
      "step": 4885
    },
    {
      "entropy": 5.148837471008301,
      "epoch": 1.3825275657336726,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000485925234867365,
      "loss": 5.1062,
      "mean_token_accuracy": 0.21084923148155213,
      "num_tokens": 3900277.0,
      "step": 4890
    },
    {
      "entropy": 5.056463050842285,
      "epoch": 1.3839411931014984,
      "grad_norm": 1.65625,
      "learning_rate": 0.00048588940373588054,
      "loss": 4.9417,
      "mean_token_accuracy": 0.24435820430517197,
      "num_tokens": 3903653.0,
      "step": 4895
    },
    {
      "entropy": 5.121997928619384,
      "epoch": 1.3853548204693242,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004858535285298148,
      "loss": 5.0884,
      "mean_token_accuracy": 0.22751763314008713,
      "num_tokens": 3908003.0,
      "step": 4900
    },
    {
      "entropy": 5.013748836517334,
      "epoch": 1.38676844783715,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048581760925666545,
      "loss": 5.0429,
      "mean_token_accuracy": 0.22613221108913423,
      "num_tokens": 3912120.0,
      "step": 4905
    },
    {
      "entropy": 5.232112646102905,
      "epoch": 1.388182075204976,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004857816459239394,
      "loss": 5.183,
      "mean_token_accuracy": 0.2071456015110016,
      "num_tokens": 3916078.0,
      "step": 4910
    },
    {
      "entropy": 5.280574893951416,
      "epoch": 1.3895957025728018,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004857456385391528,
      "loss": 5.1772,
      "mean_token_accuracy": 0.20157544165849686,
      "num_tokens": 3920214.0,
      "step": 4915
    },
    {
      "entropy": 5.170943260192871,
      "epoch": 1.3910093299406276,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000485709587109831,
      "loss": 5.0183,
      "mean_token_accuracy": 0.21981197148561477,
      "num_tokens": 3924092.0,
      "step": 4920
    },
    {
      "entropy": 5.064088535308838,
      "epoch": 1.3924229573084534,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004856734916435085,
      "loss": 5.1128,
      "mean_token_accuracy": 0.23114604502916336,
      "num_tokens": 3927999.0,
      "step": 4925
    },
    {
      "entropy": 5.144844341278076,
      "epoch": 1.3938365846762792,
      "grad_norm": 1.625,
      "learning_rate": 0.00048563735214772897,
      "loss": 5.0438,
      "mean_token_accuracy": 0.23160787373781205,
      "num_tokens": 3931804.0,
      "step": 4930
    },
    {
      "entropy": 5.261448955535888,
      "epoch": 1.3952502120441053,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048560116863004537,
      "loss": 5.1072,
      "mean_token_accuracy": 0.2227345213294029,
      "num_tokens": 3935569.0,
      "step": 4935
    },
    {
      "entropy": 5.144621706008911,
      "epoch": 1.396663839411931,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048556494109801984,
      "loss": 5.0569,
      "mean_token_accuracy": 0.2245674028992653,
      "num_tokens": 3939450.0,
      "step": 4940
    },
    {
      "entropy": 5.076720666885376,
      "epoch": 1.3980774667797569,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004855286695592238,
      "loss": 5.0453,
      "mean_token_accuracy": 0.22329609841108322,
      "num_tokens": 3943718.0,
      "step": 4945
    },
    {
      "entropy": 5.222533178329468,
      "epoch": 1.3994910941475827,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004854923540212376,
      "loss": 5.0902,
      "mean_token_accuracy": 0.23582245111465455,
      "num_tokens": 3947558.0,
      "step": 4950
    },
    {
      "entropy": 5.173509693145752,
      "epoch": 1.4009047215154085,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048545599449165106,
      "loss": 5.0225,
      "mean_token_accuracy": 0.2307034835219383,
      "num_tokens": 3951642.0,
      "step": 4955
    },
    {
      "entropy": 5.065652084350586,
      "epoch": 1.4023183488832345,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048541959097806315,
      "loss": 5.1554,
      "mean_token_accuracy": 0.22711582183837892,
      "num_tokens": 3956038.0,
      "step": 4960
    },
    {
      "entropy": 5.214606523513794,
      "epoch": 1.4037319762510603,
      "grad_norm": 1.625,
      "learning_rate": 0.00048538314348808186,
      "loss": 5.1031,
      "mean_token_accuracy": 0.21635560989379882,
      "num_tokens": 3959855.0,
      "step": 4965
    },
    {
      "entropy": 5.205858039855957,
      "epoch": 1.405145603618886,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004853466520293245,
      "loss": 5.0815,
      "mean_token_accuracy": 0.21070680618286133,
      "num_tokens": 3963819.0,
      "step": 4970
    },
    {
      "entropy": 5.0769819736480715,
      "epoch": 1.4065592309867119,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048531011660941784,
      "loss": 5.0072,
      "mean_token_accuracy": 0.22707393616437913,
      "num_tokens": 3967526.0,
      "step": 4975
    },
    {
      "entropy": 4.982313728332519,
      "epoch": 1.4079728583545377,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048527353723599716,
      "loss": 5.0868,
      "mean_token_accuracy": 0.22295337915420532,
      "num_tokens": 3971432.0,
      "step": 4980
    },
    {
      "entropy": 5.07086443901062,
      "epoch": 1.4093864857223637,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004852369139167075,
      "loss": 4.9762,
      "mean_token_accuracy": 0.24008368104696273,
      "num_tokens": 3975225.0,
      "step": 4985
    },
    {
      "entropy": 5.23497428894043,
      "epoch": 1.4108001130901895,
      "grad_norm": 1.59375,
      "learning_rate": 0.00048520024665920286,
      "loss": 5.0713,
      "mean_token_accuracy": 0.22008075714111328,
      "num_tokens": 3979265.0,
      "step": 4990
    },
    {
      "entropy": 5.194688606262207,
      "epoch": 1.4122137404580153,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004851635354711466,
      "loss": 5.0974,
      "mean_token_accuracy": 0.21924781799316406,
      "num_tokens": 3983413.0,
      "step": 4995
    },
    {
      "entropy": 5.1899094581604,
      "epoch": 1.413627367825841,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048512678036021097,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2295347958803177,
      "num_tokens": 3987304.0,
      "step": 5000
    },
    {
      "epoch": 1.413627367825841,
      "eval_entropy": 5.02764848746956,
      "eval_loss": 5.374653339385986,
      "eval_mean_token_accuracy": 0.2197306394988894,
      "eval_num_tokens": 3987304.0,
      "eval_runtime": 6.8261,
      "eval_samples_per_second": 1653.079,
      "eval_steps_per_second": 206.708,
      "step": 5000
    },
    {
      "entropy": 5.285630416870117,
      "epoch": 1.4150409951936669,
      "grad_norm": 1.3828125,
      "learning_rate": 0.00048508998133407766,
      "loss": 5.2723,
      "mean_token_accuracy": 0.2009644091129303,
      "num_tokens": 3991705.0,
      "step": 5005
    },
    {
      "entropy": 5.120524787902832,
      "epoch": 1.416454622561493,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004850531384004374,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22831859588623046,
      "num_tokens": 3995489.0,
      "step": 5010
    },
    {
      "entropy": 5.082024383544922,
      "epoch": 1.4178682499293187,
      "grad_norm": 1.625,
      "learning_rate": 0.00048501625156699006,
      "loss": 5.1176,
      "mean_token_accuracy": 0.22420299649238587,
      "num_tokens": 3999346.0,
      "step": 5015
    },
    {
      "entropy": 5.096896743774414,
      "epoch": 1.4192818772971445,
      "grad_norm": 1.609375,
      "learning_rate": 0.000484979320841445,
      "loss": 5.0973,
      "mean_token_accuracy": 0.2184542492032051,
      "num_tokens": 4003374.0,
      "step": 5020
    },
    {
      "entropy": 5.180984306335449,
      "epoch": 1.4206955046649703,
      "grad_norm": 1.390625,
      "learning_rate": 0.00048494234623152027,
      "loss": 5.1084,
      "mean_token_accuracy": 0.22045700997114182,
      "num_tokens": 4007527.0,
      "step": 5025
    },
    {
      "entropy": 5.0332879543304445,
      "epoch": 1.422109132032796,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048490532774494343,
      "loss": 4.88,
      "mean_token_accuracy": 0.24206820130348206,
      "num_tokens": 4011447.0,
      "step": 5030
    },
    {
      "entropy": 5.0520857810974125,
      "epoch": 1.4235227594006221,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004848682653894511,
      "loss": 4.96,
      "mean_token_accuracy": 0.2179910123348236,
      "num_tokens": 4015352.0,
      "step": 5035
    },
    {
      "entropy": 5.020318412780762,
      "epoch": 1.424936386768448,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048483115917278906,
      "loss": 5.1583,
      "mean_token_accuracy": 0.2114037722349167,
      "num_tokens": 4019468.0,
      "step": 5040
    },
    {
      "entropy": 5.080714559555053,
      "epoch": 1.4263500141362737,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048479400910271243,
      "loss": 5.08,
      "mean_token_accuracy": 0.22531761825084687,
      "num_tokens": 4023451.0,
      "step": 5045
    },
    {
      "entropy": 5.047916507720947,
      "epoch": 1.4277636415040995,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048475681518698517,
      "loss": 4.9607,
      "mean_token_accuracy": 0.24007776826620103,
      "num_tokens": 4027676.0,
      "step": 5050
    },
    {
      "entropy": 5.13837161064148,
      "epoch": 1.4291772688719253,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048471957743338056,
      "loss": 5.1463,
      "mean_token_accuracy": 0.21580678522586821,
      "num_tokens": 4031590.0,
      "step": 5055
    },
    {
      "entropy": 5.064213228225708,
      "epoch": 1.4305908962397513,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004846822958496812,
      "loss": 5.0001,
      "mean_token_accuracy": 0.2379024386405945,
      "num_tokens": 4035370.0,
      "step": 5060
    },
    {
      "entropy": 5.093661880493164,
      "epoch": 1.432004523607577,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00048464497044367865,
      "loss": 5.09,
      "mean_token_accuracy": 0.2248702824115753,
      "num_tokens": 4039145.0,
      "step": 5065
    },
    {
      "entropy": 5.184541034698486,
      "epoch": 1.433418150975403,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048460760122317364,
      "loss": 5.1641,
      "mean_token_accuracy": 0.22560527473688125,
      "num_tokens": 4042993.0,
      "step": 5070
    },
    {
      "entropy": 5.164267587661743,
      "epoch": 1.4348317783432287,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004845701881959762,
      "loss": 5.1188,
      "mean_token_accuracy": 0.22160816192626953,
      "num_tokens": 4047313.0,
      "step": 5075
    },
    {
      "entropy": 5.130391454696655,
      "epoch": 1.4362454057110545,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004845327313699053,
      "loss": 5.1443,
      "mean_token_accuracy": 0.21369738578796388,
      "num_tokens": 4051554.0,
      "step": 5080
    },
    {
      "entropy": 5.145473909378052,
      "epoch": 1.4376590330788805,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004844952307527893,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21098191738128663,
      "num_tokens": 4055470.0,
      "step": 5085
    },
    {
      "entropy": 5.080343103408813,
      "epoch": 1.4390726604467061,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048445768635246557,
      "loss": 5.0269,
      "mean_token_accuracy": 0.22245892584323884,
      "num_tokens": 4059352.0,
      "step": 5090
    },
    {
      "entropy": 5.116692781448364,
      "epoch": 1.4404862878145321,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004844200981767806,
      "loss": 5.1196,
      "mean_token_accuracy": 0.2258921444416046,
      "num_tokens": 4063523.0,
      "step": 5095
    },
    {
      "entropy": 5.208238840103149,
      "epoch": 1.441899915182358,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048438246623359014,
      "loss": 5.0802,
      "mean_token_accuracy": 0.21897760033607483,
      "num_tokens": 4067219.0,
      "step": 5100
    },
    {
      "entropy": 5.143711137771606,
      "epoch": 1.4433135425501837,
      "grad_norm": 1.453125,
      "learning_rate": 0.000484344790530759,
      "loss": 5.1243,
      "mean_token_accuracy": 0.22156484425067902,
      "num_tokens": 4071351.0,
      "step": 5105
    },
    {
      "entropy": 5.070819568634033,
      "epoch": 1.4447271699180095,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004843070710761613,
      "loss": 5.0817,
      "mean_token_accuracy": 0.22713102996349335,
      "num_tokens": 4075113.0,
      "step": 5110
    },
    {
      "entropy": 5.174178504943848,
      "epoch": 1.4461407972858353,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004842693078776799,
      "loss": 5.1311,
      "mean_token_accuracy": 0.21856046020984649,
      "num_tokens": 4079146.0,
      "step": 5115
    },
    {
      "entropy": 5.104150104522705,
      "epoch": 1.4475544246536614,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048423150094320733,
      "loss": 5.0259,
      "mean_token_accuracy": 0.23322485089302064,
      "num_tokens": 4083195.0,
      "step": 5120
    },
    {
      "entropy": 5.0273027420043945,
      "epoch": 1.4489680520214872,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004841936502806449,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2397835522890091,
      "num_tokens": 4087251.0,
      "step": 5125
    },
    {
      "entropy": 5.087057209014892,
      "epoch": 1.450381679389313,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048415575589790314,
      "loss": 5.093,
      "mean_token_accuracy": 0.22059950828552247,
      "num_tokens": 4091477.0,
      "step": 5130
    },
    {
      "entropy": 5.16743049621582,
      "epoch": 1.4517953067571387,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004841178178029018,
      "loss": 5.0831,
      "mean_token_accuracy": 0.22517478913068772,
      "num_tokens": 4095277.0,
      "step": 5135
    },
    {
      "entropy": 5.165636396408081,
      "epoch": 1.4532089341249645,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004840798360035697,
      "loss": 4.9819,
      "mean_token_accuracy": 0.232625612616539,
      "num_tokens": 4099157.0,
      "step": 5140
    },
    {
      "entropy": 5.18969841003418,
      "epoch": 1.4546225614927906,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048404181050784475,
      "loss": 5.0403,
      "mean_token_accuracy": 0.22336158752441407,
      "num_tokens": 4103183.0,
      "step": 5145
    },
    {
      "entropy": 5.170385551452637,
      "epoch": 1.4560361888606164,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048400374132367415,
      "loss": 5.0129,
      "mean_token_accuracy": 0.22831730395555497,
      "num_tokens": 4107265.0,
      "step": 5150
    },
    {
      "entropy": 5.096427774429321,
      "epoch": 1.4574498162284422,
      "grad_norm": 1.40625,
      "learning_rate": 0.000483965628459014,
      "loss": 4.9891,
      "mean_token_accuracy": 0.23821629136800765,
      "num_tokens": 4110919.0,
      "step": 5155
    },
    {
      "entropy": 5.0669008731842045,
      "epoch": 1.458863443596268,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004839274719218297,
      "loss": 4.9789,
      "mean_token_accuracy": 0.22245002686977386,
      "num_tokens": 4114808.0,
      "step": 5160
    },
    {
      "entropy": 5.307447385787964,
      "epoch": 1.4602770709640938,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048388927172009574,
      "loss": 5.3105,
      "mean_token_accuracy": 0.2046734422445297,
      "num_tokens": 4119299.0,
      "step": 5165
    },
    {
      "entropy": 5.103598976135254,
      "epoch": 1.4616906983319198,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004838510278617957,
      "loss": 4.9321,
      "mean_token_accuracy": 0.21666142642498015,
      "num_tokens": 4123276.0,
      "step": 5170
    },
    {
      "entropy": 5.040239048004151,
      "epoch": 1.4631043256997456,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004838127403549224,
      "loss": 5.0453,
      "mean_token_accuracy": 0.22774796336889266,
      "num_tokens": 4127286.0,
      "step": 5175
    },
    {
      "entropy": 5.097852087020874,
      "epoch": 1.4645179530675714,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048377440920747755,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2191850244998932,
      "num_tokens": 4131478.0,
      "step": 5180
    },
    {
      "entropy": 5.094895887374878,
      "epoch": 1.4659315804353972,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004837360344274723,
      "loss": 4.9438,
      "mean_token_accuracy": 0.23217518776655197,
      "num_tokens": 4135594.0,
      "step": 5185
    },
    {
      "entropy": 5.001806211471558,
      "epoch": 1.467345207803223,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004836976160229266,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2253419816493988,
      "num_tokens": 4139444.0,
      "step": 5190
    },
    {
      "entropy": 5.148755073547363,
      "epoch": 1.468758835171049,
      "grad_norm": 1.625,
      "learning_rate": 0.00048365915400186963,
      "loss": 5.1127,
      "mean_token_accuracy": 0.22827981263399125,
      "num_tokens": 4143735.0,
      "step": 5195
    },
    {
      "entropy": 5.177644348144531,
      "epoch": 1.4701724625388748,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00048362064837233985,
      "loss": 5.097,
      "mean_token_accuracy": 0.2169338971376419,
      "num_tokens": 4147737.0,
      "step": 5200
    },
    {
      "entropy": 5.147780275344848,
      "epoch": 1.4715860899067006,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048358209914238455,
      "loss": 5.0049,
      "mean_token_accuracy": 0.22870902866125106,
      "num_tokens": 4151707.0,
      "step": 5205
    },
    {
      "entropy": 5.197139024734497,
      "epoch": 1.4729997172745264,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004835435063200605,
      "loss": 5.0427,
      "mean_token_accuracy": 0.229584439098835,
      "num_tokens": 4155656.0,
      "step": 5210
    },
    {
      "entropy": 5.017662286758423,
      "epoch": 1.4744133446423522,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004835048699134331,
      "loss": 4.9919,
      "mean_token_accuracy": 0.22956017702817916,
      "num_tokens": 4159496.0,
      "step": 5215
    },
    {
      "entropy": 5.080389213562012,
      "epoch": 1.4758269720101782,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004834661899305773,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21937036663293838,
      "num_tokens": 4163496.0,
      "step": 5220
    },
    {
      "entropy": 5.178340244293213,
      "epoch": 1.477240599378004,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004834274663795769,
      "loss": 5.1888,
      "mean_token_accuracy": 0.21267613172531127,
      "num_tokens": 4167608.0,
      "step": 5225
    },
    {
      "entropy": 5.1129169940948485,
      "epoch": 1.4786542267458298,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048338869926852493,
      "loss": 4.9451,
      "mean_token_accuracy": 0.24459741711616517,
      "num_tokens": 4171482.0,
      "step": 5230
    },
    {
      "entropy": 5.082944393157959,
      "epoch": 1.4800678541136556,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004833498886055234,
      "loss": 5.1119,
      "mean_token_accuracy": 0.21556254625320434,
      "num_tokens": 4175461.0,
      "step": 5235
    },
    {
      "entropy": 5.147976064682007,
      "epoch": 1.4814814814814814,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048331103439868355,
      "loss": 5.1516,
      "mean_token_accuracy": 0.20210517793893815,
      "num_tokens": 4179378.0,
      "step": 5240
    },
    {
      "entropy": 5.121553230285644,
      "epoch": 1.4828951088493074,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048327213665612574,
      "loss": 5.0577,
      "mean_token_accuracy": 0.2167988508939743,
      "num_tokens": 4183699.0,
      "step": 5245
    },
    {
      "entropy": 5.244191884994507,
      "epoch": 1.4843087362171332,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004832331953859793,
      "loss": 5.1003,
      "mean_token_accuracy": 0.21456421613693238,
      "num_tokens": 4187637.0,
      "step": 5250
    },
    {
      "entropy": 5.061611843109131,
      "epoch": 1.485722363584959,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048319421059638256,
      "loss": 5.0526,
      "mean_token_accuracy": 0.22492918819189073,
      "num_tokens": 4191956.0,
      "step": 5255
    },
    {
      "entropy": 5.075877237319946,
      "epoch": 1.4871359909527848,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004831551822954834,
      "loss": 5.0655,
      "mean_token_accuracy": 0.22497609704732896,
      "num_tokens": 4196038.0,
      "step": 5260
    },
    {
      "entropy": 5.160483407974243,
      "epoch": 1.4885496183206106,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004831161104914382,
      "loss": 5.1375,
      "mean_token_accuracy": 0.21548450440168382,
      "num_tokens": 4200122.0,
      "step": 5265
    },
    {
      "entropy": 5.151312923431396,
      "epoch": 1.4899632456884366,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00048307699519241294,
      "loss": 4.9478,
      "mean_token_accuracy": 0.22840766310691835,
      "num_tokens": 4203923.0,
      "step": 5270
    },
    {
      "entropy": 5.160139322280884,
      "epoch": 1.4913768730562624,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004830378364065824,
      "loss": 5.0392,
      "mean_token_accuracy": 0.2347744107246399,
      "num_tokens": 4207877.0,
      "step": 5275
    },
    {
      "entropy": 5.126206541061402,
      "epoch": 1.4927905004240882,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048299863414213046,
      "loss": 4.9832,
      "mean_token_accuracy": 0.23168615400791168,
      "num_tokens": 4211625.0,
      "step": 5280
    },
    {
      "entropy": 5.187732601165772,
      "epoch": 1.494204127791914,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00048295938840725025,
      "loss": 5.2049,
      "mean_token_accuracy": 0.21745592206716538,
      "num_tokens": 4215837.0,
      "step": 5285
    },
    {
      "entropy": 5.167364692687988,
      "epoch": 1.4956177551597398,
      "grad_norm": 1.5,
      "learning_rate": 0.00048292009921014386,
      "loss": 5.0506,
      "mean_token_accuracy": 0.22552393227815629,
      "num_tokens": 4219899.0,
      "step": 5290
    },
    {
      "entropy": 5.061055755615234,
      "epoch": 1.4970313825275658,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004828807665590224,
      "loss": 4.8919,
      "mean_token_accuracy": 0.24149273335933685,
      "num_tokens": 4223609.0,
      "step": 5295
    },
    {
      "entropy": 5.0550302028656,
      "epoch": 1.4984450098953916,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048284139046210625,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22457516491413115,
      "num_tokens": 4227552.0,
      "step": 5300
    },
    {
      "entropy": 5.245283937454223,
      "epoch": 1.4998586372632174,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048280197092762475,
      "loss": 5.1568,
      "mean_token_accuracy": 0.22323370575904847,
      "num_tokens": 4231500.0,
      "step": 5305
    },
    {
      "entropy": 5.109475231170654,
      "epoch": 1.5012722646310432,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004827625079638164,
      "loss": 5.0687,
      "mean_token_accuracy": 0.23086457699537277,
      "num_tokens": 4235348.0,
      "step": 5310
    },
    {
      "entropy": 5.129686546325684,
      "epoch": 1.502685891998869,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048272300157892856,
      "loss": 5.1104,
      "mean_token_accuracy": 0.21839469224214553,
      "num_tokens": 4239255.0,
      "step": 5315
    },
    {
      "entropy": 5.122752714157104,
      "epoch": 1.504099519366695,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000482683451781218,
      "loss": 5.0529,
      "mean_token_accuracy": 0.22463565766811372,
      "num_tokens": 4243124.0,
      "step": 5320
    },
    {
      "entropy": 5.150032711029053,
      "epoch": 1.5055131467345206,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004826438585789503,
      "loss": 4.9957,
      "mean_token_accuracy": 0.22091101258993148,
      "num_tokens": 4247239.0,
      "step": 5325
    },
    {
      "entropy": 5.177624702453613,
      "epoch": 1.5069267741023467,
      "grad_norm": 1.5,
      "learning_rate": 0.0004826042219804,
      "loss": 5.1169,
      "mean_token_accuracy": 0.21934451013803483,
      "num_tokens": 4251376.0,
      "step": 5330
    },
    {
      "entropy": 5.159681224822998,
      "epoch": 1.5083404014701725,
      "grad_norm": 1.5,
      "learning_rate": 0.0004825645419938513,
      "loss": 5.0018,
      "mean_token_accuracy": 0.2272571638226509,
      "num_tokens": 4255559.0,
      "step": 5335
    },
    {
      "entropy": 5.06374888420105,
      "epoch": 1.5097540288379983,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004825248186275967,
      "loss": 5.0884,
      "mean_token_accuracy": 0.22525538355112076,
      "num_tokens": 4259778.0,
      "step": 5340
    },
    {
      "entropy": 5.068151330947876,
      "epoch": 1.5111676562058243,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004824850518899384,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22861998230218888,
      "num_tokens": 4263678.0,
      "step": 5345
    },
    {
      "entropy": 5.154883909225464,
      "epoch": 1.5125812835736498,
      "grad_norm": 1.5,
      "learning_rate": 0.0004824452417891874,
      "loss": 5.0892,
      "mean_token_accuracy": 0.22065166831016542,
      "num_tokens": 4267487.0,
      "step": 5350
    },
    {
      "entropy": 5.301064729690552,
      "epoch": 1.5139949109414759,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048240538833366355,
      "loss": 5.1276,
      "mean_token_accuracy": 0.2165652483701706,
      "num_tokens": 4271592.0,
      "step": 5355
    },
    {
      "entropy": 5.165159559249878,
      "epoch": 1.5154085383093017,
      "grad_norm": 1.640625,
      "learning_rate": 0.00048236549153169605,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21787649244070054,
      "num_tokens": 4275378.0,
      "step": 5360
    },
    {
      "entropy": 4.973621845245361,
      "epoch": 1.5168221656771275,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004823255513916232,
      "loss": 5.1194,
      "mean_token_accuracy": 0.2207758516073227,
      "num_tokens": 4279441.0,
      "step": 5365
    },
    {
      "entropy": 5.140539264678955,
      "epoch": 1.5182357930449535,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048228556792179216,
      "loss": 5.1049,
      "mean_token_accuracy": 0.22792670428752898,
      "num_tokens": 4283320.0,
      "step": 5370
    },
    {
      "entropy": 5.195977592468262,
      "epoch": 1.519649420412779,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004822455411305592,
      "loss": 5.0695,
      "mean_token_accuracy": 0.22452030330896378,
      "num_tokens": 4287380.0,
      "step": 5375
    },
    {
      "entropy": 5.060672903060913,
      "epoch": 1.521063047780605,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004822054710262898,
      "loss": 5.1336,
      "mean_token_accuracy": 0.21554662883281708,
      "num_tokens": 4291541.0,
      "step": 5380
    },
    {
      "entropy": 5.0448901653289795,
      "epoch": 1.5224766751484309,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004821653576173582,
      "loss": 5.0486,
      "mean_token_accuracy": 0.21943637877702712,
      "num_tokens": 4295649.0,
      "step": 5385
    },
    {
      "entropy": 5.058656072616577,
      "epoch": 1.5238903025162567,
      "grad_norm": 1.515625,
      "learning_rate": 0.00048212520091214794,
      "loss": 5.0549,
      "mean_token_accuracy": 0.23659731298685074,
      "num_tokens": 4299383.0,
      "step": 5390
    },
    {
      "entropy": 5.015478515625,
      "epoch": 1.5253039298840827,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004820850009190516,
      "loss": 4.9305,
      "mean_token_accuracy": 0.22999042719602586,
      "num_tokens": 4303403.0,
      "step": 5395
    },
    {
      "entropy": 5.131919240951538,
      "epoch": 1.5267175572519083,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048204475764647057,
      "loss": 5.0977,
      "mean_token_accuracy": 0.22427867650985717,
      "num_tokens": 4307458.0,
      "step": 5400
    },
    {
      "entropy": 5.038822174072266,
      "epoch": 1.5281311846197343,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00048200447110281546,
      "loss": 4.9892,
      "mean_token_accuracy": 0.2375664785504341,
      "num_tokens": 4311130.0,
      "step": 5405
    },
    {
      "entropy": 5.011002159118652,
      "epoch": 1.52954481198756,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048196414129650605,
      "loss": 5.0235,
      "mean_token_accuracy": 0.2322246477007866,
      "num_tokens": 4315271.0,
      "step": 5410
    },
    {
      "entropy": 5.031664657592773,
      "epoch": 1.530958439355386,
      "grad_norm": 1.5234375,
      "learning_rate": 0.000481923768235971,
      "loss": 5.0734,
      "mean_token_accuracy": 0.23094615191221238,
      "num_tokens": 4319044.0,
      "step": 5415
    },
    {
      "entropy": 5.147319269180298,
      "epoch": 1.532372066723212,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00048188335192964776,
      "loss": 5.0272,
      "mean_token_accuracy": 0.22528885155916215,
      "num_tokens": 4322980.0,
      "step": 5420
    },
    {
      "entropy": 5.10207953453064,
      "epoch": 1.5337856940910375,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004818428923859834,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2186853140592575,
      "num_tokens": 4326971.0,
      "step": 5425
    },
    {
      "entropy": 5.197276639938354,
      "epoch": 1.5351993214588635,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004818023896134337,
      "loss": 4.9753,
      "mean_token_accuracy": 0.236081300675869,
      "num_tokens": 4331124.0,
      "step": 5430
    },
    {
      "entropy": 5.155055952072144,
      "epoch": 1.5366129488266893,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00048176184362046335,
      "loss": 5.0601,
      "mean_token_accuracy": 0.22926076501607895,
      "num_tokens": 4334945.0,
      "step": 5435
    },
    {
      "entropy": 5.1824136734008786,
      "epoch": 1.538026576194515,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004817212544155462,
      "loss": 5.1098,
      "mean_token_accuracy": 0.21252659112215042,
      "num_tokens": 4338989.0,
      "step": 5440
    },
    {
      "entropy": 5.163645315170288,
      "epoch": 1.5394402035623411,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048168062200716524,
      "loss": 5.0888,
      "mean_token_accuracy": 0.228516985476017,
      "num_tokens": 4342748.0,
      "step": 5445
    },
    {
      "entropy": 5.177605676651001,
      "epoch": 1.5408538309301667,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004816399464038124,
      "loss": 5.091,
      "mean_token_accuracy": 0.22972729355096816,
      "num_tokens": 4346598.0,
      "step": 5450
    },
    {
      "entropy": 5.186751127243042,
      "epoch": 1.5422674582979927,
      "grad_norm": 1.546875,
      "learning_rate": 0.00048159922761398846,
      "loss": 5.0609,
      "mean_token_accuracy": 0.22715484350919724,
      "num_tokens": 4350477.0,
      "step": 5455
    },
    {
      "entropy": 5.130059099197387,
      "epoch": 1.5436810856658185,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004815584656462036,
      "loss": 4.9699,
      "mean_token_accuracy": 0.22770097404718398,
      "num_tokens": 4354556.0,
      "step": 5460
    },
    {
      "entropy": 5.1678211212158205,
      "epoch": 1.5450947130336443,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004815176605089767,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2265477493405342,
      "num_tokens": 4358575.0,
      "step": 5465
    },
    {
      "entropy": 5.268654823303223,
      "epoch": 1.5465083404014703,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004814768122108359,
      "loss": 5.1099,
      "mean_token_accuracy": 0.21699922233819963,
      "num_tokens": 4362555.0,
      "step": 5470
    },
    {
      "entropy": 5.12242546081543,
      "epoch": 1.547921967769296,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00048143592076031803,
      "loss": 4.9975,
      "mean_token_accuracy": 0.22739111334085466,
      "num_tokens": 4366714.0,
      "step": 5475
    },
    {
      "entropy": 5.236171436309815,
      "epoch": 1.549335595137122,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004813949861659694,
      "loss": 5.0316,
      "mean_token_accuracy": 0.2231994777917862,
      "num_tokens": 4370591.0,
      "step": 5480
    },
    {
      "entropy": 5.1986589431762695,
      "epoch": 1.5507492225049477,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00048135400843634497,
      "loss": 5.0276,
      "mean_token_accuracy": 0.22530472576618193,
      "num_tokens": 4374436.0,
      "step": 5485
    },
    {
      "entropy": 5.201425313949585,
      "epoch": 1.5521628498727735,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00048131298758000884,
      "loss": 5.0821,
      "mean_token_accuracy": 0.22044385075569153,
      "num_tokens": 4378655.0,
      "step": 5490
    },
    {
      "entropy": 5.229818534851074,
      "epoch": 1.5535764772405993,
      "grad_norm": 1.5,
      "learning_rate": 0.0004812719236055341,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2259032741189003,
      "num_tokens": 4382457.0,
      "step": 5495
    },
    {
      "entropy": 5.282522583007813,
      "epoch": 1.5549901046084251,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004812308165215028,
      "loss": 5.1029,
      "mean_token_accuracy": 0.22373139411211013,
      "num_tokens": 4386300.0,
      "step": 5500
    },
    {
      "entropy": 5.164923810958863,
      "epoch": 1.5564037319762511,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048118966633650613,
      "loss": 5.0473,
      "mean_token_accuracy": 0.2359399288892746,
      "num_tokens": 4390285.0,
      "step": 5505
    },
    {
      "entropy": 5.258483839035034,
      "epoch": 1.557817359344077,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004811484730591444,
      "loss": 5.061,
      "mean_token_accuracy": 0.2201920747756958,
      "num_tokens": 4394268.0,
      "step": 5510
    },
    {
      "entropy": 5.159657573699951,
      "epoch": 1.5592309867119027,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004811072366980264,
      "loss": 4.939,
      "mean_token_accuracy": 0.23556929379701613,
      "num_tokens": 4398310.0,
      "step": 5515
    },
    {
      "entropy": 5.119165277481079,
      "epoch": 1.5606446140797285,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004810659572617705,
      "loss": 4.9758,
      "mean_token_accuracy": 0.23152803033590316,
      "num_tokens": 4402306.0,
      "step": 5520
    },
    {
      "entropy": 5.205929231643677,
      "epoch": 1.5620582414475543,
      "grad_norm": 1.484375,
      "learning_rate": 0.00048102463475900386,
      "loss": 5.0037,
      "mean_token_accuracy": 0.22581120133399962,
      "num_tokens": 4406263.0,
      "step": 5525
    },
    {
      "entropy": 5.2671445369720455,
      "epoch": 1.5634718688153804,
      "grad_norm": 1.421875,
      "learning_rate": 0.00048098326919836253,
      "loss": 5.1333,
      "mean_token_accuracy": 0.2211618274450302,
      "num_tokens": 4410383.0,
      "step": 5530
    },
    {
      "entropy": 5.179918956756592,
      "epoch": 1.5648854961832062,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004809418605884917,
      "loss": 4.9703,
      "mean_token_accuracy": 0.2390846401453018,
      "num_tokens": 4414244.0,
      "step": 5535
    },
    {
      "entropy": 4.9973101139068605,
      "epoch": 1.566299123551032,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004809004089380456,
      "loss": 4.9925,
      "mean_token_accuracy": 0.23173850625753403,
      "num_tokens": 4418315.0,
      "step": 5540
    },
    {
      "entropy": 5.109961271286011,
      "epoch": 1.5677127509188578,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004808589142556872,
      "loss": 5.116,
      "mean_token_accuracy": 0.21659986078739166,
      "num_tokens": 4422333.0,
      "step": 5545
    },
    {
      "entropy": 5.2055091381073,
      "epoch": 1.5691263782866836,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004808173765500887,
      "loss": 5.1264,
      "mean_token_accuracy": 0.21076353639364243,
      "num_tokens": 4426255.0,
      "step": 5550
    },
    {
      "entropy": 5.129403591156006,
      "epoch": 1.5705400056545096,
      "grad_norm": 1.578125,
      "learning_rate": 0.00048077579582993127,
      "loss": 4.9955,
      "mean_token_accuracy": 0.23483571112155915,
      "num_tokens": 4430398.0,
      "step": 5555
    },
    {
      "entropy": 5.247654962539673,
      "epoch": 1.5719536330223352,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00048073417210390504,
      "loss": 5.2371,
      "mean_token_accuracy": 0.21894858926534652,
      "num_tokens": 4434849.0,
      "step": 5560
    },
    {
      "entropy": 5.121655035018921,
      "epoch": 1.5733672603901612,
      "grad_norm": 1.578125,
      "learning_rate": 0.000480692505380709,
      "loss": 4.9922,
      "mean_token_accuracy": 0.2310257524251938,
      "num_tokens": 4438735.0,
      "step": 5565
    },
    {
      "entropy": 5.180913496017456,
      "epoch": 1.574780887757987,
      "grad_norm": 1.5,
      "learning_rate": 0.00048065079566905126,
      "loss": 5.0331,
      "mean_token_accuracy": 0.23977220803499222,
      "num_tokens": 4442453.0,
      "step": 5570
    },
    {
      "entropy": 5.227556037902832,
      "epoch": 1.5761945151258128,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000480609042977649,
      "loss": 5.0035,
      "mean_token_accuracy": 0.23108505308628083,
      "num_tokens": 4446552.0,
      "step": 5575
    },
    {
      "entropy": 5.228785657882691,
      "epoch": 1.5776081424936388,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048056724731522824,
      "loss": 5.0381,
      "mean_token_accuracy": 0.23695874363183975,
      "num_tokens": 4450576.0,
      "step": 5580
    },
    {
      "entropy": 5.101385498046875,
      "epoch": 1.5790217698614644,
      "grad_norm": 1.453125,
      "learning_rate": 0.00048052540869052384,
      "loss": 4.991,
      "mean_token_accuracy": 0.23343684524297714,
      "num_tokens": 4454667.0,
      "step": 5585
    },
    {
      "entropy": 5.03810224533081,
      "epoch": 1.5804353972292904,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00048048352711228007,
      "loss": 4.9634,
      "mean_token_accuracy": 0.22962892800569534,
      "num_tokens": 4458337.0,
      "step": 5590
    },
    {
      "entropy": 5.157836294174194,
      "epoch": 1.5818490245971162,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004804416025892498,
      "loss": 5.0861,
      "mean_token_accuracy": 0.22233165800571442,
      "num_tokens": 4462298.0,
      "step": 5595
    },
    {
      "entropy": 5.1862602710723875,
      "epoch": 1.583262651964942,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004803996351301949,
      "loss": 5.1032,
      "mean_token_accuracy": 0.21805992275476455,
      "num_tokens": 4466586.0,
      "step": 5600
    },
    {
      "entropy": 5.101711654663086,
      "epoch": 1.584676279332768,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00048035762474388656,
      "loss": 4.9622,
      "mean_token_accuracy": 0.23858007788658142,
      "num_tokens": 4470352.0,
      "step": 5605
    },
    {
      "entropy": 5.114037322998047,
      "epoch": 1.5860899067005936,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004803155714391044,
      "loss": 5.1448,
      "mean_token_accuracy": 0.2237158253788948,
      "num_tokens": 4474741.0,
      "step": 5610
    },
    {
      "entropy": 5.1974996566772464,
      "epoch": 1.5875035340684196,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004802734752246375,
      "loss": 5.0477,
      "mean_token_accuracy": 0.22567686736583709,
      "num_tokens": 4478613.0,
      "step": 5615
    },
    {
      "entropy": 5.248621654510498,
      "epoch": 1.5889171614362454,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004802313361092837,
      "loss": 5.1078,
      "mean_token_accuracy": 0.21730149686336517,
      "num_tokens": 4482375.0,
      "step": 5620
    },
    {
      "entropy": 5.099520206451416,
      "epoch": 1.5903307888040712,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004801891541018497,
      "loss": 4.9765,
      "mean_token_accuracy": 0.22748092114925383,
      "num_tokens": 4486672.0,
      "step": 5625
    },
    {
      "entropy": 5.198351955413818,
      "epoch": 1.5917444161718972,
      "grad_norm": 1.609375,
      "learning_rate": 0.00048014692921115136,
      "loss": 5.2293,
      "mean_token_accuracy": 0.20402153730392455,
      "num_tokens": 4490606.0,
      "step": 5630
    },
    {
      "entropy": 5.227230262756348,
      "epoch": 1.5931580435397228,
      "grad_norm": 1.5625,
      "learning_rate": 0.00048010466144601344,
      "loss": 5.1776,
      "mean_token_accuracy": 0.221833735704422,
      "num_tokens": 4494690.0,
      "step": 5635
    },
    {
      "entropy": 5.248382806777954,
      "epoch": 1.5945716709075488,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004800623508152696,
      "loss": 4.9938,
      "mean_token_accuracy": 0.22964889258146287,
      "num_tokens": 4498576.0,
      "step": 5640
    },
    {
      "entropy": 5.073389959335327,
      "epoch": 1.5959852982753746,
      "grad_norm": 1.46875,
      "learning_rate": 0.00048001999732776246,
      "loss": 4.9063,
      "mean_token_accuracy": 0.2410113289952278,
      "num_tokens": 4502489.0,
      "step": 5645
    },
    {
      "entropy": 5.089180374145508,
      "epoch": 1.5973989256432004,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004799776009923437,
      "loss": 5.0639,
      "mean_token_accuracy": 0.2287627413868904,
      "num_tokens": 4506270.0,
      "step": 5650
    },
    {
      "entropy": 5.0953692436218265,
      "epoch": 1.5988125530110264,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047993516181787395,
      "loss": 5.048,
      "mean_token_accuracy": 0.2261709451675415,
      "num_tokens": 4510306.0,
      "step": 5655
    },
    {
      "entropy": 5.130039119720459,
      "epoch": 1.600226180378852,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004798926798132226,
      "loss": 4.9333,
      "mean_token_accuracy": 0.23103215545415878,
      "num_tokens": 4514353.0,
      "step": 5660
    },
    {
      "entropy": 5.100818872451782,
      "epoch": 1.601639807746678,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004798501549872681,
      "loss": 5.1269,
      "mean_token_accuracy": 0.21848572939634323,
      "num_tokens": 4518691.0,
      "step": 5665
    },
    {
      "entropy": 5.159361743927002,
      "epoch": 1.6030534351145038,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004798075873488981,
      "loss": 5.056,
      "mean_token_accuracy": 0.2305290848016739,
      "num_tokens": 4522625.0,
      "step": 5670
    },
    {
      "entropy": 5.169857788085937,
      "epoch": 1.6044670624823296,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00047976497690700883,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2349533826112747,
      "num_tokens": 4526595.0,
      "step": 5675
    },
    {
      "entropy": 5.143611669540405,
      "epoch": 1.6058806898501556,
      "grad_norm": 1.453125,
      "learning_rate": 0.00047972232367050557,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2271997883915901,
      "num_tokens": 4530928.0,
      "step": 5680
    },
    {
      "entropy": 5.131349754333496,
      "epoch": 1.6072943172179812,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004796796276483026,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2226678639650345,
      "num_tokens": 4534812.0,
      "step": 5685
    },
    {
      "entropy": 5.060449457168579,
      "epoch": 1.6087079445858072,
      "grad_norm": 1.640625,
      "learning_rate": 0.00047963688884932316,
      "loss": 5.0243,
      "mean_token_accuracy": 0.2311640426516533,
      "num_tokens": 4538869.0,
      "step": 5690
    },
    {
      "entropy": 5.107732057571411,
      "epoch": 1.610121571953633,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047959410728249947,
      "loss": 5.0811,
      "mean_token_accuracy": 0.21918047219514847,
      "num_tokens": 4543069.0,
      "step": 5695
    },
    {
      "entropy": 5.1152708530426025,
      "epoch": 1.6115351993214588,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004795512829567725,
      "loss": 5.1215,
      "mean_token_accuracy": 0.2172875478863716,
      "num_tokens": 4546990.0,
      "step": 5700
    },
    {
      "entropy": 5.076419734954834,
      "epoch": 1.6129488266892849,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004795084158810922,
      "loss": 4.9719,
      "mean_token_accuracy": 0.23624841421842574,
      "num_tokens": 4550979.0,
      "step": 5705
    },
    {
      "entropy": 5.055774784088134,
      "epoch": 1.6143624540571104,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004794655060644177,
      "loss": 4.8829,
      "mean_token_accuracy": 0.23749051690101625,
      "num_tokens": 4555002.0,
      "step": 5710
    },
    {
      "entropy": 4.931989669799805,
      "epoch": 1.6157760814249365,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004794225535157168,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23389404863119126,
      "num_tokens": 4558813.0,
      "step": 5715
    },
    {
      "entropy": 5.135082340240478,
      "epoch": 1.6171897087927622,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047937955824396634,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21693041026592255,
      "num_tokens": 4562941.0,
      "step": 5720
    },
    {
      "entropy": 5.138332176208496,
      "epoch": 1.618603336160588,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000479336520258152,
      "loss": 5.0561,
      "mean_token_accuracy": 0.2299040898680687,
      "num_tokens": 4566820.0,
      "step": 5725
    },
    {
      "entropy": 5.018996000289917,
      "epoch": 1.620016963528414,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047929343956726854,
      "loss": 4.9871,
      "mean_token_accuracy": 0.22751606106758118,
      "num_tokens": 4570577.0,
      "step": 5730
    },
    {
      "entropy": 5.09666862487793,
      "epoch": 1.6214305908962396,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047925031618031963,
      "loss": 5.1334,
      "mean_token_accuracy": 0.21439673155546188,
      "num_tokens": 4574721.0,
      "step": 5735
    },
    {
      "entropy": 5.234111690521241,
      "epoch": 1.6228442182640657,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004792071501063175,
      "loss": 5.0539,
      "mean_token_accuracy": 0.22354237884283065,
      "num_tokens": 4578499.0,
      "step": 5740
    },
    {
      "entropy": 5.135746431350708,
      "epoch": 1.6242578456318915,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047916394135428386,
      "loss": 5.0183,
      "mean_token_accuracy": 0.22844223231077193,
      "num_tokens": 4582377.0,
      "step": 5745
    },
    {
      "entropy": 5.137772274017334,
      "epoch": 1.6256714729997173,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047912068993324897,
      "loss": 5.0155,
      "mean_token_accuracy": 0.2331460729241371,
      "num_tokens": 4586396.0,
      "step": 5750
    },
    {
      "entropy": 5.125563097000122,
      "epoch": 1.6270851003675433,
      "grad_norm": 1.5,
      "learning_rate": 0.00047907739585225214,
      "loss": 5.1097,
      "mean_token_accuracy": 0.2287369191646576,
      "num_tokens": 4590605.0,
      "step": 5755
    },
    {
      "entropy": 5.20036039352417,
      "epoch": 1.6284987277353689,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004790340591203416,
      "loss": 5.0205,
      "mean_token_accuracy": 0.22584520727396012,
      "num_tokens": 4594759.0,
      "step": 5760
    },
    {
      "entropy": 5.0811755657196045,
      "epoch": 1.6299123551031949,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004789906797465744,
      "loss": 4.9104,
      "mean_token_accuracy": 0.23817975372076033,
      "num_tokens": 4598738.0,
      "step": 5765
    },
    {
      "entropy": 5.065733289718628,
      "epoch": 1.6313259824710207,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004789472577400165,
      "loss": 5.192,
      "mean_token_accuracy": 0.21387971192598343,
      "num_tokens": 4602843.0,
      "step": 5770
    },
    {
      "entropy": 5.017878198623658,
      "epoch": 1.6327396098388465,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047890379310974295,
      "loss": 4.8955,
      "mean_token_accuracy": 0.23911944031715393,
      "num_tokens": 4606826.0,
      "step": 5775
    },
    {
      "entropy": 5.1054637908935545,
      "epoch": 1.6341532372066723,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004788602858648375,
      "loss": 5.0091,
      "mean_token_accuracy": 0.2160083070397377,
      "num_tokens": 4610476.0,
      "step": 5780
    },
    {
      "entropy": 5.069868755340576,
      "epoch": 1.635566864574498,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047881673601439306,
      "loss": 5.0108,
      "mean_token_accuracy": 0.22067209482192993,
      "num_tokens": 4614287.0,
      "step": 5785
    },
    {
      "entropy": 5.162918472290039,
      "epoch": 1.636980491942324,
      "grad_norm": 1.609375,
      "learning_rate": 0.000478773143567511,
      "loss": 5.0847,
      "mean_token_accuracy": 0.21399689465761185,
      "num_tokens": 4618279.0,
      "step": 5790
    },
    {
      "entropy": 5.09878978729248,
      "epoch": 1.6383941193101499,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004787295085333021,
      "loss": 4.9943,
      "mean_token_accuracy": 0.23267503827810287,
      "num_tokens": 4622491.0,
      "step": 5795
    },
    {
      "entropy": 5.049604892730713,
      "epoch": 1.6398077466779757,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004786858309208858,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22849075347185135,
      "num_tokens": 4626314.0,
      "step": 5800
    },
    {
      "entropy": 5.059546041488647,
      "epoch": 1.6412213740458015,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047864211073939036,
      "loss": 5.0303,
      "mean_token_accuracy": 0.22859601825475692,
      "num_tokens": 4630593.0,
      "step": 5805
    },
    {
      "entropy": 5.072523355484009,
      "epoch": 1.6426350014136273,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047859834799795305,
      "loss": 5.0404,
      "mean_token_accuracy": 0.23431884199380876,
      "num_tokens": 4634562.0,
      "step": 5810
    },
    {
      "entropy": 5.163015365600586,
      "epoch": 1.6440486287814533,
      "grad_norm": 1.375,
      "learning_rate": 0.00047855454270571995,
      "loss": 5.0212,
      "mean_token_accuracy": 0.22630929350852966,
      "num_tokens": 4638689.0,
      "step": 5815
    },
    {
      "entropy": 5.029438161849976,
      "epoch": 1.645462256149279,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004785106948718463,
      "loss": 4.9734,
      "mean_token_accuracy": 0.23704889714717864,
      "num_tokens": 4642544.0,
      "step": 5820
    },
    {
      "entropy": 4.923280668258667,
      "epoch": 1.646875883517105,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004784668045054958,
      "loss": 4.9266,
      "mean_token_accuracy": 0.24414554685354234,
      "num_tokens": 4646247.0,
      "step": 5825
    },
    {
      "entropy": 5.037514925003052,
      "epoch": 1.6482895108849307,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047842287161584135,
      "loss": 4.9689,
      "mean_token_accuracy": 0.22659680843353272,
      "num_tokens": 4650503.0,
      "step": 5830
    },
    {
      "entropy": 5.031885147094727,
      "epoch": 1.6497031382527565,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004783788962120647,
      "loss": 5.0796,
      "mean_token_accuracy": 0.22651676535606385,
      "num_tokens": 4654541.0,
      "step": 5835
    },
    {
      "entropy": 5.097152328491211,
      "epoch": 1.6511167656205825,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047833487830335644,
      "loss": 5.0715,
      "mean_token_accuracy": 0.22631967216730117,
      "num_tokens": 4658619.0,
      "step": 5840
    },
    {
      "entropy": 4.903795480728149,
      "epoch": 1.652530392988408,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000478290817898916,
      "loss": 4.926,
      "mean_token_accuracy": 0.24210842102766036,
      "num_tokens": 4662417.0,
      "step": 5845
    },
    {
      "entropy": 5.013822412490844,
      "epoch": 1.6539440203562341,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047824671500795165,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20652547776699065,
      "num_tokens": 4666779.0,
      "step": 5850
    },
    {
      "entropy": 5.0866453647613525,
      "epoch": 1.65535764772406,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047820256963968075,
      "loss": 5.07,
      "mean_token_accuracy": 0.2242145597934723,
      "num_tokens": 4671010.0,
      "step": 5855
    },
    {
      "entropy": 5.19718656539917,
      "epoch": 1.6567712750918857,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047815838180332936,
      "loss": 5.1297,
      "mean_token_accuracy": 0.21102940291166306,
      "num_tokens": 4675225.0,
      "step": 5860
    },
    {
      "entropy": 5.03183217048645,
      "epoch": 1.6581849024597117,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00047811415150813245,
      "loss": 4.9964,
      "mean_token_accuracy": 0.22597301304340361,
      "num_tokens": 4679488.0,
      "step": 5865
    },
    {
      "entropy": 5.079188442230224,
      "epoch": 1.6595985298275373,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047806987876333393,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22046170383691788,
      "num_tokens": 4683271.0,
      "step": 5870
    },
    {
      "entropy": 4.982157182693482,
      "epoch": 1.6610121571953633,
      "grad_norm": 1.5,
      "learning_rate": 0.00047802556357818643,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2270784318447113,
      "num_tokens": 4687240.0,
      "step": 5875
    },
    {
      "entropy": 4.998562479019165,
      "epoch": 1.6624257845631891,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047798120596195164,
      "loss": 4.9615,
      "mean_token_accuracy": 0.22471271455287933,
      "num_tokens": 4691322.0,
      "step": 5880
    },
    {
      "entropy": 5.034221553802491,
      "epoch": 1.663839411931015,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004779368059239,
      "loss": 5.0793,
      "mean_token_accuracy": 0.22571154534816742,
      "num_tokens": 4695369.0,
      "step": 5885
    },
    {
      "entropy": 5.1495583057403564,
      "epoch": 1.665253039298841,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004778923634733109,
      "loss": 5.0604,
      "mean_token_accuracy": 0.20926564633846284,
      "num_tokens": 4699392.0,
      "step": 5890
    },
    {
      "entropy": 5.114596891403198,
      "epoch": 1.6666666666666665,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004778478786194724,
      "loss": 5.0526,
      "mean_token_accuracy": 0.22140302658081054,
      "num_tokens": 4703112.0,
      "step": 5895
    },
    {
      "entropy": 5.0327517032623295,
      "epoch": 1.6680802940344925,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00047780335137168176,
      "loss": 4.9459,
      "mean_token_accuracy": 0.24414481222629547,
      "num_tokens": 4707201.0,
      "step": 5900
    },
    {
      "entropy": 5.055890607833862,
      "epoch": 1.6694939214023183,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047775878173924463,
      "loss": 5.0683,
      "mean_token_accuracy": 0.22704073488712312,
      "num_tokens": 4711257.0,
      "step": 5905
    },
    {
      "entropy": 5.0120703220367435,
      "epoch": 1.6709075487701441,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047771416973147607,
      "loss": 5.0204,
      "mean_token_accuracy": 0.2356857255101204,
      "num_tokens": 4715438.0,
      "step": 5910
    },
    {
      "entropy": 5.10405421257019,
      "epoch": 1.6723211761379702,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047766951535769956,
      "loss": 5.0668,
      "mean_token_accuracy": 0.22063805758953095,
      "num_tokens": 4719524.0,
      "step": 5915
    },
    {
      "entropy": 5.070660924911499,
      "epoch": 1.6737348035057957,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047762481862724754,
      "loss": 5.0187,
      "mean_token_accuracy": 0.22822391241788864,
      "num_tokens": 4723350.0,
      "step": 5920
    },
    {
      "entropy": 5.1049707412719725,
      "epoch": 1.6751484308736218,
      "grad_norm": 1.6875,
      "learning_rate": 0.00047758007954946144,
      "loss": 5.1049,
      "mean_token_accuracy": 0.22477101981639863,
      "num_tokens": 4727225.0,
      "step": 5925
    },
    {
      "entropy": 4.971397399902344,
      "epoch": 1.6765620582414476,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047753529813369147,
      "loss": 4.915,
      "mean_token_accuracy": 0.24205376207828522,
      "num_tokens": 4731252.0,
      "step": 5930
    },
    {
      "entropy": 5.034673976898193,
      "epoch": 1.6779756856092733,
      "grad_norm": 1.5,
      "learning_rate": 0.00047749047438929663,
      "loss": 4.9999,
      "mean_token_accuracy": 0.22744559794664382,
      "num_tokens": 4735428.0,
      "step": 5935
    },
    {
      "entropy": 5.029658508300781,
      "epoch": 1.6793893129770994,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047744560832564475,
      "loss": 5.0099,
      "mean_token_accuracy": 0.2266997367143631,
      "num_tokens": 4739480.0,
      "step": 5940
    },
    {
      "entropy": 5.191289043426513,
      "epoch": 1.680802940344925,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004774006999521127,
      "loss": 5.1387,
      "mean_token_accuracy": 0.2114172250032425,
      "num_tokens": 4744145.0,
      "step": 5945
    },
    {
      "entropy": 5.150563430786133,
      "epoch": 1.682216567712751,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004773557492780859,
      "loss": 4.9978,
      "mean_token_accuracy": 0.23651132434606553,
      "num_tokens": 4748088.0,
      "step": 5950
    },
    {
      "entropy": 5.1101823329925535,
      "epoch": 1.6836301950805768,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004773107563129588,
      "loss": 5.0068,
      "mean_token_accuracy": 0.23149034678936004,
      "num_tokens": 4751874.0,
      "step": 5955
    },
    {
      "entropy": 5.115221929550171,
      "epoch": 1.6850438224484026,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004772657210661348,
      "loss": 5.1122,
      "mean_token_accuracy": 0.21590612083673477,
      "num_tokens": 4755934.0,
      "step": 5960
    },
    {
      "entropy": 5.1689688205719,
      "epoch": 1.6864574498162286,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004772206435470258,
      "loss": 5.0613,
      "mean_token_accuracy": 0.22585527896881102,
      "num_tokens": 4759554.0,
      "step": 5965
    },
    {
      "entropy": 5.089975786209107,
      "epoch": 1.6878710771840542,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004771755237650528,
      "loss": 4.9767,
      "mean_token_accuracy": 0.23322198539972305,
      "num_tokens": 4763508.0,
      "step": 5970
    },
    {
      "entropy": 5.042976236343383,
      "epoch": 1.6892847045518802,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004771303617296455,
      "loss": 4.907,
      "mean_token_accuracy": 0.23039111942052842,
      "num_tokens": 4767408.0,
      "step": 5975
    },
    {
      "entropy": 5.058781290054322,
      "epoch": 1.690698331919706,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004770851574502426,
      "loss": 5.0466,
      "mean_token_accuracy": 0.230085851252079,
      "num_tokens": 4771347.0,
      "step": 5980
    },
    {
      "entropy": 5.156385660171509,
      "epoch": 1.6921119592875318,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00047703991093629147,
      "loss": 5.0408,
      "mean_token_accuracy": 0.22459974139928818,
      "num_tokens": 4775541.0,
      "step": 5985
    },
    {
      "entropy": 5.087676858901977,
      "epoch": 1.6935255866553578,
      "grad_norm": 1.5,
      "learning_rate": 0.0004769946221972483,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23913984447717668,
      "num_tokens": 4779485.0,
      "step": 5990
    },
    {
      "entropy": 5.163230514526367,
      "epoch": 1.6949392140231834,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004769492912425782,
      "loss": 5.059,
      "mean_token_accuracy": 0.21983974128961564,
      "num_tokens": 4783619.0,
      "step": 5995
    },
    {
      "entropy": 5.1115906715393065,
      "epoch": 1.6963528413910094,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00047690391808175497,
      "loss": 5.0116,
      "mean_token_accuracy": 0.22564171999692917,
      "num_tokens": 4787648.0,
      "step": 6000
    },
    {
      "epoch": 1.6963528413910094,
      "eval_entropy": 4.925529229564924,
      "eval_loss": 5.301758766174316,
      "eval_mean_token_accuracy": 0.2223591335714497,
      "eval_num_tokens": 4787648.0,
      "eval_runtime": 6.5862,
      "eval_samples_per_second": 1713.287,
      "eval_steps_per_second": 214.237,
      "step": 6000
    },
    {
      "entropy": 5.027353048324585,
      "epoch": 1.6977664687588352,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004768585027242615,
      "loss": 5.0066,
      "mean_token_accuracy": 0.238006529211998,
      "num_tokens": 4791302.0,
      "step": 6005
    },
    {
      "entropy": 5.0390783786773685,
      "epoch": 1.699180096126661,
      "grad_norm": 1.390625,
      "learning_rate": 0.00047681304517958916,
      "loss": 4.9724,
      "mean_token_accuracy": 0.23309302628040313,
      "num_tokens": 4795481.0,
      "step": 6010
    },
    {
      "entropy": 5.149922132492065,
      "epoch": 1.700593723494487,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047676754545723836,
      "loss": 5.1323,
      "mean_token_accuracy": 0.2138197675347328,
      "num_tokens": 4799380.0,
      "step": 6015
    },
    {
      "entropy": 5.036032962799072,
      "epoch": 1.7020073508623126,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004767220035667183,
      "loss": 4.9276,
      "mean_token_accuracy": 0.2285904586315155,
      "num_tokens": 4802946.0,
      "step": 6020
    },
    {
      "entropy": 5.146701002120972,
      "epoch": 1.7034209782301386,
      "grad_norm": 1.5,
      "learning_rate": 0.0004766764195175469,
      "loss": 5.0835,
      "mean_token_accuracy": 0.22148032039403914,
      "num_tokens": 4806925.0,
      "step": 6025
    },
    {
      "entropy": 4.973399353027344,
      "epoch": 1.7048346055979644,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047663079331925086,
      "loss": 4.9291,
      "mean_token_accuracy": 0.23884583562612532,
      "num_tokens": 4810879.0,
      "step": 6030
    },
    {
      "entropy": 4.8970386505126955,
      "epoch": 1.7062482329657902,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000476585124981366,
      "loss": 4.8758,
      "mean_token_accuracy": 0.23640986084938048,
      "num_tokens": 4814863.0,
      "step": 6035
    },
    {
      "entropy": 4.980345344543457,
      "epoch": 1.7076618603336162,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004765394145134365,
      "loss": 4.9729,
      "mean_token_accuracy": 0.2318740889430046,
      "num_tokens": 4818548.0,
      "step": 6040
    },
    {
      "entropy": 5.045832681655884,
      "epoch": 1.7090754877014418,
      "grad_norm": 1.890625,
      "learning_rate": 0.0004764936619250156,
      "loss": 5.0184,
      "mean_token_accuracy": 0.2202947571873665,
      "num_tokens": 4822693.0,
      "step": 6045
    },
    {
      "entropy": 5.133233499526978,
      "epoch": 1.7104891150692678,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00047644786722566537,
      "loss": 5.0311,
      "mean_token_accuracy": 0.22884614020586014,
      "num_tokens": 4827110.0,
      "step": 6050
    },
    {
      "entropy": 5.078573560714721,
      "epoch": 1.7119027424370936,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047640203042495667,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2241154968738556,
      "num_tokens": 4831327.0,
      "step": 6055
    },
    {
      "entropy": 5.047287511825561,
      "epoch": 1.7133163698049194,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00047635615153246907,
      "loss": 4.896,
      "mean_token_accuracy": 0.2353956937789917,
      "num_tokens": 4835767.0,
      "step": 6060
    },
    {
      "entropy": 5.15030460357666,
      "epoch": 1.7147299971727452,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004763102305577908,
      "loss": 5.002,
      "mean_token_accuracy": 0.2306875616312027,
      "num_tokens": 4840163.0,
      "step": 6065
    },
    {
      "entropy": 5.187795877456665,
      "epoch": 1.716143624540571,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004762642675105194,
      "loss": 5.0302,
      "mean_token_accuracy": 0.21816271245479585,
      "num_tokens": 4843874.0,
      "step": 6070
    },
    {
      "entropy": 5.1544421195983885,
      "epoch": 1.717557251908397,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004762182624002606,
      "loss": 4.9901,
      "mean_token_accuracy": 0.23147399723529816,
      "num_tokens": 4848042.0,
      "step": 6075
    },
    {
      "entropy": 5.0734031200408936,
      "epoch": 1.7189708792762228,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047617221523662916,
      "loss": 4.9855,
      "mean_token_accuracy": 0.2388760268688202,
      "num_tokens": 4851998.0,
      "step": 6080
    },
    {
      "entropy": 5.134505844116211,
      "epoch": 1.7203845066440486,
      "grad_norm": 1.625,
      "learning_rate": 0.00047612612602924883,
      "loss": 4.9689,
      "mean_token_accuracy": 0.23264792412519456,
      "num_tokens": 4856141.0,
      "step": 6085
    },
    {
      "entropy": 5.058994197845459,
      "epoch": 1.7217981340118744,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004760799947877518,
      "loss": 5.0566,
      "mean_token_accuracy": 0.22185839265584945,
      "num_tokens": 4860069.0,
      "step": 6090
    },
    {
      "entropy": 5.049883651733398,
      "epoch": 1.7232117613797002,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004760338215217795,
      "loss": 5.028,
      "mean_token_accuracy": 0.2247716709971428,
      "num_tokens": 4864042.0,
      "step": 6095
    },
    {
      "entropy": 5.084624576568603,
      "epoch": 1.7246253887475262,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004759876062409815,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2460511550307274,
      "num_tokens": 4868086.0,
      "step": 6100
    },
    {
      "entropy": 5.188075304031372,
      "epoch": 1.726039016115352,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004759413489550167,
      "loss": 5.1429,
      "mean_token_accuracy": 0.21462565213441848,
      "num_tokens": 4872034.0,
      "step": 6105
    },
    {
      "entropy": 5.155353212356568,
      "epoch": 1.7274526434831778,
      "grad_norm": 1.4375,
      "learning_rate": 0.00047589504967355254,
      "loss": 5.0328,
      "mean_token_accuracy": 0.22590787410736085,
      "num_tokens": 4876293.0,
      "step": 6110
    },
    {
      "entropy": 5.115995740890503,
      "epoch": 1.7288662708510036,
      "grad_norm": 1.5,
      "learning_rate": 0.0004758487084062653,
      "loss": 4.9557,
      "mean_token_accuracy": 0.22829659730196,
      "num_tokens": 4880573.0,
      "step": 6115
    },
    {
      "entropy": 5.138199949264527,
      "epoch": 1.7302798982188294,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047580232516284,
      "loss": 5.0585,
      "mean_token_accuracy": 0.22750245034694672,
      "num_tokens": 4884598.0,
      "step": 6120
    },
    {
      "entropy": 5.115506887435913,
      "epoch": 1.7316935255866555,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047575589995297045,
      "loss": 5.0398,
      "mean_token_accuracy": 0.23209791183471679,
      "num_tokens": 4888343.0,
      "step": 6125
    },
    {
      "entropy": 5.162092638015747,
      "epoch": 1.733107152954481,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047570943278635937,
      "loss": 5.0078,
      "mean_token_accuracy": 0.22737011611461638,
      "num_tokens": 4892876.0,
      "step": 6130
    },
    {
      "entropy": 5.1325681686401365,
      "epoch": 1.734520780322307,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047566292367271786,
      "loss": 5.0002,
      "mean_token_accuracy": 0.22855199724435807,
      "num_tokens": 4896589.0,
      "step": 6135
    },
    {
      "entropy": 5.115910005569458,
      "epoch": 1.7359344076901329,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047561637262176623,
      "loss": 4.9798,
      "mean_token_accuracy": 0.225470632314682,
      "num_tokens": 4900868.0,
      "step": 6140
    },
    {
      "entropy": 5.158762979507446,
      "epoch": 1.7373480350579587,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004755697796432333,
      "loss": 5.0943,
      "mean_token_accuracy": 0.23128358274698257,
      "num_tokens": 4904689.0,
      "step": 6145
    },
    {
      "entropy": 5.049402904510498,
      "epoch": 1.7387616624257847,
      "grad_norm": 1.609375,
      "learning_rate": 0.00047552314474685664,
      "loss": 4.8515,
      "mean_token_accuracy": 0.2319832667708397,
      "num_tokens": 4908421.0,
      "step": 6150
    },
    {
      "entropy": 4.9776286602020265,
      "epoch": 1.7401752897936102,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047547646794238277,
      "loss": 4.9337,
      "mean_token_accuracy": 0.23013666719198228,
      "num_tokens": 4912481.0,
      "step": 6155
    },
    {
      "entropy": 5.104219102859497,
      "epoch": 1.7415889171614363,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004754297492395668,
      "loss": 4.9868,
      "mean_token_accuracy": 0.22566144317388534,
      "num_tokens": 4916430.0,
      "step": 6160
    },
    {
      "entropy": 5.204376459121704,
      "epoch": 1.743002544529262,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004753829886481727,
      "loss": 4.9364,
      "mean_token_accuracy": 0.23793880194425582,
      "num_tokens": 4920322.0,
      "step": 6165
    },
    {
      "entropy": 5.1002038478851315,
      "epoch": 1.7444161718970879,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000475336186177973,
      "loss": 4.9961,
      "mean_token_accuracy": 0.22953584641218186,
      "num_tokens": 4924386.0,
      "step": 6170
    },
    {
      "entropy": 5.144687747955322,
      "epoch": 1.7458297992649139,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00047528934183874936,
      "loss": 5.0978,
      "mean_token_accuracy": 0.2217543825507164,
      "num_tokens": 4928711.0,
      "step": 6175
    },
    {
      "entropy": 5.150707626342774,
      "epoch": 1.7472434266327395,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004752424556402918,
      "loss": 4.9236,
      "mean_token_accuracy": 0.2297823578119278,
      "num_tokens": 4932810.0,
      "step": 6180
    },
    {
      "entropy": 5.024611091613769,
      "epoch": 1.7486570540005655,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004751955275923992,
      "loss": 4.8904,
      "mean_token_accuracy": 0.2405124709010124,
      "num_tokens": 4936893.0,
      "step": 6185
    },
    {
      "entropy": 4.999659490585327,
      "epoch": 1.7500706813683913,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047514855770487935,
      "loss": 4.9474,
      "mean_token_accuracy": 0.23304235190153122,
      "num_tokens": 4940942.0,
      "step": 6190
    },
    {
      "entropy": 5.156673383712769,
      "epoch": 1.751484308736217,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047510154598754865,
      "loss": 5.1461,
      "mean_token_accuracy": 0.21325360089540482,
      "num_tokens": 4945366.0,
      "step": 6195
    },
    {
      "entropy": 5.2605711936950685,
      "epoch": 1.752897936104043,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047505449245023216,
      "loss": 4.9732,
      "mean_token_accuracy": 0.23185187131166457,
      "num_tokens": 4949276.0,
      "step": 6200
    },
    {
      "entropy": 5.197354507446289,
      "epoch": 1.7543115634718687,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004750073971027639,
      "loss": 5.0028,
      "mean_token_accuracy": 0.23088634610176087,
      "num_tokens": 4953068.0,
      "step": 6205
    },
    {
      "entropy": 5.104533863067627,
      "epoch": 1.7557251908396947,
      "grad_norm": 1.5,
      "learning_rate": 0.0004749602599549865,
      "loss": 4.907,
      "mean_token_accuracy": 0.23583677113056184,
      "num_tokens": 4957148.0,
      "step": 6210
    },
    {
      "entropy": 5.135370111465454,
      "epoch": 1.7571388182075205,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004749130810167511,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2292344629764557,
      "num_tokens": 4961070.0,
      "step": 6215
    },
    {
      "entropy": 5.186055040359497,
      "epoch": 1.7585524455753463,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047486586029791823,
      "loss": 5.0448,
      "mean_token_accuracy": 0.21900554895401,
      "num_tokens": 4964883.0,
      "step": 6220
    },
    {
      "entropy": 5.070383310317993,
      "epoch": 1.7599660729431723,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047481859780835637,
      "loss": 5.0986,
      "mean_token_accuracy": 0.2199846014380455,
      "num_tokens": 4968825.0,
      "step": 6225
    },
    {
      "entropy": 5.1314915180206295,
      "epoch": 1.7613797003109979,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004747712935579431,
      "loss": 5.0692,
      "mean_token_accuracy": 0.22489855736494063,
      "num_tokens": 4973067.0,
      "step": 6230
    },
    {
      "entropy": 5.2085777759552006,
      "epoch": 1.762793327678824,
      "grad_norm": 1.5,
      "learning_rate": 0.000474723947556565,
      "loss": 5.0526,
      "mean_token_accuracy": 0.22984699308872222,
      "num_tokens": 4977215.0,
      "step": 6235
    },
    {
      "entropy": 5.188218021392823,
      "epoch": 1.7642069550466497,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004746765598141169,
      "loss": 5.009,
      "mean_token_accuracy": 0.21720004975795745,
      "num_tokens": 4980749.0,
      "step": 6240
    },
    {
      "entropy": 5.097091102600098,
      "epoch": 1.7656205824144755,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047462913034050245,
      "loss": 5.002,
      "mean_token_accuracy": 0.23229551166296006,
      "num_tokens": 4984757.0,
      "step": 6245
    },
    {
      "entropy": 5.10816535949707,
      "epoch": 1.7670342097823015,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047458165914563436,
      "loss": 5.0761,
      "mean_token_accuracy": 0.23077977448701859,
      "num_tokens": 4988657.0,
      "step": 6250
    },
    {
      "entropy": 5.142723655700683,
      "epoch": 1.768447837150127,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004745341462394336,
      "loss": 5.0707,
      "mean_token_accuracy": 0.23065107613801955,
      "num_tokens": 4992479.0,
      "step": 6255
    },
    {
      "entropy": 5.177563667297363,
      "epoch": 1.7698614645179531,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004744865916318302,
      "loss": 5.1391,
      "mean_token_accuracy": 0.2161106824874878,
      "num_tokens": 4996495.0,
      "step": 6260
    },
    {
      "entropy": 5.1951744079589846,
      "epoch": 1.771275091885779,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00047443899533276276,
      "loss": 5.064,
      "mean_token_accuracy": 0.22435715198516845,
      "num_tokens": 5000503.0,
      "step": 6265
    },
    {
      "entropy": 5.054482364654541,
      "epoch": 1.7726887192536047,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047439135735217845,
      "loss": 4.9418,
      "mean_token_accuracy": 0.23666587471961975,
      "num_tokens": 5004402.0,
      "step": 6270
    },
    {
      "entropy": 4.973062372207641,
      "epoch": 1.7741023466214307,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004743436777000336,
      "loss": 5.0406,
      "mean_token_accuracy": 0.22305506765842437,
      "num_tokens": 5008287.0,
      "step": 6275
    },
    {
      "entropy": 4.9560168266296385,
      "epoch": 1.7755159739892563,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047429595638629276,
      "loss": 4.9612,
      "mean_token_accuracy": 0.24013022780418397,
      "num_tokens": 5012217.0,
      "step": 6280
    },
    {
      "entropy": 5.1114496231079105,
      "epoch": 1.7769296013570823,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004742481934209295,
      "loss": 5.1234,
      "mean_token_accuracy": 0.2103919580578804,
      "num_tokens": 5016454.0,
      "step": 6285
    },
    {
      "entropy": 5.055655574798584,
      "epoch": 1.7783432287249081,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004742003888139258,
      "loss": 5.0566,
      "mean_token_accuracy": 0.22753520905971528,
      "num_tokens": 5020483.0,
      "step": 6290
    },
    {
      "entropy": 4.994224882125854,
      "epoch": 1.779756856092734,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004741525425752727,
      "loss": 4.9091,
      "mean_token_accuracy": 0.2405051991343498,
      "num_tokens": 5024091.0,
      "step": 6295
    },
    {
      "entropy": 5.031513118743897,
      "epoch": 1.78117048346056,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004741046547149699,
      "loss": 5.0545,
      "mean_token_accuracy": 0.22602088302373885,
      "num_tokens": 5028090.0,
      "step": 6300
    },
    {
      "entropy": 5.108484172821045,
      "epoch": 1.7825841108283855,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047405672524302524,
      "loss": 5.0568,
      "mean_token_accuracy": 0.22583350241184236,
      "num_tokens": 5032045.0,
      "step": 6305
    },
    {
      "entropy": 5.045698022842407,
      "epoch": 1.7839977381962115,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004740087541694561,
      "loss": 4.9465,
      "mean_token_accuracy": 0.225492425262928,
      "num_tokens": 5035996.0,
      "step": 6310
    },
    {
      "entropy": 5.027023649215698,
      "epoch": 1.7854113655640373,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004739607415042879,
      "loss": 4.9631,
      "mean_token_accuracy": 0.23109336644411088,
      "num_tokens": 5040001.0,
      "step": 6315
    },
    {
      "entropy": 4.89773268699646,
      "epoch": 1.7868249929318631,
      "grad_norm": 1.5,
      "learning_rate": 0.0004739126872575551,
      "loss": 4.826,
      "mean_token_accuracy": 0.2527568221092224,
      "num_tokens": 5043888.0,
      "step": 6320
    },
    {
      "entropy": 5.035908126831055,
      "epoch": 1.7882386202996892,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047386459143930064,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2338537871837616,
      "num_tokens": 5047760.0,
      "step": 6325
    },
    {
      "entropy": 5.193585348129273,
      "epoch": 1.7896522476675147,
      "grad_norm": 1.546875,
      "learning_rate": 0.00047381645405957635,
      "loss": 5.1285,
      "mean_token_accuracy": 0.21914148032665254,
      "num_tokens": 5051551.0,
      "step": 6330
    },
    {
      "entropy": 5.118231678009034,
      "epoch": 1.7910658750353408,
      "grad_norm": 1.484375,
      "learning_rate": 0.00047376827512844266,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2443781942129135,
      "num_tokens": 5055566.0,
      "step": 6335
    },
    {
      "entropy": 5.151401662826538,
      "epoch": 1.7924795024031666,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004737200546559685,
      "loss": 5.0734,
      "mean_token_accuracy": 0.20828960835933685,
      "num_tokens": 5059599.0,
      "step": 6340
    },
    {
      "entropy": 5.089802551269531,
      "epoch": 1.7938931297709924,
      "grad_norm": 1.34375,
      "learning_rate": 0.00047367179265223183,
      "loss": 4.9419,
      "mean_token_accuracy": 0.22843153029680252,
      "num_tokens": 5063541.0,
      "step": 6345
    },
    {
      "entropy": 5.008627128601074,
      "epoch": 1.7953067571388182,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000473623489127319,
      "loss": 4.9478,
      "mean_token_accuracy": 0.23515840172767638,
      "num_tokens": 5067615.0,
      "step": 6350
    },
    {
      "entropy": 5.020669412612915,
      "epoch": 1.796720384506644,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004735751440913253,
      "loss": 5.0178,
      "mean_token_accuracy": 0.21532591581344604,
      "num_tokens": 5071825.0,
      "step": 6355
    },
    {
      "entropy": 5.140336608886718,
      "epoch": 1.79813401187447,
      "grad_norm": 1.578125,
      "learning_rate": 0.00047352675755435436,
      "loss": 5.0298,
      "mean_token_accuracy": 0.22102514654397964,
      "num_tokens": 5075806.0,
      "step": 6360
    },
    {
      "entropy": 5.151509571075439,
      "epoch": 1.7995476392422958,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004734783295265188,
      "loss": 5.0248,
      "mean_token_accuracy": 0.2197818860411644,
      "num_tokens": 5079909.0,
      "step": 6365
    },
    {
      "entropy": 5.130090188980103,
      "epoch": 1.8009612666101216,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047342986001793966,
      "loss": 5.2183,
      "mean_token_accuracy": 0.21301428973674774,
      "num_tokens": 5083995.0,
      "step": 6370
    },
    {
      "entropy": 5.124442195892334,
      "epoch": 1.8023748939779474,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004733813490387468,
      "loss": 4.999,
      "mean_token_accuracy": 0.22955162227153778,
      "num_tokens": 5088086.0,
      "step": 6375
    },
    {
      "entropy": 5.182482147216797,
      "epoch": 1.8037885213457732,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004733327965990788,
      "loss": 5.0259,
      "mean_token_accuracy": 0.22501018643379211,
      "num_tokens": 5091816.0,
      "step": 6380
    },
    {
      "entropy": 5.156447410583496,
      "epoch": 1.8052021487135992,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004732842027090828,
      "loss": 4.9906,
      "mean_token_accuracy": 0.22198696434497833,
      "num_tokens": 5095672.0,
      "step": 6385
    },
    {
      "entropy": 5.037242126464844,
      "epoch": 1.806615776081425,
      "grad_norm": 1.5,
      "learning_rate": 0.0004732355673789145,
      "loss": 5.0997,
      "mean_token_accuracy": 0.21783643066883088,
      "num_tokens": 5099590.0,
      "step": 6390
    },
    {
      "entropy": 5.024095869064331,
      "epoch": 1.8080294034492508,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004731868906187385,
      "loss": 4.9639,
      "mean_token_accuracy": 0.23914585560560225,
      "num_tokens": 5103604.0,
      "step": 6395
    },
    {
      "entropy": 5.045601463317871,
      "epoch": 1.8094430308170766,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00047313817243872786,
      "loss": 4.946,
      "mean_token_accuracy": 0.23477836102247238,
      "num_tokens": 5107548.0,
      "step": 6400
    },
    {
      "entropy": 5.053776788711548,
      "epoch": 1.8108566581849024,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00047308941284906456,
      "loss": 5.0093,
      "mean_token_accuracy": 0.2315283641219139,
      "num_tokens": 5111535.0,
      "step": 6405
    },
    {
      "entropy": 5.156882810592651,
      "epoch": 1.8122702855527284,
      "grad_norm": 1.671875,
      "learning_rate": 0.00047304061185993884,
      "loss": 5.044,
      "mean_token_accuracy": 0.2238496333360672,
      "num_tokens": 5115314.0,
      "step": 6410
    },
    {
      "entropy": 5.1238726615905765,
      "epoch": 1.813683912920554,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004729917694815499,
      "loss": 4.965,
      "mean_token_accuracy": 0.23334530144929885,
      "num_tokens": 5119219.0,
      "step": 6415
    },
    {
      "entropy": 4.920544576644898,
      "epoch": 1.81509754028838,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004729428857241055,
      "loss": 4.8393,
      "mean_token_accuracy": 0.24091110080480577,
      "num_tokens": 5122851.0,
      "step": 6420
    },
    {
      "entropy": 4.99084882736206,
      "epoch": 1.8165111676562058,
      "grad_norm": 1.625,
      "learning_rate": 0.000472893960597822,
      "loss": 5.0506,
      "mean_token_accuracy": 0.22518180608749389,
      "num_tokens": 5126783.0,
      "step": 6425
    },
    {
      "entropy": 5.00876898765564,
      "epoch": 1.8179247950240316,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004728449941129244,
      "loss": 4.9149,
      "mean_token_accuracy": 0.24661077558994293,
      "num_tokens": 5130443.0,
      "step": 6430
    },
    {
      "entropy": 5.08847942352295,
      "epoch": 1.8193384223918576,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004727959862796466,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2358928307890892,
      "num_tokens": 5134324.0,
      "step": 6435
    },
    {
      "entropy": 5.131006288528442,
      "epoch": 1.8207520497596832,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047274693710823077,
      "loss": 4.9912,
      "mean_token_accuracy": 0.2229037970304489,
      "num_tokens": 5137950.0,
      "step": 6440
    },
    {
      "entropy": 5.050856304168701,
      "epoch": 1.8221656771275092,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004726978466089278,
      "loss": 4.8372,
      "mean_token_accuracy": 0.23366211205720902,
      "num_tokens": 5141849.0,
      "step": 6445
    },
    {
      "entropy": 5.006933307647705,
      "epoch": 1.823579304495335,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004726487147919976,
      "loss": 4.9889,
      "mean_token_accuracy": 0.2394751325249672,
      "num_tokens": 5145653.0,
      "step": 6450
    },
    {
      "entropy": 5.051721858978271,
      "epoch": 1.8249929318631608,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004725995416677081,
      "loss": 4.9429,
      "mean_token_accuracy": 0.23016407191753388,
      "num_tokens": 5149517.0,
      "step": 6455
    },
    {
      "entropy": 5.169414949417114,
      "epoch": 1.8264065592309868,
      "grad_norm": 1.53125,
      "learning_rate": 0.00047255032724633633,
      "loss": 5.1044,
      "mean_token_accuracy": 0.2209499403834343,
      "num_tokens": 5153525.0,
      "step": 6460
    },
    {
      "entropy": 5.062844610214233,
      "epoch": 1.8278201865988124,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004725010715381679,
      "loss": 5.0175,
      "mean_token_accuracy": 0.22005246728658676,
      "num_tokens": 5157762.0,
      "step": 6465
    },
    {
      "entropy": 5.16579270362854,
      "epoch": 1.8292338139666384,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00047245177455349673,
      "loss": 5.2011,
      "mean_token_accuracy": 0.21809534430503846,
      "num_tokens": 5161740.0,
      "step": 6470
    },
    {
      "entropy": 5.071216011047364,
      "epoch": 1.8306474413344642,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00047240243630262567,
      "loss": 4.8916,
      "mean_token_accuracy": 0.2420902281999588,
      "num_tokens": 5165755.0,
      "step": 6475
    },
    {
      "entropy": 5.077785921096802,
      "epoch": 1.83206106870229,
      "grad_norm": 1.515625,
      "learning_rate": 0.00047235305679586614,
      "loss": 4.9714,
      "mean_token_accuracy": 0.22164665311574935,
      "num_tokens": 5169805.0,
      "step": 6480
    },
    {
      "entropy": 5.084092426300049,
      "epoch": 1.833474696070116,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00047230363604353813,
      "loss": 5.0193,
      "mean_token_accuracy": 0.22743782848119737,
      "num_tokens": 5173983.0,
      "step": 6485
    },
    {
      "entropy": 5.223385763168335,
      "epoch": 1.8348883234379416,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00047225417405597023,
      "loss": 5.1395,
      "mean_token_accuracy": 0.22291941195726395,
      "num_tokens": 5178159.0,
      "step": 6490
    },
    {
      "entropy": 5.173782777786255,
      "epoch": 1.8363019508057676,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00047220467084349983,
      "loss": 4.9519,
      "mean_token_accuracy": 0.23269151002168656,
      "num_tokens": 5182204.0,
      "step": 6495
    },
    {
      "entropy": 5.136368131637573,
      "epoch": 1.8377155781735934,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047215512641647265,
      "loss": 4.8881,
      "mean_token_accuracy": 0.23314682096242906,
      "num_tokens": 5186179.0,
      "step": 6500
    },
    {
      "entropy": 5.045179271697998,
      "epoch": 1.8391292055414192,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004721055407852432,
      "loss": 4.9736,
      "mean_token_accuracy": 0.22942371815443038,
      "num_tokens": 5190184.0,
      "step": 6505
    },
    {
      "entropy": 5.186350345611572,
      "epoch": 1.8405428329092453,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004720559139601746,
      "loss": 5.1543,
      "mean_token_accuracy": 0.21901514232158661,
      "num_tokens": 5194126.0,
      "step": 6510
    },
    {
      "entropy": 5.142795848846435,
      "epoch": 1.8419564602770708,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004720062459516385,
      "loss": 4.9785,
      "mean_token_accuracy": 0.23701722770929337,
      "num_tokens": 5198214.0,
      "step": 6515
    },
    {
      "entropy": 5.118045234680176,
      "epoch": 1.8433700876448968,
      "grad_norm": 1.46875,
      "learning_rate": 0.00047195653677001536,
      "loss": 4.9482,
      "mean_token_accuracy": 0.23547864556312562,
      "num_tokens": 5202278.0,
      "step": 6520
    },
    {
      "entropy": 5.193373012542724,
      "epoch": 1.8447837150127226,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004719067864256939,
      "loss": 4.9998,
      "mean_token_accuracy": 0.22849097698926926,
      "num_tokens": 5206475.0,
      "step": 6525
    },
    {
      "entropy": 5.1291832447052,
      "epoch": 1.8461973423805484,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047185699492907174,
      "loss": 4.9997,
      "mean_token_accuracy": 0.22309511601924897,
      "num_tokens": 5210386.0,
      "step": 6530
    },
    {
      "entropy": 5.03418197631836,
      "epoch": 1.8476109697483745,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000471807162290555,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2296585276722908,
      "num_tokens": 5214342.0,
      "step": 6535
    },
    {
      "entropy": 4.972331857681274,
      "epoch": 1.8490245971162,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047175728852055825,
      "loss": 4.8382,
      "mean_token_accuracy": 0.23592381328344345,
      "num_tokens": 5218328.0,
      "step": 6540
    },
    {
      "entropy": 5.0296563625335695,
      "epoch": 1.850438224484026,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004717073736295051,
      "loss": 5.0185,
      "mean_token_accuracy": 0.22969842702150345,
      "num_tokens": 5222580.0,
      "step": 6545
    },
    {
      "entropy": 5.15314793586731,
      "epoch": 1.8518518518518519,
      "grad_norm": 1.5625,
      "learning_rate": 0.00047165741762782707,
      "loss": 4.9889,
      "mean_token_accuracy": 0.22327495515346527,
      "num_tokens": 5226568.0,
      "step": 6550
    },
    {
      "entropy": 5.097716331481934,
      "epoch": 1.8532654792196777,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004716074205259649,
      "loss": 4.9648,
      "mean_token_accuracy": 0.22715803384780883,
      "num_tokens": 5230830.0,
      "step": 6555
    },
    {
      "entropy": 5.288179349899292,
      "epoch": 1.8546791065875037,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004715573823343676,
      "loss": 5.1393,
      "mean_token_accuracy": 0.2170239046216011,
      "num_tokens": 5234648.0,
      "step": 6560
    },
    {
      "entropy": 5.221612787246704,
      "epoch": 1.8560927339553293,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004715073030634929,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2350559189915657,
      "num_tokens": 5238592.0,
      "step": 6565
    },
    {
      "entropy": 5.023984146118164,
      "epoch": 1.8575063613231553,
      "grad_norm": 1.625,
      "learning_rate": 0.00047145718272380706,
      "loss": 4.993,
      "mean_token_accuracy": 0.23441555202007294,
      "num_tokens": 5242596.0,
      "step": 6570
    },
    {
      "entropy": 5.119027519226075,
      "epoch": 1.858919988690981,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004714070213257849,
      "loss": 5.0153,
      "mean_token_accuracy": 0.21780619025230408,
      "num_tokens": 5246494.0,
      "step": 6575
    },
    {
      "entropy": 5.207915449142456,
      "epoch": 1.8603336160588069,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004713568188799097,
      "loss": 5.0368,
      "mean_token_accuracy": 0.21946613490581512,
      "num_tokens": 5250456.0,
      "step": 6580
    },
    {
      "entropy": 5.197103452682495,
      "epoch": 1.861747243426633,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004713065753966737,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2410424381494522,
      "num_tokens": 5254526.0,
      "step": 6585
    },
    {
      "entropy": 5.192329835891724,
      "epoch": 1.8631608707944585,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004712562908865774,
      "loss": 5.1091,
      "mean_token_accuracy": 0.22582194805145264,
      "num_tokens": 5258393.0,
      "step": 6590
    },
    {
      "entropy": 5.131658458709717,
      "epoch": 1.8645744981622845,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047120596536012976,
      "loss": 4.9633,
      "mean_token_accuracy": 0.2395295485854149,
      "num_tokens": 5262272.0,
      "step": 6595
    },
    {
      "entropy": 5.008411598205567,
      "epoch": 1.8659881255301103,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004711555988278488,
      "loss": 4.8953,
      "mean_token_accuracy": 0.2407392218708992,
      "num_tokens": 5266175.0,
      "step": 6600
    },
    {
      "entropy": 5.015675258636475,
      "epoch": 1.867401752897936,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00047110519130026066,
      "loss": 4.9361,
      "mean_token_accuracy": 0.23503223359584807,
      "num_tokens": 5269992.0,
      "step": 6605
    },
    {
      "entropy": 5.118558073043824,
      "epoch": 1.868815380265762,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004710547427879002,
      "loss": 4.9482,
      "mean_token_accuracy": 0.22060782909393312,
      "num_tokens": 5274235.0,
      "step": 6610
    },
    {
      "entropy": 5.042722415924072,
      "epoch": 1.8702290076335877,
      "grad_norm": 1.5,
      "learning_rate": 0.0004710042533013108,
      "loss": 4.9999,
      "mean_token_accuracy": 0.23334646373987197,
      "num_tokens": 5278104.0,
      "step": 6615
    },
    {
      "entropy": 5.218494462966919,
      "epoch": 1.8716426350014137,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004709537228510446,
      "loss": 5.1514,
      "mean_token_accuracy": 0.2111768215894699,
      "num_tokens": 5282017.0,
      "step": 6620
    },
    {
      "entropy": 5.26529221534729,
      "epoch": 1.8730562623692395,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000470903151447662,
      "loss": 5.1188,
      "mean_token_accuracy": 0.22293411046266556,
      "num_tokens": 5286509.0,
      "step": 6625
    },
    {
      "entropy": 5.204425430297851,
      "epoch": 1.8744698897370653,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004708525391017321,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22236587554216386,
      "num_tokens": 5290297.0,
      "step": 6630
    },
    {
      "entropy": 5.190113162994384,
      "epoch": 1.875883517104891,
      "grad_norm": 1.421875,
      "learning_rate": 0.00047080188582383287,
      "loss": 5.0672,
      "mean_token_accuracy": 0.22365201860666276,
      "num_tokens": 5294738.0,
      "step": 6635
    },
    {
      "entropy": 5.136478233337402,
      "epoch": 1.877297144472717,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047075119162455015,
      "loss": 5.0174,
      "mean_token_accuracy": 0.23802433758974076,
      "num_tokens": 5298676.0,
      "step": 6640
    },
    {
      "entropy": 5.006561183929444,
      "epoch": 1.878710771840543,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004707004565144789,
      "loss": 4.9773,
      "mean_token_accuracy": 0.23810056895017623,
      "num_tokens": 5302429.0,
      "step": 6645
    },
    {
      "entropy": 5.088086128234863,
      "epoch": 1.8801243992083687,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004706496805042225,
      "loss": 4.9903,
      "mean_token_accuracy": 0.23756980746984482,
      "num_tokens": 5306218.0,
      "step": 6650
    },
    {
      "entropy": 5.036809396743775,
      "epoch": 1.8815380265761945,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00047059886360439256,
      "loss": 4.916,
      "mean_token_accuracy": 0.24103206545114517,
      "num_tokens": 5310294.0,
      "step": 6655
    },
    {
      "entropy": 5.106139898300171,
      "epoch": 1.8829516539440203,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004705480058256098,
      "loss": 5.0414,
      "mean_token_accuracy": 0.22901678681373597,
      "num_tokens": 5313858.0,
      "step": 6660
    },
    {
      "entropy": 5.036658811569214,
      "epoch": 1.884365281311846,
      "grad_norm": 1.75,
      "learning_rate": 0.00047049710717850295,
      "loss": 4.9606,
      "mean_token_accuracy": 0.23796511590480804,
      "num_tokens": 5318083.0,
      "step": 6665
    },
    {
      "entropy": 5.107324886322021,
      "epoch": 1.8857789086796721,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047044616767370973,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2258594200015068,
      "num_tokens": 5322024.0,
      "step": 6670
    },
    {
      "entropy": 5.1088666915893555,
      "epoch": 1.887192536047498,
      "grad_norm": 1.484375,
      "learning_rate": 0.000470395187321876,
      "loss": 5.036,
      "mean_token_accuracy": 0.2214437887072563,
      "num_tokens": 5325903.0,
      "step": 6675
    },
    {
      "entropy": 5.104984188079834,
      "epoch": 1.8886061634153237,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004703441661336565,
      "loss": 4.9759,
      "mean_token_accuracy": 0.2373169481754303,
      "num_tokens": 5329502.0,
      "step": 6680
    },
    {
      "entropy": 5.1473531246185305,
      "epoch": 1.8900197907831495,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004702931041197142,
      "loss": 5.0352,
      "mean_token_accuracy": 0.2260161891579628,
      "num_tokens": 5333629.0,
      "step": 6685
    },
    {
      "entropy": 5.144501256942749,
      "epoch": 1.8914334181509753,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00047024200129072073,
      "loss": 5.0501,
      "mean_token_accuracy": 0.22809749096632004,
      "num_tokens": 5337551.0,
      "step": 6690
    },
    {
      "entropy": 5.1778223514556885,
      "epoch": 1.8928470455188013,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004701908576573564,
      "loss": 5.0163,
      "mean_token_accuracy": 0.2388780489563942,
      "num_tokens": 5341498.0,
      "step": 6695
    },
    {
      "entropy": 5.153984928131104,
      "epoch": 1.8942606728866271,
      "grad_norm": 1.59375,
      "learning_rate": 0.00047013967323030983,
      "loss": 5.0423,
      "mean_token_accuracy": 0.22395086288452148,
      "num_tokens": 5345574.0,
      "step": 6700
    },
    {
      "entropy": 4.982226037979126,
      "epoch": 1.895674300254453,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004700884480202783,
      "loss": 4.9241,
      "mean_token_accuracy": 0.24417694360017778,
      "num_tokens": 5349371.0,
      "step": 6705
    },
    {
      "entropy": 4.965596914291382,
      "epoch": 1.8970879276222787,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0004700371820379674,
      "loss": 5.0452,
      "mean_token_accuracy": 0.2340289145708084,
      "num_tokens": 5353485.0,
      "step": 6710
    },
    {
      "entropy": 5.014316654205322,
      "epoch": 1.8985015549901045,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004699858752940916,
      "loss": 4.9812,
      "mean_token_accuracy": 0.2299184247851372,
      "num_tokens": 5357320.0,
      "step": 6715
    },
    {
      "entropy": 5.141608810424804,
      "epoch": 1.8999151823579306,
      "grad_norm": 1.625,
      "learning_rate": 0.0004699345277993737,
      "loss": 5.1023,
      "mean_token_accuracy": 0.22708941996097565,
      "num_tokens": 5360879.0,
      "step": 6720
    },
    {
      "entropy": 5.155096483230591,
      "epoch": 1.9013288097257561,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004698831395645448,
      "loss": 5.022,
      "mean_token_accuracy": 0.22912110835313798,
      "num_tokens": 5365029.0,
      "step": 6725
    },
    {
      "entropy": 5.0258971691131595,
      "epoch": 1.9027424370935822,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046983171060034493,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2333139255642891,
      "num_tokens": 5369020.0,
      "step": 6730
    },
    {
      "entropy": 5.017714166641236,
      "epoch": 1.904156064461408,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004697802409175223,
      "loss": 4.9886,
      "mean_token_accuracy": 0.22181347459554673,
      "num_tokens": 5373092.0,
      "step": 6735
    },
    {
      "entropy": 5.160233020782471,
      "epoch": 1.9055696918292337,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046972873052683384,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2276818186044693,
      "num_tokens": 5376880.0,
      "step": 6740
    },
    {
      "entropy": 4.987977790832519,
      "epoch": 1.9069833191970598,
      "grad_norm": 1.5,
      "learning_rate": 0.0004696771794390449,
      "loss": 4.8936,
      "mean_token_accuracy": 0.23205955624580382,
      "num_tokens": 5380654.0,
      "step": 6745
    },
    {
      "entropy": 4.979429721832275,
      "epoch": 1.9083969465648853,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046962558766492924,
      "loss": 4.9668,
      "mean_token_accuracy": 0.2346696972846985,
      "num_tokens": 5384762.0,
      "step": 6750
    },
    {
      "entropy": 4.9728655338287355,
      "epoch": 1.9098105739327114,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004695739552152693,
      "loss": 4.851,
      "mean_token_accuracy": 0.24139469712972642,
      "num_tokens": 5388503.0,
      "step": 6755
    },
    {
      "entropy": 5.102569913864135,
      "epoch": 1.9112242013005372,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046952228210085596,
      "loss": 5.0939,
      "mean_token_accuracy": 0.21765656024217606,
      "num_tokens": 5392279.0,
      "step": 6760
    },
    {
      "entropy": 5.128580141067505,
      "epoch": 1.912637828668363,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046947056833248854,
      "loss": 5.1105,
      "mean_token_accuracy": 0.22523009777069092,
      "num_tokens": 5396326.0,
      "step": 6765
    },
    {
      "entropy": 5.0750995635986325,
      "epoch": 1.914051456036189,
      "grad_norm": 1.5,
      "learning_rate": 0.0004694188139209748,
      "loss": 5.0347,
      "mean_token_accuracy": 0.22917448431253434,
      "num_tokens": 5400557.0,
      "step": 6770
    },
    {
      "entropy": 5.1119132995605465,
      "epoch": 1.9154650834040146,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046936701887713135,
      "loss": 5.0581,
      "mean_token_accuracy": 0.21875028014183046,
      "num_tokens": 5404418.0,
      "step": 6775
    },
    {
      "entropy": 5.036420392990112,
      "epoch": 1.9168787107718406,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004693151832117828,
      "loss": 4.8346,
      "mean_token_accuracy": 0.23584452271461487,
      "num_tokens": 5408467.0,
      "step": 6780
    },
    {
      "entropy": 5.014028596878052,
      "epoch": 1.9182923381396664,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004692633069357626,
      "loss": 5.0245,
      "mean_token_accuracy": 0.2318779557943344,
      "num_tokens": 5412603.0,
      "step": 6785
    },
    {
      "entropy": 5.050658130645752,
      "epoch": 1.9197059655074922,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004692113900599125,
      "loss": 4.9889,
      "mean_token_accuracy": 0.23316234797239305,
      "num_tokens": 5416529.0,
      "step": 6790
    },
    {
      "entropy": 5.020613431930542,
      "epoch": 1.9211195928753182,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000469159432595083,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2330153152346611,
      "num_tokens": 5420514.0,
      "step": 6795
    },
    {
      "entropy": 4.993831396102905,
      "epoch": 1.9225332202431438,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004691074345521326,
      "loss": 4.9285,
      "mean_token_accuracy": 0.23852734118700028,
      "num_tokens": 5424539.0,
      "step": 6800
    },
    {
      "entropy": 4.9680588722229,
      "epoch": 1.9239468476109698,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004690553959419286,
      "loss": 4.954,
      "mean_token_accuracy": 0.23611341863870622,
      "num_tokens": 5428507.0,
      "step": 6805
    },
    {
      "entropy": 4.946990966796875,
      "epoch": 1.9253604749787956,
      "grad_norm": 1.4765625,
      "learning_rate": 0.000469003316775347,
      "loss": 4.9133,
      "mean_token_accuracy": 0.24908787459135057,
      "num_tokens": 5432168.0,
      "step": 6810
    },
    {
      "entropy": 5.085066175460815,
      "epoch": 1.9267741023466214,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046895119706327185,
      "loss": 5.0551,
      "mean_token_accuracy": 0.22704901397228242,
      "num_tokens": 5435848.0,
      "step": 6815
    },
    {
      "entropy": 4.982158470153808,
      "epoch": 1.9281877297144474,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046889903681659583,
      "loss": 4.9967,
      "mean_token_accuracy": 0.23254324048757552,
      "num_tokens": 5439823.0,
      "step": 6820
    },
    {
      "entropy": 5.0059116840362545,
      "epoch": 1.929601357082273,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004688468360462202,
      "loss": 5.0802,
      "mean_token_accuracy": 0.23209846019744873,
      "num_tokens": 5443836.0,
      "step": 6825
    },
    {
      "entropy": 5.048571825027466,
      "epoch": 1.931014984450099,
      "grad_norm": 1.625,
      "learning_rate": 0.0004687945947630546,
      "loss": 4.9855,
      "mean_token_accuracy": 0.23392627090215684,
      "num_tokens": 5447638.0,
      "step": 6830
    },
    {
      "entropy": 4.954223012924194,
      "epoch": 1.9324286118179248,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046874231297801695,
      "loss": 5.0266,
      "mean_token_accuracy": 0.23423477858304978,
      "num_tokens": 5451444.0,
      "step": 6835
    },
    {
      "entropy": 5.140878772735595,
      "epoch": 1.9338422391857506,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004686899907020341,
      "loss": 5.2253,
      "mean_token_accuracy": 0.2171703964471817,
      "num_tokens": 5455687.0,
      "step": 6840
    },
    {
      "entropy": 5.157349300384522,
      "epoch": 1.9352558665535766,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004686376279460408,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22813319712877272,
      "num_tokens": 5459406.0,
      "step": 6845
    },
    {
      "entropy": 4.993503522872925,
      "epoch": 1.9366694939214022,
      "grad_norm": 1.4375,
      "learning_rate": 0.00046858522472098074,
      "loss": 4.8308,
      "mean_token_accuracy": 0.2526347920298576,
      "num_tokens": 5463227.0,
      "step": 6850
    },
    {
      "entropy": 5.255039882659912,
      "epoch": 1.9380831212892282,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046853278103780585,
      "loss": 5.3385,
      "mean_token_accuracy": 0.19318314492702485,
      "num_tokens": 5467154.0,
      "step": 6855
    },
    {
      "entropy": 5.166647815704346,
      "epoch": 1.939496748657054,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004684802969074764,
      "loss": 5.0333,
      "mean_token_accuracy": 0.22455305904150008,
      "num_tokens": 5470977.0,
      "step": 6860
    },
    {
      "entropy": 5.063953447341919,
      "epoch": 1.9409103760248798,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004684277723409614,
      "loss": 4.9719,
      "mean_token_accuracy": 0.2294907808303833,
      "num_tokens": 5474824.0,
      "step": 6865
    },
    {
      "entropy": 4.938325214385986,
      "epoch": 1.9423240033927058,
      "grad_norm": 1.546875,
      "learning_rate": 0.000468375207349238,
      "loss": 4.929,
      "mean_token_accuracy": 0.24076665192842484,
      "num_tokens": 5478873.0,
      "step": 6870
    },
    {
      "entropy": 5.038755178451538,
      "epoch": 1.9437376307605314,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004683226019432922,
      "loss": 4.9438,
      "mean_token_accuracy": 0.23168932497501374,
      "num_tokens": 5483208.0,
      "step": 6875
    },
    {
      "entropy": 5.021457815170288,
      "epoch": 1.9451512581283574,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046826995613411806,
      "loss": 4.9296,
      "mean_token_accuracy": 0.22730178982019425,
      "num_tokens": 5487387.0,
      "step": 6880
    },
    {
      "entropy": 4.996645116806031,
      "epoch": 1.9465648854961832,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004682172699327182,
      "loss": 4.8868,
      "mean_token_accuracy": 0.23785894215106965,
      "num_tokens": 5491486.0,
      "step": 6885
    },
    {
      "entropy": 5.093940258026123,
      "epoch": 1.947978512864009,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046816454335010365,
      "loss": 4.9707,
      "mean_token_accuracy": 0.22918916642665862,
      "num_tokens": 5495333.0,
      "step": 6890
    },
    {
      "entropy": 5.090713214874268,
      "epoch": 1.949392140231835,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046811177639729417,
      "loss": 5.016,
      "mean_token_accuracy": 0.22775486409664153,
      "num_tokens": 5499351.0,
      "step": 6895
    },
    {
      "entropy": 4.97871880531311,
      "epoch": 1.9508057675996606,
      "grad_norm": 1.515625,
      "learning_rate": 0.00046805896908531756,
      "loss": 4.9801,
      "mean_token_accuracy": 0.23614903837442397,
      "num_tokens": 5503260.0,
      "step": 6900
    },
    {
      "entropy": 5.035440731048584,
      "epoch": 1.9522193949674866,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004680061214252102,
      "loss": 4.9822,
      "mean_token_accuracy": 0.22942784130573274,
      "num_tokens": 5507223.0,
      "step": 6905
    },
    {
      "entropy": 5.044527149200439,
      "epoch": 1.9536330223353124,
      "grad_norm": 1.5078125,
      "learning_rate": 0.000467953233428017,
      "loss": 4.9623,
      "mean_token_accuracy": 0.22872766703367234,
      "num_tokens": 5511347.0,
      "step": 6910
    },
    {
      "entropy": 5.011592674255371,
      "epoch": 1.9550466497031382,
      "grad_norm": 1.5,
      "learning_rate": 0.0004679003051047912,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2232879638671875,
      "num_tokens": 5515350.0,
      "step": 6915
    },
    {
      "entropy": 4.959643316268921,
      "epoch": 1.9564602770709643,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004678473364665945,
      "loss": 4.9327,
      "mean_token_accuracy": 0.22906329333782197,
      "num_tokens": 5519262.0,
      "step": 6920
    },
    {
      "entropy": 4.9847430229187015,
      "epoch": 1.9578739044387898,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046779432752449705,
      "loss": 5.0135,
      "mean_token_accuracy": 0.22321516573429107,
      "num_tokens": 5523209.0,
      "step": 6925
    },
    {
      "entropy": 5.012733411788941,
      "epoch": 1.9592875318066159,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004677412782895773,
      "loss": 4.9406,
      "mean_token_accuracy": 0.234261591732502,
      "num_tokens": 5526903.0,
      "step": 6930
    },
    {
      "entropy": 5.011602067947388,
      "epoch": 1.9607011591744417,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004676881887729222,
      "loss": 4.8991,
      "mean_token_accuracy": 0.24087762385606765,
      "num_tokens": 5530667.0,
      "step": 6935
    },
    {
      "entropy": 4.950695419311524,
      "epoch": 1.9621147865422675,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046763505898562735,
      "loss": 4.949,
      "mean_token_accuracy": 0.23225943893194198,
      "num_tokens": 5534413.0,
      "step": 6940
    },
    {
      "entropy": 4.959881544113159,
      "epoch": 1.9635284139100933,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00046758188893879633,
      "loss": 5.0127,
      "mean_token_accuracy": 0.22849936336278914,
      "num_tokens": 5538662.0,
      "step": 6945
    },
    {
      "entropy": 5.0057579517364506,
      "epoch": 1.964942041277919,
      "grad_norm": 1.453125,
      "learning_rate": 0.00046752867864354136,
      "loss": 4.835,
      "mean_token_accuracy": 0.25231233835220335,
      "num_tokens": 5542602.0,
      "step": 6950
    },
    {
      "entropy": 5.094666242599487,
      "epoch": 1.966355668645745,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046747542811098315,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22542366832494737,
      "num_tokens": 5546244.0,
      "step": 6955
    },
    {
      "entropy": 4.997423267364502,
      "epoch": 1.9677692960135709,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004674221373522507,
      "loss": 5.0143,
      "mean_token_accuracy": 0.22740250080823898,
      "num_tokens": 5550116.0,
      "step": 6960
    },
    {
      "entropy": 4.988827037811279,
      "epoch": 1.9691829233813967,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004673688063784815,
      "loss": 4.937,
      "mean_token_accuracy": 0.23751620948314667,
      "num_tokens": 5554116.0,
      "step": 6965
    },
    {
      "entropy": 4.937193965911865,
      "epoch": 1.9705965507492225,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046731543520082123,
      "loss": 4.8389,
      "mean_token_accuracy": 0.24395109564065934,
      "num_tokens": 5557911.0,
      "step": 6970
    },
    {
      "entropy": 4.988838243484497,
      "epoch": 1.9720101781170483,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004672620238304243,
      "loss": 4.9476,
      "mean_token_accuracy": 0.2380879819393158,
      "num_tokens": 5561828.0,
      "step": 6975
    },
    {
      "entropy": 5.062899351119995,
      "epoch": 1.9734238054848743,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046720857227845324,
      "loss": 4.9252,
      "mean_token_accuracy": 0.24040958434343337,
      "num_tokens": 5565976.0,
      "step": 6980
    },
    {
      "entropy": 4.972138977050781,
      "epoch": 1.9748374328527,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046715508055607925,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22846325039863585,
      "num_tokens": 5569654.0,
      "step": 6985
    },
    {
      "entropy": 4.8895775318145756,
      "epoch": 1.9762510602205259,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004671015486744817,
      "loss": 4.9753,
      "mean_token_accuracy": 0.22724231332540512,
      "num_tokens": 5573440.0,
      "step": 6990
    },
    {
      "entropy": 4.99875545501709,
      "epoch": 1.9776646875883517,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004670479766448483,
      "loss": 4.9933,
      "mean_token_accuracy": 0.23137834519147873,
      "num_tokens": 5577667.0,
      "step": 6995
    },
    {
      "entropy": 4.967022562026978,
      "epoch": 1.9790783149561775,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004669943644783754,
      "loss": 4.9089,
      "mean_token_accuracy": 0.2317409783601761,
      "num_tokens": 5581520.0,
      "step": 7000
    },
    {
      "epoch": 1.9790783149561775,
      "eval_entropy": 4.846932500410722,
      "eval_loss": 5.232217788696289,
      "eval_mean_token_accuracy": 0.22721486390270637,
      "eval_num_tokens": 5581520.0,
      "eval_runtime": 6.7497,
      "eval_samples_per_second": 1671.782,
      "eval_steps_per_second": 209.047,
      "step": 7000
    },
    {
      "entropy": 4.9216163635253904,
      "epoch": 1.9804919423240035,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004669407121862676,
      "loss": 4.8197,
      "mean_token_accuracy": 0.24163173884153366,
      "num_tokens": 5585338.0,
      "step": 7005
    },
    {
      "entropy": 5.10053014755249,
      "epoch": 1.981905569691829,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046688701977973783,
      "loss": 5.0331,
      "mean_token_accuracy": 0.22461846470832825,
      "num_tokens": 5589332.0,
      "step": 7010
    },
    {
      "entropy": 4.976572322845459,
      "epoch": 1.983319197059655,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004668332872700075,
      "loss": 4.9018,
      "mean_token_accuracy": 0.23370043337345123,
      "num_tokens": 5593048.0,
      "step": 7015
    },
    {
      "entropy": 5.000931024551392,
      "epoch": 1.984732824427481,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004667795146683065,
      "loss": 4.9814,
      "mean_token_accuracy": 0.238823401927948,
      "num_tokens": 5596984.0,
      "step": 7020
    },
    {
      "entropy": 5.057454442977905,
      "epoch": 1.9861464517953067,
      "grad_norm": 2.484375,
      "learning_rate": 0.00046672570198587283,
      "loss": 5.1053,
      "mean_token_accuracy": 0.22054912596940995,
      "num_tokens": 5600661.0,
      "step": 7025
    },
    {
      "entropy": 5.024930334091186,
      "epoch": 1.9875600791631327,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046667184923395306,
      "loss": 4.9073,
      "mean_token_accuracy": 0.23663445860147475,
      "num_tokens": 5604479.0,
      "step": 7030
    },
    {
      "entropy": 5.091260623931885,
      "epoch": 1.9889737065309583,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000466617956423802,
      "loss": 4.9606,
      "mean_token_accuracy": 0.23344213962554933,
      "num_tokens": 5608603.0,
      "step": 7035
    },
    {
      "entropy": 4.935049962997437,
      "epoch": 1.9903873338987843,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000466564023566683,
      "loss": 4.8885,
      "mean_token_accuracy": 0.2323440656065941,
      "num_tokens": 5612579.0,
      "step": 7040
    },
    {
      "entropy": 5.055939149856568,
      "epoch": 1.99180096126661,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046651005067386767,
      "loss": 4.9965,
      "mean_token_accuracy": 0.2178351551294327,
      "num_tokens": 5616768.0,
      "step": 7045
    },
    {
      "entropy": 5.0345381736755375,
      "epoch": 1.993214588634436,
      "grad_norm": 1.625,
      "learning_rate": 0.000466456037756636,
      "loss": 4.8787,
      "mean_token_accuracy": 0.24052711874246596,
      "num_tokens": 5620574.0,
      "step": 7050
    },
    {
      "entropy": 5.070065641403199,
      "epoch": 1.994628216002262,
      "grad_norm": 1.625,
      "learning_rate": 0.0004664019848262764,
      "loss": 5.105,
      "mean_token_accuracy": 0.22373791188001632,
      "num_tokens": 5624740.0,
      "step": 7055
    },
    {
      "entropy": 4.904339790344238,
      "epoch": 1.9960418433700875,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00046634789189408554,
      "loss": 5.0008,
      "mean_token_accuracy": 0.2316736862063408,
      "num_tokens": 5628361.0,
      "step": 7060
    },
    {
      "entropy": 5.02471432685852,
      "epoch": 1.9974554707379135,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046629375897136856,
      "loss": 4.9074,
      "mean_token_accuracy": 0.2363758221268654,
      "num_tokens": 5632431.0,
      "step": 7065
    },
    {
      "entropy": 5.124822378158569,
      "epoch": 1.9988690981057393,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004662395860694388,
      "loss": 5.0238,
      "mean_token_accuracy": 0.22634965032339097,
      "num_tokens": 5636495.0,
      "step": 7070
    },
    {
      "entropy": 5.166530704498291,
      "epoch": 2.000282725473565,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004661853731996182,
      "loss": 5.1133,
      "mean_token_accuracy": 0.2082611948251724,
      "num_tokens": 5640304.0,
      "step": 7075
    },
    {
      "entropy": 5.075640344619751,
      "epoch": 2.001696352841391,
      "grad_norm": 1.53125,
      "learning_rate": 0.00046613112037323693,
      "loss": 4.7033,
      "mean_token_accuracy": 0.24516110867261887,
      "num_tokens": 5644231.0,
      "step": 7080
    },
    {
      "entropy": 4.830364799499511,
      "epoch": 2.0031099802092167,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046607682760163333,
      "loss": 4.6062,
      "mean_token_accuracy": 0.24832604676485062,
      "num_tokens": 5648266.0,
      "step": 7085
    },
    {
      "entropy": 4.882409286499024,
      "epoch": 2.0045236075770427,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004660224948961544,
      "loss": 4.7542,
      "mean_token_accuracy": 0.22261690646409987,
      "num_tokens": 5651856.0,
      "step": 7090
    },
    {
      "entropy": 4.768161296844482,
      "epoch": 2.0059372349448688,
      "grad_norm": 1.546875,
      "learning_rate": 0.00046596812226815516,
      "loss": 4.585,
      "mean_token_accuracy": 0.249821212887764,
      "num_tokens": 5655556.0,
      "step": 7095
    },
    {
      "entropy": 4.88502721786499,
      "epoch": 2.0073508623126943,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004659137097289994,
      "loss": 4.7194,
      "mean_token_accuracy": 0.2401501163840294,
      "num_tokens": 5659647.0,
      "step": 7100
    },
    {
      "entropy": 4.967523384094238,
      "epoch": 2.0087644896805203,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046585925729005883,
      "loss": 4.7467,
      "mean_token_accuracy": 0.22838964611291884,
      "num_tokens": 5663240.0,
      "step": 7105
    },
    {
      "entropy": 4.9127099990844725,
      "epoch": 2.010178117048346,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004658047649627138,
      "loss": 4.7059,
      "mean_token_accuracy": 0.2456618294119835,
      "num_tokens": 5667333.0,
      "step": 7110
    },
    {
      "entropy": 4.948296308517456,
      "epoch": 2.011591744416172,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046575023275835276,
      "loss": 4.7785,
      "mean_token_accuracy": 0.23617770969867707,
      "num_tokens": 5671261.0,
      "step": 7115
    },
    {
      "entropy": 4.914106655120849,
      "epoch": 2.0130053717839975,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046569566068837263,
      "loss": 4.7095,
      "mean_token_accuracy": 0.23749754577875137,
      "num_tokens": 5675454.0,
      "step": 7120
    },
    {
      "entropy": 4.929000329971314,
      "epoch": 2.0144189991518235,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004656410487641785,
      "loss": 4.7514,
      "mean_token_accuracy": 0.24219815135002137,
      "num_tokens": 5679490.0,
      "step": 7125
    },
    {
      "entropy": 5.005086469650268,
      "epoch": 2.0158326265196496,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004655863969971843,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23792520314455032,
      "num_tokens": 5683574.0,
      "step": 7130
    },
    {
      "entropy": 4.930145931243897,
      "epoch": 2.017246253887475,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046553170539881156,
      "loss": 4.707,
      "mean_token_accuracy": 0.246723972260952,
      "num_tokens": 5687542.0,
      "step": 7135
    },
    {
      "entropy": 4.877951383590698,
      "epoch": 2.018659881255301,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004654769739804906,
      "loss": 4.6612,
      "mean_token_accuracy": 0.2518881171941757,
      "num_tokens": 5691202.0,
      "step": 7140
    },
    {
      "entropy": 4.848537969589233,
      "epoch": 2.0200735086231267,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046542220275366,
      "loss": 4.7683,
      "mean_token_accuracy": 0.24416338354349137,
      "num_tokens": 5695476.0,
      "step": 7145
    },
    {
      "entropy": 4.861104249954224,
      "epoch": 2.0214871359909528,
      "grad_norm": 1.5,
      "learning_rate": 0.00046536739172976647,
      "loss": 4.7532,
      "mean_token_accuracy": 0.24856139868497848,
      "num_tokens": 5699530.0,
      "step": 7150
    },
    {
      "entropy": 4.927355098724365,
      "epoch": 2.0229007633587788,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046531254092026535,
      "loss": 4.6614,
      "mean_token_accuracy": 0.25240073502063753,
      "num_tokens": 5703573.0,
      "step": 7155
    },
    {
      "entropy": 4.835448789596557,
      "epoch": 2.0243143907266044,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00046525765033662007,
      "loss": 4.6931,
      "mean_token_accuracy": 0.24382925480604173,
      "num_tokens": 5707953.0,
      "step": 7160
    },
    {
      "entropy": 4.8902770519256595,
      "epoch": 2.0257280180944304,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046520271999030225,
      "loss": 4.7244,
      "mean_token_accuracy": 0.24274213165044783,
      "num_tokens": 5712018.0,
      "step": 7165
    },
    {
      "entropy": 4.898256063461304,
      "epoch": 2.027141645462256,
      "grad_norm": 1.8828125,
      "learning_rate": 0.0004651477498927922,
      "loss": 4.7335,
      "mean_token_accuracy": 0.23825664222240447,
      "num_tokens": 5715785.0,
      "step": 7170
    },
    {
      "entropy": 4.891660690307617,
      "epoch": 2.028555272830082,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004650927400555783,
      "loss": 4.6665,
      "mean_token_accuracy": 0.24565665274858475,
      "num_tokens": 5719699.0,
      "step": 7175
    },
    {
      "entropy": 4.8627808570861815,
      "epoch": 2.029968900197908,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00046503769049015714,
      "loss": 4.6801,
      "mean_token_accuracy": 0.24087597131729127,
      "num_tokens": 5723594.0,
      "step": 7180
    },
    {
      "entropy": 4.841994190216065,
      "epoch": 2.0313825275657336,
      "grad_norm": 1.75,
      "learning_rate": 0.00046498260120803395,
      "loss": 4.6714,
      "mean_token_accuracy": 0.24189623594284057,
      "num_tokens": 5727474.0,
      "step": 7185
    },
    {
      "entropy": 4.897206354141235,
      "epoch": 2.0327961549335596,
      "grad_norm": 1.625,
      "learning_rate": 0.00046492747222072185,
      "loss": 4.6345,
      "mean_token_accuracy": 0.2509519994258881,
      "num_tokens": 5731264.0,
      "step": 7190
    },
    {
      "entropy": 4.873613739013672,
      "epoch": 2.034209782301385,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004648723035397425,
      "loss": 4.796,
      "mean_token_accuracy": 0.24578043520450593,
      "num_tokens": 5735065.0,
      "step": 7195
    },
    {
      "entropy": 4.831220054626465,
      "epoch": 2.035623409669211,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004648170951766259,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24227618873119355,
      "num_tokens": 5738974.0,
      "step": 7200
    },
    {
      "entropy": 4.8045690059661865,
      "epoch": 2.037037037037037,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004647618471429102,
      "loss": 4.5912,
      "mean_token_accuracy": 0.24921780377626418,
      "num_tokens": 5742782.0,
      "step": 7205
    },
    {
      "entropy": 4.917854261398316,
      "epoch": 2.0384506644048628,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004647065594501418,
      "loss": 4.76,
      "mean_token_accuracy": 0.22202645242214203,
      "num_tokens": 5746991.0,
      "step": 7210
    },
    {
      "entropy": 4.944791316986084,
      "epoch": 2.039864291772689,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046465123210987576,
      "loss": 4.7706,
      "mean_token_accuracy": 0.23736041337251662,
      "num_tokens": 5750955.0,
      "step": 7215
    },
    {
      "entropy": 4.920100450515747,
      "epoch": 2.0412779191405144,
      "grad_norm": 1.75,
      "learning_rate": 0.00046459586513367487,
      "loss": 4.6504,
      "mean_token_accuracy": 0.24948371350765228,
      "num_tokens": 5754809.0,
      "step": 7220
    },
    {
      "entropy": 4.919025659561157,
      "epoch": 2.0426915465083404,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004645404585331106,
      "loss": 4.7559,
      "mean_token_accuracy": 0.240752175450325,
      "num_tokens": 5758658.0,
      "step": 7225
    },
    {
      "entropy": 4.966362953186035,
      "epoch": 2.0441051738761664,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004644850123197626,
      "loss": 4.7961,
      "mean_token_accuracy": 0.2339584544301033,
      "num_tokens": 5762965.0,
      "step": 7230
    },
    {
      "entropy": 4.958462333679199,
      "epoch": 2.045518801243992,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004644295265052187,
      "loss": 4.7653,
      "mean_token_accuracy": 0.23920557349920274,
      "num_tokens": 5767204.0,
      "step": 7235
    },
    {
      "entropy": 5.09040265083313,
      "epoch": 2.046932428611818,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046437400110107527,
      "loss": 4.7766,
      "mean_token_accuracy": 0.23008474558591843,
      "num_tokens": 5771239.0,
      "step": 7240
    },
    {
      "entropy": 4.925770425796509,
      "epoch": 2.0483460559796436,
      "grad_norm": 1.5,
      "learning_rate": 0.0004643184361189366,
      "loss": 4.7821,
      "mean_token_accuracy": 0.23429293185472488,
      "num_tokens": 5775345.0,
      "step": 7245
    },
    {
      "entropy": 4.964169073104858,
      "epoch": 2.0497596833474696,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004642628315704155,
      "loss": 4.6943,
      "mean_token_accuracy": 0.24238999485969542,
      "num_tokens": 5779294.0,
      "step": 7250
    },
    {
      "entropy": 4.930596733093262,
      "epoch": 2.0511733107152956,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046420718746713297,
      "loss": 4.6772,
      "mean_token_accuracy": 0.24291148483753205,
      "num_tokens": 5783362.0,
      "step": 7255
    },
    {
      "entropy": 4.937539720535279,
      "epoch": 2.052586938083121,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004641515038207182,
      "loss": 4.6456,
      "mean_token_accuracy": 0.24585285037755966,
      "num_tokens": 5787134.0,
      "step": 7260
    },
    {
      "entropy": 4.859331655502319,
      "epoch": 2.0540005654509472,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004640957806428089,
      "loss": 4.7461,
      "mean_token_accuracy": 0.24490924030542374,
      "num_tokens": 5791116.0,
      "step": 7265
    },
    {
      "entropy": 4.844360399246216,
      "epoch": 2.055414192818773,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004640400179450508,
      "loss": 4.7457,
      "mean_token_accuracy": 0.23751633614301682,
      "num_tokens": 5795292.0,
      "step": 7270
    },
    {
      "entropy": 4.870352172851563,
      "epoch": 2.056827820186599,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004639842157390979,
      "loss": 4.6254,
      "mean_token_accuracy": 0.25084561556577684,
      "num_tokens": 5799138.0,
      "step": 7275
    },
    {
      "entropy": 4.832441186904907,
      "epoch": 2.058241447554425,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046392837403661245,
      "loss": 4.7753,
      "mean_token_accuracy": 0.23748939484357834,
      "num_tokens": 5803289.0,
      "step": 7280
    },
    {
      "entropy": 4.790563774108887,
      "epoch": 2.0596550749222504,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00046387249284926535,
      "loss": 4.6533,
      "mean_token_accuracy": 0.2480671986937523,
      "num_tokens": 5807286.0,
      "step": 7285
    },
    {
      "entropy": 4.851416730880738,
      "epoch": 2.0610687022900764,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000463816572188735,
      "loss": 4.7695,
      "mean_token_accuracy": 0.24022811502218247,
      "num_tokens": 5811170.0,
      "step": 7290
    },
    {
      "entropy": 4.849501132965088,
      "epoch": 2.062482329657902,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004637606120667087,
      "loss": 4.5033,
      "mean_token_accuracy": 0.26361607760190964,
      "num_tokens": 5815059.0,
      "step": 7295
    },
    {
      "entropy": 5.007969570159912,
      "epoch": 2.063895957025728,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004637046124948818,
      "loss": 4.814,
      "mean_token_accuracy": 0.23017711341381072,
      "num_tokens": 5819109.0,
      "step": 7300
    },
    {
      "entropy": 4.857758712768555,
      "epoch": 2.065309584393554,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004636485734849578,
      "loss": 4.7568,
      "mean_token_accuracy": 0.2413598984479904,
      "num_tokens": 5823249.0,
      "step": 7305
    },
    {
      "entropy": 4.883078002929688,
      "epoch": 2.0667232117613796,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004635924950486485,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23116111606359482,
      "num_tokens": 5826936.0,
      "step": 7310
    },
    {
      "entropy": 5.090041399002075,
      "epoch": 2.0681368391292057,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000463536377197674,
      "loss": 4.8439,
      "mean_token_accuracy": 0.23178970962762832,
      "num_tokens": 5830884.0,
      "step": 7315
    },
    {
      "entropy": 5.001302909851074,
      "epoch": 2.0695504664970312,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004634802199437625,
      "loss": 4.7187,
      "mean_token_accuracy": 0.2317226454615593,
      "num_tokens": 5834741.0,
      "step": 7320
    },
    {
      "entropy": 4.865066242218018,
      "epoch": 2.0709640938648572,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046342402329865054,
      "loss": 4.7192,
      "mean_token_accuracy": 0.24653877466917037,
      "num_tokens": 5838592.0,
      "step": 7325
    },
    {
      "entropy": 4.8440440654754635,
      "epoch": 2.0723777212326833,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046336778727408285,
      "loss": 4.7409,
      "mean_token_accuracy": 0.23850802332162857,
      "num_tokens": 5842745.0,
      "step": 7330
    },
    {
      "entropy": 4.9063512802124025,
      "epoch": 2.073791348600509,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004633115118818125,
      "loss": 4.8098,
      "mean_token_accuracy": 0.23030341118574144,
      "num_tokens": 5846611.0,
      "step": 7335
    },
    {
      "entropy": 4.941003227233887,
      "epoch": 2.075204975968335,
      "grad_norm": 1.5625,
      "learning_rate": 0.00046325519713360066,
      "loss": 4.7737,
      "mean_token_accuracy": 0.2425389662384987,
      "num_tokens": 5850982.0,
      "step": 7340
    },
    {
      "entropy": 4.93012056350708,
      "epoch": 2.0766186033361604,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046319884304121676,
      "loss": 4.7172,
      "mean_token_accuracy": 0.23295947015285492,
      "num_tokens": 5855034.0,
      "step": 7345
    },
    {
      "entropy": 5.014405727386475,
      "epoch": 2.0780322307039865,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004631424496164384,
      "loss": 4.6862,
      "mean_token_accuracy": 0.24196864068508148,
      "num_tokens": 5858912.0,
      "step": 7350
    },
    {
      "entropy": 5.004891967773437,
      "epoch": 2.0794458580718125,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046308601687105157,
      "loss": 4.7929,
      "mean_token_accuracy": 0.23566136956214906,
      "num_tokens": 5862721.0,
      "step": 7355
    },
    {
      "entropy": 4.970847082138062,
      "epoch": 2.080859485439638,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004630295448168502,
      "loss": 4.7326,
      "mean_token_accuracy": 0.23737827688455582,
      "num_tokens": 5866767.0,
      "step": 7360
    },
    {
      "entropy": 5.0594807147979735,
      "epoch": 2.082273112807464,
      "grad_norm": 1.578125,
      "learning_rate": 0.00046297303346563677,
      "loss": 4.8881,
      "mean_token_accuracy": 0.21784915775060654,
      "num_tokens": 5870733.0,
      "step": 7365
    },
    {
      "entropy": 4.969377756118774,
      "epoch": 2.0836867401752897,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00046291648282922176,
      "loss": 4.7203,
      "mean_token_accuracy": 0.23940921127796172,
      "num_tokens": 5874571.0,
      "step": 7370
    },
    {
      "entropy": 4.799388360977173,
      "epoch": 2.0851003675431157,
      "grad_norm": 1.625,
      "learning_rate": 0.0004628598929194239,
      "loss": 4.6419,
      "mean_token_accuracy": 0.2524953752756119,
      "num_tokens": 5878562.0,
      "step": 7375
    },
    {
      "entropy": 4.8628462791442875,
      "epoch": 2.0865139949109412,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046280326374807006,
      "loss": 4.6974,
      "mean_token_accuracy": 0.23603426963090895,
      "num_tokens": 5882601.0,
      "step": 7380
    },
    {
      "entropy": 4.843268489837646,
      "epoch": 2.0879276222787673,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004627465953269955,
      "loss": 4.7141,
      "mean_token_accuracy": 0.24187173545360566,
      "num_tokens": 5886647.0,
      "step": 7385
    },
    {
      "entropy": 4.9335462093353275,
      "epoch": 2.0893412496465933,
      "grad_norm": 1.75,
      "learning_rate": 0.0004626898876680434,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24601833075284957,
      "num_tokens": 5890985.0,
      "step": 7390
    },
    {
      "entropy": 4.983161735534668,
      "epoch": 2.090754877014419,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004626331407830655,
      "loss": 4.8841,
      "mean_token_accuracy": 0.22796189785003662,
      "num_tokens": 5894735.0,
      "step": 7395
    },
    {
      "entropy": 4.905942583084107,
      "epoch": 2.092168504382245,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004625763546839214,
      "loss": 4.6522,
      "mean_token_accuracy": 0.2427350953221321,
      "num_tokens": 5898955.0,
      "step": 7400
    },
    {
      "entropy": 4.868928050994873,
      "epoch": 2.093582131750071,
      "grad_norm": 1.84375,
      "learning_rate": 0.00046251952938247915,
      "loss": 4.6946,
      "mean_token_accuracy": 0.24402108341455458,
      "num_tokens": 5902875.0,
      "step": 7405
    },
    {
      "entropy": 4.877893733978271,
      "epoch": 2.0949957591178965,
      "grad_norm": 1.796875,
      "learning_rate": 0.00046246266489061486,
      "loss": 4.7641,
      "mean_token_accuracy": 0.2461784675717354,
      "num_tokens": 5906801.0,
      "step": 7410
    },
    {
      "entropy": 4.930834770202637,
      "epoch": 2.0964093864857225,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004624057612202128,
      "loss": 4.7806,
      "mean_token_accuracy": 0.23830197751522064,
      "num_tokens": 5910930.0,
      "step": 7415
    },
    {
      "entropy": 4.940243577957153,
      "epoch": 2.097823013853548,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046234881838316556,
      "loss": 4.7412,
      "mean_token_accuracy": 0.2399543359875679,
      "num_tokens": 5914963.0,
      "step": 7420
    },
    {
      "entropy": 4.8621608257293705,
      "epoch": 2.099236641221374,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046229183639137384,
      "loss": 4.8368,
      "mean_token_accuracy": 0.23376625776290894,
      "num_tokens": 5918919.0,
      "step": 7425
    },
    {
      "entropy": 4.969300127029419,
      "epoch": 2.1006502685891997,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046223481525674645,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23370543718338013,
      "num_tokens": 5922914.0,
      "step": 7430
    },
    {
      "entropy": 5.0128199577331545,
      "epoch": 2.1020638959570257,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046217775499120045,
      "loss": 4.8399,
      "mean_token_accuracy": 0.22617774903774263,
      "num_tokens": 5926737.0,
      "step": 7435
    },
    {
      "entropy": 4.830695867538452,
      "epoch": 2.1034775233248517,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00046212065560666125,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2541534170508385,
      "num_tokens": 5930979.0,
      "step": 7440
    },
    {
      "entropy": 5.050378847122192,
      "epoch": 2.1048911506926773,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000462063517115062,
      "loss": 5.006,
      "mean_token_accuracy": 0.21643176823854446,
      "num_tokens": 5935325.0,
      "step": 7445
    },
    {
      "entropy": 5.003580808639526,
      "epoch": 2.1063047780605033,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004620063395283445,
      "loss": 4.7351,
      "mean_token_accuracy": 0.23645782023668288,
      "num_tokens": 5939445.0,
      "step": 7450
    },
    {
      "entropy": 5.032155466079712,
      "epoch": 2.107718405428329,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00046194912285845845,
      "loss": 4.8299,
      "mean_token_accuracy": 0.23628872036933898,
      "num_tokens": 5943125.0,
      "step": 7455
    },
    {
      "entropy": 4.9045721054077145,
      "epoch": 2.109132032796155,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004618918671173618,
      "loss": 4.7561,
      "mean_token_accuracy": 0.23904254883527756,
      "num_tokens": 5947342.0,
      "step": 7460
    },
    {
      "entropy": 4.789917659759522,
      "epoch": 2.110545660163981,
      "grad_norm": 1.6875,
      "learning_rate": 0.00046183457231702065,
      "loss": 4.6517,
      "mean_token_accuracy": 0.24576762020587922,
      "num_tokens": 5951269.0,
      "step": 7465
    },
    {
      "entropy": 4.802541303634643,
      "epoch": 2.1119592875318065,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004617772384694092,
      "loss": 4.6386,
      "mean_token_accuracy": 0.25787283182144166,
      "num_tokens": 5955280.0,
      "step": 7470
    },
    {
      "entropy": 4.826529693603516,
      "epoch": 2.1133729148996325,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004617198655865099,
      "loss": 4.7206,
      "mean_token_accuracy": 0.2458101823925972,
      "num_tokens": 5959662.0,
      "step": 7475
    },
    {
      "entropy": 4.895381689071655,
      "epoch": 2.114786542267458,
      "grad_norm": 1.640625,
      "learning_rate": 0.00046166245368031337,
      "loss": 4.64,
      "mean_token_accuracy": 0.25078244805336,
      "num_tokens": 5963471.0,
      "step": 7480
    },
    {
      "entropy": 4.987657308578491,
      "epoch": 2.116200169635284,
      "grad_norm": 1.65625,
      "learning_rate": 0.00046160500276281826,
      "loss": 4.8568,
      "mean_token_accuracy": 0.23446590602397918,
      "num_tokens": 5967678.0,
      "step": 7485
    },
    {
      "entropy": 4.946350383758545,
      "epoch": 2.11761379700311,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004615475128460315,
      "loss": 4.8395,
      "mean_token_accuracy": 0.22628939747810364,
      "num_tokens": 5971900.0,
      "step": 7490
    },
    {
      "entropy": 5.03000841140747,
      "epoch": 2.1190274243709357,
      "grad_norm": 1.484375,
      "learning_rate": 0.00046148998394196825,
      "loss": 4.8084,
      "mean_token_accuracy": 0.23481720685958862,
      "num_tokens": 5975922.0,
      "step": 7495
    },
    {
      "entropy": 4.896151065826416,
      "epoch": 2.1204410517387617,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004614324160626514,
      "loss": 4.7024,
      "mean_token_accuracy": 0.2536986216902733,
      "num_tokens": 5979791.0,
      "step": 7500
    },
    {
      "entropy": 4.938299226760864,
      "epoch": 2.1218546791065873,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004613748092201126,
      "loss": 4.7213,
      "mean_token_accuracy": 0.2509587794542313,
      "num_tokens": 5983698.0,
      "step": 7505
    },
    {
      "entropy": 5.048752737045288,
      "epoch": 2.1232683064744133,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004613171634263911,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23095273673534394,
      "num_tokens": 5987670.0,
      "step": 7510
    },
    {
      "entropy": 5.015461254119873,
      "epoch": 2.1246819338422394,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004612594786935346,
      "loss": 4.7961,
      "mean_token_accuracy": 0.23643084466457367,
      "num_tokens": 5991592.0,
      "step": 7515
    },
    {
      "entropy": 4.812802124023437,
      "epoch": 2.126095561210065,
      "grad_norm": 1.90625,
      "learning_rate": 0.00046120175503359885,
      "loss": 4.6966,
      "mean_token_accuracy": 0.24089805781841278,
      "num_tokens": 5995743.0,
      "step": 7520
    },
    {
      "entropy": 4.90613865852356,
      "epoch": 2.127509188577891,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004611439924586477,
      "loss": 4.7165,
      "mean_token_accuracy": 0.24349485337734222,
      "num_tokens": 5999789.0,
      "step": 7525
    },
    {
      "entropy": 4.888037061691284,
      "epoch": 2.1289228159457165,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004610861909807533,
      "loss": 4.6335,
      "mean_token_accuracy": 0.25321493446826937,
      "num_tokens": 6003479.0,
      "step": 7530
    },
    {
      "entropy": 4.90562071800232,
      "epoch": 2.1303364433135425,
      "grad_norm": 1.609375,
      "learning_rate": 0.00046102835061199565,
      "loss": 4.7065,
      "mean_token_accuracy": 0.2391906648874283,
      "num_tokens": 6007396.0,
      "step": 7535
    },
    {
      "entropy": 4.912233686447143,
      "epoch": 2.1317500706813686,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00046097047136446314,
      "loss": 4.7241,
      "mean_token_accuracy": 0.228609898686409,
      "num_tokens": 6011164.0,
      "step": 7540
    },
    {
      "entropy": 4.938687181472778,
      "epoch": 2.133163698049194,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00046091255325025204,
      "loss": 4.744,
      "mean_token_accuracy": 0.2367461934685707,
      "num_tokens": 6015359.0,
      "step": 7545
    },
    {
      "entropy": 4.96913776397705,
      "epoch": 2.13457732541702,
      "grad_norm": 1.5625,
      "learning_rate": 0.000460854596281467,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2284056708216667,
      "num_tokens": 6019548.0,
      "step": 7550
    },
    {
      "entropy": 5.003738164901733,
      "epoch": 2.1359909527848457,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004607966004702206,
      "loss": 4.7905,
      "mean_token_accuracy": 0.23629674613475798,
      "num_tokens": 6023368.0,
      "step": 7555
    },
    {
      "entropy": 5.012578248977661,
      "epoch": 2.1374045801526718,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046073856582863365,
      "loss": 4.7262,
      "mean_token_accuracy": 0.23502033203840256,
      "num_tokens": 6026917.0,
      "step": 7560
    },
    {
      "entropy": 4.9917925834655765,
      "epoch": 2.138818207520498,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004606804923688351,
      "loss": 4.7816,
      "mean_token_accuracy": 0.23860219717025757,
      "num_tokens": 6031241.0,
      "step": 7565
    },
    {
      "entropy": 4.871499967575073,
      "epoch": 2.1402318348883234,
      "grad_norm": 1.59375,
      "learning_rate": 0.00046062238010296174,
      "loss": 4.7215,
      "mean_token_accuracy": 0.23876003175973892,
      "num_tokens": 6035295.0,
      "step": 7570
    },
    {
      "entropy": 4.8571858406066895,
      "epoch": 2.1416454622561494,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004605642290431588,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2295125052332878,
      "num_tokens": 6039344.0,
      "step": 7575
    },
    {
      "entropy": 4.958080577850342,
      "epoch": 2.143059089623975,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00046050603920157955,
      "loss": 4.7917,
      "mean_token_accuracy": 0.2456389307975769,
      "num_tokens": 6043369.0,
      "step": 7580
    },
    {
      "entropy": 4.981222867965698,
      "epoch": 2.144472716991801,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004604478105903852,
      "loss": 4.7165,
      "mean_token_accuracy": 0.24173346012830735,
      "num_tokens": 6047265.0,
      "step": 7585
    },
    {
      "entropy": 4.909101247787476,
      "epoch": 2.145886344359627,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00046038954322174527,
      "loss": 4.7977,
      "mean_token_accuracy": 0.24193304777145386,
      "num_tokens": 6051224.0,
      "step": 7590
    },
    {
      "entropy": 4.959563302993774,
      "epoch": 2.1472999717274526,
      "grad_norm": 1.671875,
      "learning_rate": 0.00046033123710783713,
      "loss": 4.7454,
      "mean_token_accuracy": 0.22938533425331115,
      "num_tokens": 6055139.0,
      "step": 7595
    },
    {
      "entropy": 4.910887813568115,
      "epoch": 2.1487135990952786,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046027289226084655,
      "loss": 4.7117,
      "mean_token_accuracy": 0.23569464534521103,
      "num_tokens": 6059264.0,
      "step": 7600
    },
    {
      "entropy": 4.968990230560303,
      "epoch": 2.150127226463104,
      "grad_norm": 1.625,
      "learning_rate": 0.00046021450869296717,
      "loss": 4.8636,
      "mean_token_accuracy": 0.22430213242769242,
      "num_tokens": 6062886.0,
      "step": 7605
    },
    {
      "entropy": 4.930490016937256,
      "epoch": 2.15154085383093,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00046015608641640076,
      "loss": 4.7892,
      "mean_token_accuracy": 0.2360200896859169,
      "num_tokens": 6066856.0,
      "step": 7610
    },
    {
      "entropy": 4.924584007263183,
      "epoch": 2.152954481198756,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046009762544335734,
      "loss": 4.7441,
      "mean_token_accuracy": 0.2334244355559349,
      "num_tokens": 6070702.0,
      "step": 7615
    },
    {
      "entropy": 4.971550321578979,
      "epoch": 2.154368108566582,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004600391257860547,
      "loss": 4.7991,
      "mean_token_accuracy": 0.24147284924983978,
      "num_tokens": 6074867.0,
      "step": 7620
    },
    {
      "entropy": 4.941831159591675,
      "epoch": 2.155781735934408,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004599805874567191,
      "loss": 4.6311,
      "mean_token_accuracy": 0.2485829010605812,
      "num_tokens": 6078979.0,
      "step": 7625
    },
    {
      "entropy": 4.805006790161133,
      "epoch": 2.1571953633022334,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045992201046758457,
      "loss": 4.5847,
      "mean_token_accuracy": 0.26612347215414045,
      "num_tokens": 6082750.0,
      "step": 7630
    },
    {
      "entropy": 4.8402608871459964,
      "epoch": 2.1586089906700594,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004598633948308935,
      "loss": 4.7327,
      "mean_token_accuracy": 0.24039724618196487,
      "num_tokens": 6087001.0,
      "step": 7635
    },
    {
      "entropy": 4.9053552627563475,
      "epoch": 2.1600226180378854,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00045980474055889585,
      "loss": 4.7183,
      "mean_token_accuracy": 0.241668176651001,
      "num_tokens": 6091030.0,
      "step": 7640
    },
    {
      "entropy": 4.887157964706421,
      "epoch": 2.161436245405711,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004597460476638504,
      "loss": 4.7263,
      "mean_token_accuracy": 0.2404913619160652,
      "num_tokens": 6095185.0,
      "step": 7645
    },
    {
      "entropy": 4.9058538436889645,
      "epoch": 2.162849872773537,
      "grad_norm": 1.75,
      "learning_rate": 0.00045968731615802334,
      "loss": 4.7533,
      "mean_token_accuracy": 0.24335065633058547,
      "num_tokens": 6099154.0,
      "step": 7650
    },
    {
      "entropy": 4.992227363586426,
      "epoch": 2.1642635001413626,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045962854605368926,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22699621319770813,
      "num_tokens": 6103084.0,
      "step": 7655
    },
    {
      "entropy": 4.842662715911866,
      "epoch": 2.1656771275091886,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004595697373631309,
      "loss": 4.6158,
      "mean_token_accuracy": 0.25087072402238847,
      "num_tokens": 6107163.0,
      "step": 7660
    },
    {
      "entropy": 4.8408431053161625,
      "epoch": 2.167090754877014,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004595108900986386,
      "loss": 4.7078,
      "mean_token_accuracy": 0.24570512771606445,
      "num_tokens": 6110978.0,
      "step": 7665
    },
    {
      "entropy": 4.807547664642334,
      "epoch": 2.16850438224484,
      "grad_norm": 1.65625,
      "learning_rate": 0.00045945200427251133,
      "loss": 4.7415,
      "mean_token_accuracy": 0.25210558623075485,
      "num_tokens": 6115109.0,
      "step": 7670
    },
    {
      "entropy": 4.92478837966919,
      "epoch": 2.1699180096126662,
      "grad_norm": 1.625,
      "learning_rate": 0.0004593930798970558,
      "loss": 4.8239,
      "mean_token_accuracy": 0.24337360709905625,
      "num_tokens": 6119068.0,
      "step": 7675
    },
    {
      "entropy": 4.950503301620484,
      "epoch": 2.171331636980492,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00045933411698458683,
      "loss": 4.7724,
      "mean_token_accuracy": 0.2432636708021164,
      "num_tokens": 6122773.0,
      "step": 7680
    },
    {
      "entropy": 4.914773607254029,
      "epoch": 2.172745264348318,
      "grad_norm": 1.625,
      "learning_rate": 0.00045927511554742724,
      "loss": 4.736,
      "mean_token_accuracy": 0.24321763813495637,
      "num_tokens": 6127009.0,
      "step": 7685
    },
    {
      "entropy": 4.886823558807373,
      "epoch": 2.174158891716144,
      "grad_norm": 1.703125,
      "learning_rate": 0.00045921607559790806,
      "loss": 4.736,
      "mean_token_accuracy": 0.2491409718990326,
      "num_tokens": 6131002.0,
      "step": 7690
    },
    {
      "entropy": 4.837510967254639,
      "epoch": 2.1755725190839694,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004591569971483682,
      "loss": 4.7391,
      "mean_token_accuracy": 0.23707708418369294,
      "num_tokens": 6134827.0,
      "step": 7695
    },
    {
      "entropy": 4.849724149703979,
      "epoch": 2.1769861464517954,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004590978802111548,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23529824912548064,
      "num_tokens": 6138992.0,
      "step": 7700
    },
    {
      "entropy": 4.840382719039917,
      "epoch": 2.178399773819621,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00045903872479862286,
      "loss": 4.7214,
      "mean_token_accuracy": 0.2460593655705452,
      "num_tokens": 6143206.0,
      "step": 7705
    },
    {
      "entropy": 4.837607097625733,
      "epoch": 2.179813401187447,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004589795309231354,
      "loss": 4.6431,
      "mean_token_accuracy": 0.2511198177933693,
      "num_tokens": 6146981.0,
      "step": 7710
    },
    {
      "entropy": 4.853651142120361,
      "epoch": 2.1812270285552726,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045892029859706377,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23793757408857347,
      "num_tokens": 6151010.0,
      "step": 7715
    },
    {
      "entropy": 4.897086811065674,
      "epoch": 2.1826406559230986,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045886102783278704,
      "loss": 4.7087,
      "mean_token_accuracy": 0.24578996896743774,
      "num_tokens": 6154765.0,
      "step": 7720
    },
    {
      "entropy": 5.0166059017181395,
      "epoch": 2.1840542832909247,
      "grad_norm": 1.515625,
      "learning_rate": 0.00045880171864269245,
      "loss": 4.7621,
      "mean_token_accuracy": 0.22592788487672805,
      "num_tokens": 6158493.0,
      "step": 7725
    },
    {
      "entropy": 4.918649053573608,
      "epoch": 2.1854679106587502,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004587423710391753,
      "loss": 4.7088,
      "mean_token_accuracy": 0.2454855114221573,
      "num_tokens": 6162659.0,
      "step": 7730
    },
    {
      "entropy": 4.921793842315674,
      "epoch": 2.1868815380265763,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004586829850346388,
      "loss": 4.7497,
      "mean_token_accuracy": 0.23512034565210344,
      "num_tokens": 6166761.0,
      "step": 7735
    },
    {
      "entropy": 4.900226163864136,
      "epoch": 2.188295165394402,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00045862356064149424,
      "loss": 4.7381,
      "mean_token_accuracy": 0.24324047565460205,
      "num_tokens": 6170719.0,
      "step": 7740
    },
    {
      "entropy": 4.9172369003295895,
      "epoch": 2.189708792762228,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000458564097872161,
      "loss": 4.8295,
      "mean_token_accuracy": 0.22815432250499726,
      "num_tokens": 6175013.0,
      "step": 7745
    },
    {
      "entropy": 4.84746150970459,
      "epoch": 2.191122420130054,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004585045967390665,
      "loss": 4.726,
      "mean_token_accuracy": 0.24184510558843614,
      "num_tokens": 6178897.0,
      "step": 7750
    },
    {
      "entropy": 4.864307260513305,
      "epoch": 2.1925360474978794,
      "grad_norm": 1.5625,
      "learning_rate": 0.000458445057254646,
      "loss": 4.7289,
      "mean_token_accuracy": 0.2511356472969055,
      "num_tokens": 6182685.0,
      "step": 7755
    },
    {
      "entropy": 4.864558553695678,
      "epoch": 2.1939496748657055,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004583854794313429,
      "loss": 4.6733,
      "mean_token_accuracy": 0.24554864317178726,
      "num_tokens": 6186277.0,
      "step": 7760
    },
    {
      "entropy": 4.968964338302612,
      "epoch": 2.195363302233531,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004583258632816087,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23463668972253798,
      "num_tokens": 6190409.0,
      "step": 7765
    },
    {
      "entropy": 4.976002168655396,
      "epoch": 2.196776929601357,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004582662088179027,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23525602966547013,
      "num_tokens": 6194582.0,
      "step": 7770
    },
    {
      "entropy": 4.9860350608825685,
      "epoch": 2.198190556969183,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045820651605269223,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22998220324516297,
      "num_tokens": 6198593.0,
      "step": 7775
    },
    {
      "entropy": 4.938497686386109,
      "epoch": 2.1996041843370087,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004581467849984529,
      "loss": 4.8247,
      "mean_token_accuracy": 0.23661979883909226,
      "num_tokens": 6202639.0,
      "step": 7780
    },
    {
      "entropy": 4.901692819595337,
      "epoch": 2.2010178117048347,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004580870156676681,
      "loss": 4.7907,
      "mean_token_accuracy": 0.24077287167310715,
      "num_tokens": 6206674.0,
      "step": 7785
    },
    {
      "entropy": 4.904650163650513,
      "epoch": 2.2024314390726603,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045802720807282916,
      "loss": 4.7471,
      "mean_token_accuracy": 0.24332133084535598,
      "num_tokens": 6210707.0,
      "step": 7790
    },
    {
      "entropy": 4.984134006500244,
      "epoch": 2.2038450664404863,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004579673622264356,
      "loss": 4.7959,
      "mean_token_accuracy": 0.23955468088388443,
      "num_tokens": 6214711.0,
      "step": 7795
    },
    {
      "entropy": 4.972776746749878,
      "epoch": 2.2052586938083123,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045790747814099464,
      "loss": 4.7601,
      "mean_token_accuracy": 0.23523485660552979,
      "num_tokens": 6219088.0,
      "step": 7800
    },
    {
      "entropy": 4.991784906387329,
      "epoch": 2.206672321176138,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045784755582902186,
      "loss": 4.742,
      "mean_token_accuracy": 0.23102276474237443,
      "num_tokens": 6223365.0,
      "step": 7805
    },
    {
      "entropy": 5.043491220474243,
      "epoch": 2.208085948543964,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045778759530304067,
      "loss": 4.9446,
      "mean_token_accuracy": 0.2276066079735756,
      "num_tokens": 6227469.0,
      "step": 7810
    },
    {
      "entropy": 4.9494105815887455,
      "epoch": 2.2094995759117895,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00045772759657558234,
      "loss": 4.6084,
      "mean_token_accuracy": 0.248030486702919,
      "num_tokens": 6231428.0,
      "step": 7815
    },
    {
      "entropy": 4.9713630199432375,
      "epoch": 2.2109132032796155,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045766755965918627,
      "loss": 4.789,
      "mean_token_accuracy": 0.24161217361688614,
      "num_tokens": 6235244.0,
      "step": 7820
    },
    {
      "entropy": 4.840947151184082,
      "epoch": 2.2123268306474415,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004576074845663999,
      "loss": 4.7367,
      "mean_token_accuracy": 0.2523552879691124,
      "num_tokens": 6239434.0,
      "step": 7825
    },
    {
      "entropy": 4.869684886932373,
      "epoch": 2.213740458015267,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004575473713097784,
      "loss": 4.7296,
      "mean_token_accuracy": 0.23898506462574004,
      "num_tokens": 6243480.0,
      "step": 7830
    },
    {
      "entropy": 4.909815979003906,
      "epoch": 2.215154085383093,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004574872199018851,
      "loss": 4.7441,
      "mean_token_accuracy": 0.2340654894709587,
      "num_tokens": 6247488.0,
      "step": 7835
    },
    {
      "entropy": 4.924313116073608,
      "epoch": 2.2165677127509187,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004574270303552913,
      "loss": 4.6454,
      "mean_token_accuracy": 0.25280262231826783,
      "num_tokens": 6251505.0,
      "step": 7840
    },
    {
      "entropy": 4.88576602935791,
      "epoch": 2.2179813401187447,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004573668026825763,
      "loss": 4.7276,
      "mean_token_accuracy": 0.2412440523505211,
      "num_tokens": 6255458.0,
      "step": 7845
    },
    {
      "entropy": 4.986235523223877,
      "epoch": 2.2193949674865707,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004573065368963273,
      "loss": 4.8999,
      "mean_token_accuracy": 0.22221068441867828,
      "num_tokens": 6259095.0,
      "step": 7850
    },
    {
      "entropy": 4.951801490783692,
      "epoch": 2.2208085948543963,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00045724623300913935,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22970976084470748,
      "num_tokens": 6263014.0,
      "step": 7855
    },
    {
      "entropy": 4.830119323730469,
      "epoch": 2.2222222222222223,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004571858910336157,
      "loss": 4.7025,
      "mean_token_accuracy": 0.24116699099540712,
      "num_tokens": 6266993.0,
      "step": 7860
    },
    {
      "entropy": 5.041571426391601,
      "epoch": 2.223635849590048,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004571255109823675,
      "loss": 4.8172,
      "mean_token_accuracy": 0.2278713181614876,
      "num_tokens": 6270957.0,
      "step": 7865
    },
    {
      "entropy": 4.907837820053101,
      "epoch": 2.225049476957874,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004570650928680136,
      "loss": 4.7058,
      "mean_token_accuracy": 0.24028218388557435,
      "num_tokens": 6274928.0,
      "step": 7870
    },
    {
      "entropy": 4.948648309707641,
      "epoch": 2.2264631043257,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004570046367031813,
      "loss": 4.8202,
      "mean_token_accuracy": 0.23769910037517547,
      "num_tokens": 6278955.0,
      "step": 7875
    },
    {
      "entropy": 4.8111656665802,
      "epoch": 2.2278767316935255,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004569441425005053,
      "loss": 4.6212,
      "mean_token_accuracy": 0.25501487851142884,
      "num_tokens": 6283021.0,
      "step": 7880
    },
    {
      "entropy": 4.985872983932495,
      "epoch": 2.2292903590613515,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00045688361027262865,
      "loss": 4.7909,
      "mean_token_accuracy": 0.23332791179418563,
      "num_tokens": 6287372.0,
      "step": 7885
    },
    {
      "entropy": 4.995982837677002,
      "epoch": 2.230703986429177,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004568230400322022,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2292018160223961,
      "num_tokens": 6291754.0,
      "step": 7890
    },
    {
      "entropy": 4.809036064147949,
      "epoch": 2.232117613797003,
      "grad_norm": 1.65625,
      "learning_rate": 0.00045676243179188473,
      "loss": 4.7188,
      "mean_token_accuracy": 0.24786302894353868,
      "num_tokens": 6295405.0,
      "step": 7895
    },
    {
      "entropy": 4.89035930633545,
      "epoch": 2.233531241164829,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004567017855643431,
      "loss": 4.8184,
      "mean_token_accuracy": 0.2395893454551697,
      "num_tokens": 6299409.0,
      "step": 7900
    },
    {
      "entropy": 4.878594446182251,
      "epoch": 2.2349448685326547,
      "grad_norm": 1.796875,
      "learning_rate": 0.0004566411013622518,
      "loss": 4.853,
      "mean_token_accuracy": 0.2327413409948349,
      "num_tokens": 6303678.0,
      "step": 7905
    },
    {
      "entropy": 4.982954692840576,
      "epoch": 2.2363584959004807,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045658037919829353,
      "loss": 4.8174,
      "mean_token_accuracy": 0.23659908324480056,
      "num_tokens": 6307541.0,
      "step": 7910
    },
    {
      "entropy": 4.893229389190674,
      "epoch": 2.2377721232683063,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004565196190851589,
      "loss": 4.6361,
      "mean_token_accuracy": 0.25792349576950074,
      "num_tokens": 6311630.0,
      "step": 7915
    },
    {
      "entropy": 4.964516115188599,
      "epoch": 2.2391857506361323,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004564588210355464,
      "loss": 4.7814,
      "mean_token_accuracy": 0.24428862780332566,
      "num_tokens": 6315521.0,
      "step": 7920
    },
    {
      "entropy": 4.996157598495484,
      "epoch": 2.2405993780039584,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00045639798506216236,
      "loss": 4.7856,
      "mean_token_accuracy": 0.23350564241409302,
      "num_tokens": 6319463.0,
      "step": 7925
    },
    {
      "entropy": 4.957121992111206,
      "epoch": 2.242013005371784,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045633711117772113,
      "loss": 4.7647,
      "mean_token_accuracy": 0.24551494121551515,
      "num_tokens": 6323818.0,
      "step": 7930
    },
    {
      "entropy": 4.868045949935913,
      "epoch": 2.24342663273961,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004562761993949451,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22543315291404725,
      "num_tokens": 6327339.0,
      "step": 7935
    },
    {
      "entropy": 4.91867299079895,
      "epoch": 2.2448402601074355,
      "grad_norm": 1.5,
      "learning_rate": 0.00045621524972656435,
      "loss": 4.5975,
      "mean_token_accuracy": 0.2491591081023216,
      "num_tokens": 6331549.0,
      "step": 7940
    },
    {
      "entropy": 4.738259744644165,
      "epoch": 2.2462538874752616,
      "grad_norm": 1.625,
      "learning_rate": 0.00045615426218531703,
      "loss": 4.6845,
      "mean_token_accuracy": 0.25024494528770447,
      "num_tokens": 6335606.0,
      "step": 7945
    },
    {
      "entropy": 4.879965114593506,
      "epoch": 2.247667514843087,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045609323678394914,
      "loss": 4.7518,
      "mean_token_accuracy": 0.24529256224632262,
      "num_tokens": 6339752.0,
      "step": 7950
    },
    {
      "entropy": 4.9291314601898195,
      "epoch": 2.249081142210913,
      "grad_norm": 1.625,
      "learning_rate": 0.0004560321735352146,
      "loss": 4.6589,
      "mean_token_accuracy": 0.25463327914476397,
      "num_tokens": 6343312.0,
      "step": 7955
    },
    {
      "entropy": 4.939134407043457,
      "epoch": 2.250494769578739,
      "grad_norm": 1.609375,
      "learning_rate": 0.00045597107245187523,
      "loss": 4.7595,
      "mean_token_accuracy": 0.23730673044919967,
      "num_tokens": 6347090.0,
      "step": 7960
    },
    {
      "entropy": 4.802748489379883,
      "epoch": 2.2519083969465647,
      "grad_norm": 1.625,
      "learning_rate": 0.0004559099335467008,
      "loss": 4.6318,
      "mean_token_accuracy": 0.2489236757159233,
      "num_tokens": 6350945.0,
      "step": 7965
    },
    {
      "entropy": 4.834428882598877,
      "epoch": 2.2533220243143908,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000455848756832469,
      "loss": 4.6829,
      "mean_token_accuracy": 0.24327393621206284,
      "num_tokens": 6354830.0,
      "step": 7970
    },
    {
      "entropy": 4.909896945953369,
      "epoch": 2.254735651682217,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00045578754232196536,
      "loss": 4.6815,
      "mean_token_accuracy": 0.24918185472488402,
      "num_tokens": 6358659.0,
      "step": 7975
    },
    {
      "entropy": 5.037668180465698,
      "epoch": 2.2561492790500424,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004557262900279834,
      "loss": 4.7896,
      "mean_token_accuracy": 0.22988000959157945,
      "num_tokens": 6362890.0,
      "step": 7980
    },
    {
      "entropy": 4.9425195217132565,
      "epoch": 2.2575629064178684,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00045566499996332427,
      "loss": 4.7804,
      "mean_token_accuracy": 0.23381834626197814,
      "num_tokens": 6367212.0,
      "step": 7985
    },
    {
      "entropy": 4.96451325416565,
      "epoch": 2.258976533785694,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045560367214079746,
      "loss": 4.6409,
      "mean_token_accuracy": 0.2497054159641266,
      "num_tokens": 6371058.0,
      "step": 7990
    },
    {
      "entropy": 4.99442195892334,
      "epoch": 2.26039016115352,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004555423065732199,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2164274737238884,
      "num_tokens": 6374931.0,
      "step": 7995
    },
    {
      "entropy": 4.970412158966065,
      "epoch": 2.2618037885213456,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004554809032734168,
      "loss": 4.84,
      "mean_token_accuracy": 0.22535085529088975,
      "num_tokens": 6379069.0,
      "step": 8000
    },
    {
      "epoch": 2.2618037885213456,
      "eval_entropy": 4.745825303183811,
      "eval_loss": 5.216935157775879,
      "eval_mean_token_accuracy": 0.22887288644899298,
      "eval_num_tokens": 6379069.0,
      "eval_runtime": 6.5252,
      "eval_samples_per_second": 1729.29,
      "eval_steps_per_second": 216.238,
      "step": 8000
    },
    {
      "entropy": 4.857929039001465,
      "epoch": 2.2632174158891716,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004554194622542209,
      "loss": 4.6737,
      "mean_token_accuracy": 0.25069428235292435,
      "num_tokens": 6383241.0,
      "step": 8005
    },
    {
      "entropy": 4.869005966186523,
      "epoch": 2.2646310432569976,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00045535798352847306,
      "loss": 4.6717,
      "mean_token_accuracy": 0.2507780626416206,
      "num_tokens": 6387303.0,
      "step": 8010
    },
    {
      "entropy": 4.98301854133606,
      "epoch": 2.266044670624823,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004552964671090219,
      "loss": 4.8172,
      "mean_token_accuracy": 0.23257110565900802,
      "num_tokens": 6391604.0,
      "step": 8015
    },
    {
      "entropy": 4.956060791015625,
      "epoch": 2.267458297992649,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00045523491300872423,
      "loss": 4.7704,
      "mean_token_accuracy": 0.22800056934356688,
      "num_tokens": 6395672.0,
      "step": 8020
    },
    {
      "entropy": 4.944966411590576,
      "epoch": 2.268871925360475,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00045517332124044414,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23206809014081956,
      "num_tokens": 6399962.0,
      "step": 8025
    },
    {
      "entropy": 4.91999659538269,
      "epoch": 2.270285552728301,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004551116918170541,
      "loss": 4.6793,
      "mean_token_accuracy": 0.25140474587678907,
      "num_tokens": 6404012.0,
      "step": 8030
    },
    {
      "entropy": 4.786691379547119,
      "epoch": 2.271699180096127,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00045505002475143433,
      "loss": 4.6981,
      "mean_token_accuracy": 0.24575625509023666,
      "num_tokens": 6407773.0,
      "step": 8035
    },
    {
      "entropy": 4.839750623703003,
      "epoch": 2.2731128074639524,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045498832005647267,
      "loss": 4.7828,
      "mean_token_accuracy": 0.23506145775318146,
      "num_tokens": 6411681.0,
      "step": 8040
    },
    {
      "entropy": 4.950056076049805,
      "epoch": 2.2745264348317784,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004549265777450652,
      "loss": 4.7874,
      "mean_token_accuracy": 0.22764758616685868,
      "num_tokens": 6415374.0,
      "step": 8045
    },
    {
      "entropy": 4.975764656066895,
      "epoch": 2.275940062199604,
      "grad_norm": 1.5625,
      "learning_rate": 0.00045486479783011553,
      "loss": 4.6986,
      "mean_token_accuracy": 0.24839378297328948,
      "num_tokens": 6419321.0,
      "step": 8050
    },
    {
      "entropy": 4.884446716308593,
      "epoch": 2.27735368956743,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00045480298032453544,
      "loss": 4.6701,
      "mean_token_accuracy": 0.24067514538764953,
      "num_tokens": 6423064.0,
      "step": 8055
    },
    {
      "entropy": 4.81669397354126,
      "epoch": 2.278767316935256,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004547411252412443,
      "loss": 4.8406,
      "mean_token_accuracy": 0.2297661319375038,
      "num_tokens": 6427130.0,
      "step": 8060
    },
    {
      "entropy": 4.820209074020386,
      "epoch": 2.2801809443030816,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004546792325931695,
      "loss": 4.6752,
      "mean_token_accuracy": 0.23365880846977233,
      "num_tokens": 6431401.0,
      "step": 8065
    },
    {
      "entropy": 4.827763891220092,
      "epoch": 2.2815945716709076,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004546173023932461,
      "loss": 4.7153,
      "mean_token_accuracy": 0.24070453345775605,
      "num_tokens": 6434993.0,
      "step": 8070
    },
    {
      "entropy": 4.877689504623413,
      "epoch": 2.283008199038733,
      "grad_norm": 1.640625,
      "learning_rate": 0.00045455533465441734,
      "loss": 4.7286,
      "mean_token_accuracy": 0.23689131885766984,
      "num_tokens": 6438819.0,
      "step": 8075
    },
    {
      "entropy": 4.918686580657959,
      "epoch": 2.284421826406559,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004544933293896339,
      "loss": 4.748,
      "mean_token_accuracy": 0.23657878786325454,
      "num_tokens": 6442849.0,
      "step": 8080
    },
    {
      "entropy": 4.930405950546264,
      "epoch": 2.2858354537743852,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00045443128661185457,
      "loss": 4.7744,
      "mean_token_accuracy": 0.2418770745396614,
      "num_tokens": 6446853.0,
      "step": 8085
    },
    {
      "entropy": 5.017850065231324,
      "epoch": 2.287249081142211,
      "grad_norm": 1.65625,
      "learning_rate": 0.00045436920633404597,
      "loss": 4.7948,
      "mean_token_accuracy": 0.24375513195991516,
      "num_tokens": 6451312.0,
      "step": 8090
    },
    {
      "entropy": 4.939690971374512,
      "epoch": 2.288662708510037,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004543070885691824,
      "loss": 4.808,
      "mean_token_accuracy": 0.2359660267829895,
      "num_tokens": 6455356.0,
      "step": 8095
    },
    {
      "entropy": 4.883692932128906,
      "epoch": 2.2900763358778624,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045424493333024607,
      "loss": 4.6695,
      "mean_token_accuracy": 0.253852504491806,
      "num_tokens": 6459102.0,
      "step": 8100
    },
    {
      "entropy": 4.848534774780274,
      "epoch": 2.2914899632456884,
      "grad_norm": 1.71875,
      "learning_rate": 0.00045418274063022715,
      "loss": 4.6495,
      "mean_token_accuracy": 0.2508886009454727,
      "num_tokens": 6462864.0,
      "step": 8105
    },
    {
      "entropy": 4.904107618331909,
      "epoch": 2.2929035906135145,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004541205104821235,
      "loss": 4.7223,
      "mean_token_accuracy": 0.23758445084095,
      "num_tokens": 6466870.0,
      "step": 8110
    },
    {
      "entropy": 4.96296968460083,
      "epoch": 2.29431721798134,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004540582428989408,
      "loss": 4.7594,
      "mean_token_accuracy": 0.2400449827313423,
      "num_tokens": 6470793.0,
      "step": 8115
    },
    {
      "entropy": 4.961249208450317,
      "epoch": 2.295730845349166,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00045399593789369246,
      "loss": 4.8394,
      "mean_token_accuracy": 0.23255587816238404,
      "num_tokens": 6474851.0,
      "step": 8120
    },
    {
      "entropy": 4.8786375522613525,
      "epoch": 2.2971444727169916,
      "grad_norm": 1.671875,
      "learning_rate": 0.00045393359547940017,
      "loss": 4.6885,
      "mean_token_accuracy": 0.2462855711579323,
      "num_tokens": 6478827.0,
      "step": 8125
    },
    {
      "entropy": 5.058320808410644,
      "epoch": 2.2985581000848176,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004538712156690927,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2231368288397789,
      "num_tokens": 6482771.0,
      "step": 8130
    },
    {
      "entropy": 4.975544881820679,
      "epoch": 2.2999717274526437,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00045380879847580743,
      "loss": 4.7322,
      "mean_token_accuracy": 0.24580084085464476,
      "num_tokens": 6486876.0,
      "step": 8135
    },
    {
      "entropy": 4.8193159103393555,
      "epoch": 2.3013853548204692,
      "grad_norm": 1.578125,
      "learning_rate": 0.000453746343912589,
      "loss": 4.6441,
      "mean_token_accuracy": 0.2498108074069023,
      "num_tokens": 6490741.0,
      "step": 8140
    },
    {
      "entropy": 4.838828802108765,
      "epoch": 2.3027989821882953,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045368385199249004,
      "loss": 4.7591,
      "mean_token_accuracy": 0.24246631413698197,
      "num_tokens": 6494387.0,
      "step": 8145
    },
    {
      "entropy": 4.897405195236206,
      "epoch": 2.304212609556121,
      "grad_norm": 1.484375,
      "learning_rate": 0.00045362132272857084,
      "loss": 4.7668,
      "mean_token_accuracy": 0.2433604344725609,
      "num_tokens": 6498384.0,
      "step": 8150
    },
    {
      "entropy": 4.956755352020264,
      "epoch": 2.305626236923947,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004535587561338998,
      "loss": 4.8537,
      "mean_token_accuracy": 0.2395369753241539,
      "num_tokens": 6502371.0,
      "step": 8155
    },
    {
      "entropy": 4.928529357910156,
      "epoch": 2.307039864291773,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004534961522215529,
      "loss": 4.7986,
      "mean_token_accuracy": 0.23546857982873917,
      "num_tokens": 6506660.0,
      "step": 8160
    },
    {
      "entropy": 4.9114875316619875,
      "epoch": 2.3084534916595985,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000453433511004614,
      "loss": 4.7426,
      "mean_token_accuracy": 0.24340589046478273,
      "num_tokens": 6510691.0,
      "step": 8165
    },
    {
      "entropy": 5.057585191726685,
      "epoch": 2.3098671190274245,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004533708324961747,
      "loss": 4.8369,
      "mean_token_accuracy": 0.22615054100751877,
      "num_tokens": 6514525.0,
      "step": 8170
    },
    {
      "entropy": 4.977590465545655,
      "epoch": 2.31128074639525,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045330811670933434,
      "loss": 4.7117,
      "mean_token_accuracy": 0.24224456995725632,
      "num_tokens": 6518532.0,
      "step": 8175
    },
    {
      "entropy": 4.789402103424072,
      "epoch": 2.312694373763076,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00045324536365720034,
      "loss": 4.6815,
      "mean_token_accuracy": 0.2452653869986534,
      "num_tokens": 6522600.0,
      "step": 8180
    },
    {
      "entropy": 4.8989004611969,
      "epoch": 2.3141080011309016,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004531825733528875,
      "loss": 4.8338,
      "mean_token_accuracy": 0.23871913701295852,
      "num_tokens": 6527131.0,
      "step": 8185
    },
    {
      "entropy": 4.970239210128784,
      "epoch": 2.3155216284987277,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004531197458095187,
      "loss": 4.8607,
      "mean_token_accuracy": 0.22604157030582428,
      "num_tokens": 6531400.0,
      "step": 8190
    },
    {
      "entropy": 4.910205316543579,
      "epoch": 2.3169352558665537,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004530568810402245,
      "loss": 4.6367,
      "mean_token_accuracy": 0.2566689506173134,
      "num_tokens": 6535298.0,
      "step": 8195
    },
    {
      "entropy": 4.944407606124878,
      "epoch": 2.3183488832343793,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045299397905814316,
      "loss": 4.7674,
      "mean_token_accuracy": 0.23807455748319625,
      "num_tokens": 6539272.0,
      "step": 8200
    },
    {
      "entropy": 4.914580249786377,
      "epoch": 2.3197625106022053,
      "grad_norm": 1.609375,
      "learning_rate": 0.00045293103987642096,
      "loss": 4.7241,
      "mean_token_accuracy": 0.23718222677707673,
      "num_tokens": 6543107.0,
      "step": 8205
    },
    {
      "entropy": 5.100191974639893,
      "epoch": 2.3211761379700313,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004528680635082117,
      "loss": 4.8226,
      "mean_token_accuracy": 0.23096920251846315,
      "num_tokens": 6547096.0,
      "step": 8210
    },
    {
      "entropy": 4.945192813873291,
      "epoch": 2.322589765337857,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00045280504996667704,
      "loss": 4.666,
      "mean_token_accuracy": 0.23566138595342637,
      "num_tokens": 6550863.0,
      "step": 8215
    },
    {
      "entropy": 4.907277297973633,
      "epoch": 2.324003392705683,
      "grad_norm": 1.71875,
      "learning_rate": 0.00045274199926498645,
      "loss": 4.7642,
      "mean_token_accuracy": 0.23366111367940903,
      "num_tokens": 6554403.0,
      "step": 8220
    },
    {
      "entropy": 4.806126499176026,
      "epoch": 2.3254170200735085,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004526789114163172,
      "loss": 4.6765,
      "mean_token_accuracy": 0.2501532584428787,
      "num_tokens": 6558542.0,
      "step": 8225
    },
    {
      "entropy": 4.975665283203125,
      "epoch": 2.3268306474413345,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004526157864338541,
      "loss": 4.6565,
      "mean_token_accuracy": 0.2418670043349266,
      "num_tokens": 6562471.0,
      "step": 8230
    },
    {
      "entropy": 4.893019437789917,
      "epoch": 2.32824427480916,
      "grad_norm": 1.625,
      "learning_rate": 0.00045255262433079003,
      "loss": 4.7559,
      "mean_token_accuracy": 0.2380644753575325,
      "num_tokens": 6566459.0,
      "step": 8235
    },
    {
      "entropy": 4.867609262466431,
      "epoch": 2.329657902176986,
      "grad_norm": 1.6875,
      "learning_rate": 0.00045248942512032543,
      "loss": 4.7015,
      "mean_token_accuracy": 0.2457309067249298,
      "num_tokens": 6570380.0,
      "step": 8240
    },
    {
      "entropy": 4.8301746368408205,
      "epoch": 2.331071529544812,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045242618881566847,
      "loss": 4.7184,
      "mean_token_accuracy": 0.24481467753648758,
      "num_tokens": 6574142.0,
      "step": 8245
    },
    {
      "entropy": 4.904778528213501,
      "epoch": 2.3324851569126377,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004523629154300353,
      "loss": 4.6749,
      "mean_token_accuracy": 0.24547937959432603,
      "num_tokens": 6578286.0,
      "step": 8250
    },
    {
      "entropy": 4.914330530166626,
      "epoch": 2.3338987842804637,
      "grad_norm": 1.75,
      "learning_rate": 0.00045229960497664964,
      "loss": 4.6668,
      "mean_token_accuracy": 0.24352151453495025,
      "num_tokens": 6582045.0,
      "step": 8255
    },
    {
      "entropy": 4.954193830490112,
      "epoch": 2.3353124116482897,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004522362574687428,
      "loss": 4.7678,
      "mean_token_accuracy": 0.2400249019265175,
      "num_tokens": 6586246.0,
      "step": 8260
    },
    {
      "entropy": 5.000493907928467,
      "epoch": 2.3367260390161153,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004521728729195542,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23822900503873826,
      "num_tokens": 6590274.0,
      "step": 8265
    },
    {
      "entropy": 4.95618200302124,
      "epoch": 2.3381396663839413,
      "grad_norm": 1.6875,
      "learning_rate": 0.00045210945134233077,
      "loss": 4.7861,
      "mean_token_accuracy": 0.24167825579643248,
      "num_tokens": 6594327.0,
      "step": 8270
    },
    {
      "entropy": 4.974596548080444,
      "epoch": 2.339553293751767,
      "grad_norm": 1.59375,
      "learning_rate": 0.00045204599275032724,
      "loss": 4.832,
      "mean_token_accuracy": 0.22525953203439714,
      "num_tokens": 6597873.0,
      "step": 8275
    },
    {
      "entropy": 4.933299493789673,
      "epoch": 2.340966921119593,
      "grad_norm": 1.46875,
      "learning_rate": 0.000451982497156806,
      "loss": 4.6977,
      "mean_token_accuracy": 0.25642334669828415,
      "num_tokens": 6601919.0,
      "step": 8280
    },
    {
      "entropy": 4.967727851867676,
      "epoch": 2.3423805484874185,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004519189645750373,
      "loss": 4.8183,
      "mean_token_accuracy": 0.2429963916540146,
      "num_tokens": 6606008.0,
      "step": 8285
    },
    {
      "entropy": 4.943296909332275,
      "epoch": 2.3437941758552445,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000451855395018299,
      "loss": 4.7908,
      "mean_token_accuracy": 0.24030357897281646,
      "num_tokens": 6609870.0,
      "step": 8290
    },
    {
      "entropy": 4.892520761489868,
      "epoch": 2.3452078032230705,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004517917884998768,
      "loss": 4.6726,
      "mean_token_accuracy": 0.25001850724220276,
      "num_tokens": 6613678.0,
      "step": 8295
    },
    {
      "entropy": 4.788475751876831,
      "epoch": 2.346621430590896,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004517281450330642,
      "loss": 4.7565,
      "mean_token_accuracy": 0.23538568168878554,
      "num_tokens": 6617648.0,
      "step": 8300
    },
    {
      "entropy": 4.833195352554322,
      "epoch": 2.348035057958722,
      "grad_norm": 1.46875,
      "learning_rate": 0.00045166446463116195,
      "loss": 4.678,
      "mean_token_accuracy": 0.2520333990454674,
      "num_tokens": 6621736.0,
      "step": 8305
    },
    {
      "entropy": 4.8603565216064455,
      "epoch": 2.349448685326548,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004516007473074791,
      "loss": 4.705,
      "mean_token_accuracy": 0.2461240291595459,
      "num_tokens": 6625542.0,
      "step": 8310
    },
    {
      "entropy": 4.935512447357178,
      "epoch": 2.3508623126943737,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004515369930753322,
      "loss": 4.8131,
      "mean_token_accuracy": 0.22625891864299774,
      "num_tokens": 6629790.0,
      "step": 8315
    },
    {
      "entropy": 4.906849479675293,
      "epoch": 2.3522759400621998,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004514732019480453,
      "loss": 4.7288,
      "mean_token_accuracy": 0.2453140765428543,
      "num_tokens": 6633758.0,
      "step": 8320
    },
    {
      "entropy": 4.910966491699218,
      "epoch": 2.3536895674300253,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004514093739389506,
      "loss": 4.7536,
      "mean_token_accuracy": 0.254069447517395,
      "num_tokens": 6638069.0,
      "step": 8325
    },
    {
      "entropy": 4.837586069107056,
      "epoch": 2.3551031947978514,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004513455090613876,
      "loss": 4.7762,
      "mean_token_accuracy": 0.24584570676088333,
      "num_tokens": 6641811.0,
      "step": 8330
    },
    {
      "entropy": 4.753280782699585,
      "epoch": 2.356516822165677,
      "grad_norm": 1.5,
      "learning_rate": 0.0004512816073287036,
      "loss": 4.7239,
      "mean_token_accuracy": 0.2515147224068642,
      "num_tokens": 6645839.0,
      "step": 8335
    },
    {
      "entropy": 4.992108917236328,
      "epoch": 2.357930449533503,
      "grad_norm": 1.5,
      "learning_rate": 0.00045121766875425386,
      "loss": 4.8214,
      "mean_token_accuracy": 0.23518812358379365,
      "num_tokens": 6649878.0,
      "step": 8340
    },
    {
      "entropy": 4.876882886886596,
      "epoch": 2.359344076901329,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00045115369335140096,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2406069815158844,
      "num_tokens": 6653775.0,
      "step": 8345
    },
    {
      "entropy": 4.88419394493103,
      "epoch": 2.3607577042691545,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004510896811335155,
      "loss": 4.7958,
      "mean_token_accuracy": 0.24225910902023315,
      "num_tokens": 6657905.0,
      "step": 8350
    },
    {
      "entropy": 4.97317419052124,
      "epoch": 2.3621713316369806,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045102563211397556,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2353278562426567,
      "num_tokens": 6662067.0,
      "step": 8355
    },
    {
      "entropy": 4.8347900867462155,
      "epoch": 2.363584959004806,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004509615463061671,
      "loss": 4.711,
      "mean_token_accuracy": 0.23949723690748215,
      "num_tokens": 6665774.0,
      "step": 8360
    },
    {
      "entropy": 4.835385513305664,
      "epoch": 2.364998586372632,
      "grad_norm": 1.546875,
      "learning_rate": 0.00045089742372348345,
      "loss": 4.667,
      "mean_token_accuracy": 0.2482277810573578,
      "num_tokens": 6669566.0,
      "step": 8365
    },
    {
      "entropy": 4.898536491394043,
      "epoch": 2.366412213740458,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004508332643793261,
      "loss": 4.8433,
      "mean_token_accuracy": 0.23185713440179825,
      "num_tokens": 6673505.0,
      "step": 8370
    },
    {
      "entropy": 4.969058465957642,
      "epoch": 2.3678258411082838,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004507690682871037,
      "loss": 4.7679,
      "mean_token_accuracy": 0.2410995200276375,
      "num_tokens": 6677115.0,
      "step": 8375
    },
    {
      "entropy": 4.871894264221192,
      "epoch": 2.3692394684761098,
      "grad_norm": 1.625,
      "learning_rate": 0.000450704835460233,
      "loss": 4.769,
      "mean_token_accuracy": 0.2436786636710167,
      "num_tokens": 6680763.0,
      "step": 8380
    },
    {
      "entropy": 4.993407201766968,
      "epoch": 2.3706530958439354,
      "grad_norm": 1.578125,
      "learning_rate": 0.00045064056591213826,
      "loss": 4.844,
      "mean_token_accuracy": 0.22968405932188035,
      "num_tokens": 6685082.0,
      "step": 8385
    },
    {
      "entropy": 4.881499338150024,
      "epoch": 2.3720667232117614,
      "grad_norm": 1.703125,
      "learning_rate": 0.00045057625965625126,
      "loss": 4.7703,
      "mean_token_accuracy": 0.23090616911649703,
      "num_tokens": 6688925.0,
      "step": 8390
    },
    {
      "entropy": 4.760347890853882,
      "epoch": 2.3734803505795874,
      "grad_norm": 1.5,
      "learning_rate": 0.00045051191670601184,
      "loss": 4.6051,
      "mean_token_accuracy": 0.24676554501056672,
      "num_tokens": 6692832.0,
      "step": 8395
    },
    {
      "entropy": 4.831993198394775,
      "epoch": 2.374893977947413,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004504475370748671,
      "loss": 4.7484,
      "mean_token_accuracy": 0.23567487746477128,
      "num_tokens": 6696758.0,
      "step": 8400
    },
    {
      "entropy": 4.825489187240601,
      "epoch": 2.376307605315239,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00045038312077627207,
      "loss": 4.7376,
      "mean_token_accuracy": 0.2414853975176811,
      "num_tokens": 6700968.0,
      "step": 8405
    },
    {
      "entropy": 4.897916555404663,
      "epoch": 2.3777212326830646,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004503186678236893,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23469944596290587,
      "num_tokens": 6705137.0,
      "step": 8410
    },
    {
      "entropy": 4.974789047241211,
      "epoch": 2.3791348600508906,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00045025417823058906,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2210345223546028,
      "num_tokens": 6709051.0,
      "step": 8415
    },
    {
      "entropy": 4.771886539459229,
      "epoch": 2.3805484874187166,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004501896520104494,
      "loss": 4.588,
      "mean_token_accuracy": 0.26057340353727343,
      "num_tokens": 6713029.0,
      "step": 8420
    },
    {
      "entropy": 4.820526504516602,
      "epoch": 2.381962114786542,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004501250891767558,
      "loss": 4.8164,
      "mean_token_accuracy": 0.24143949300050735,
      "num_tokens": 6717420.0,
      "step": 8425
    },
    {
      "entropy": 4.828280019760132,
      "epoch": 2.383375742154368,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004500604897430015,
      "loss": 4.6375,
      "mean_token_accuracy": 0.24870380461215974,
      "num_tokens": 6721458.0,
      "step": 8430
    },
    {
      "entropy": 4.780137443542481,
      "epoch": 2.384789369522194,
      "grad_norm": 1.625,
      "learning_rate": 0.0004499958537226873,
      "loss": 4.688,
      "mean_token_accuracy": 0.2500035151839256,
      "num_tokens": 6725584.0,
      "step": 8435
    },
    {
      "entropy": 4.801298904418945,
      "epoch": 2.38620299689002,
      "grad_norm": 1.6875,
      "learning_rate": 0.00044993118112932187,
      "loss": 4.7785,
      "mean_token_accuracy": 0.24658534824848174,
      "num_tokens": 6730125.0,
      "step": 8440
    },
    {
      "entropy": 4.814269351959228,
      "epoch": 2.387616624257846,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004498664719764213,
      "loss": 4.6712,
      "mean_token_accuracy": 0.24236520677804946,
      "num_tokens": 6734182.0,
      "step": 8445
    },
    {
      "entropy": 4.808274936676026,
      "epoch": 2.3890302516256714,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00044980172627750954,
      "loss": 4.7305,
      "mean_token_accuracy": 0.24563776850700378,
      "num_tokens": 6738163.0,
      "step": 8450
    },
    {
      "entropy": 4.818344640731811,
      "epoch": 2.3904438789934974,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00044973694404611776,
      "loss": 4.7045,
      "mean_token_accuracy": 0.24322300255298615,
      "num_tokens": 6741861.0,
      "step": 8455
    },
    {
      "entropy": 4.8598674774169925,
      "epoch": 2.391857506361323,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044967212529578533,
      "loss": 4.8468,
      "mean_token_accuracy": 0.2410425513982773,
      "num_tokens": 6745776.0,
      "step": 8460
    },
    {
      "entropy": 4.882827472686768,
      "epoch": 2.393271133729149,
      "grad_norm": 1.625,
      "learning_rate": 0.0004496072700400588,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23479101210832595,
      "num_tokens": 6749775.0,
      "step": 8465
    },
    {
      "entropy": 4.94600658416748,
      "epoch": 2.3946847610969746,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00044954237829249264,
      "loss": 4.7949,
      "mean_token_accuracy": 0.23425700813531875,
      "num_tokens": 6753535.0,
      "step": 8470
    },
    {
      "entropy": 4.861195039749146,
      "epoch": 2.3960983884648006,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004494774500666487,
      "loss": 4.7979,
      "mean_token_accuracy": 0.21959861367940903,
      "num_tokens": 6757733.0,
      "step": 8475
    },
    {
      "entropy": 4.829150533676147,
      "epoch": 2.3975120158326266,
      "grad_norm": 1.5,
      "learning_rate": 0.00044941248537609663,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23843462765216827,
      "num_tokens": 6762026.0,
      "step": 8480
    },
    {
      "entropy": 4.908278226852417,
      "epoch": 2.398925643200452,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004493474842344137,
      "loss": 4.7622,
      "mean_token_accuracy": 0.23106739521026612,
      "num_tokens": 6766095.0,
      "step": 8485
    },
    {
      "entropy": 4.847249698638916,
      "epoch": 2.4003392705682782,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044928244665518464,
      "loss": 4.7164,
      "mean_token_accuracy": 0.23815023452043532,
      "num_tokens": 6770185.0,
      "step": 8490
    },
    {
      "entropy": 4.861820650100708,
      "epoch": 2.4017528979361042,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00044921737265200194,
      "loss": 4.7682,
      "mean_token_accuracy": 0.23911094069480895,
      "num_tokens": 6773735.0,
      "step": 8495
    },
    {
      "entropy": 4.883049011230469,
      "epoch": 2.40316652530393,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004491522622384658,
      "loss": 4.8079,
      "mean_token_accuracy": 0.24237620383501052,
      "num_tokens": 6777546.0,
      "step": 8500
    },
    {
      "entropy": 4.7721583366394045,
      "epoch": 2.404580152671756,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004490871154281836,
      "loss": 4.727,
      "mean_token_accuracy": 0.24927582889795302,
      "num_tokens": 6781181.0,
      "step": 8505
    },
    {
      "entropy": 4.8119298934936525,
      "epoch": 2.4059937800395814,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044902193223477096,
      "loss": 4.6402,
      "mean_token_accuracy": 0.25188308507204055,
      "num_tokens": 6785123.0,
      "step": 8510
    },
    {
      "entropy": 4.750468778610229,
      "epoch": 2.4074074074074074,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004489567126718505,
      "loss": 4.7019,
      "mean_token_accuracy": 0.24906525760889053,
      "num_tokens": 6789024.0,
      "step": 8515
    },
    {
      "entropy": 4.832190036773682,
      "epoch": 2.408821034775233,
      "grad_norm": 1.515625,
      "learning_rate": 0.00044889145675305284,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24265960305929185,
      "num_tokens": 6792893.0,
      "step": 8520
    },
    {
      "entropy": 4.982015228271484,
      "epoch": 2.410234662143059,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004488261644920161,
      "loss": 4.8601,
      "mean_token_accuracy": 0.23162841796875,
      "num_tokens": 6796722.0,
      "step": 8525
    },
    {
      "entropy": 4.918169975280762,
      "epoch": 2.411648289510885,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00044876083590238575,
      "loss": 4.7573,
      "mean_token_accuracy": 0.2423504039645195,
      "num_tokens": 6800662.0,
      "step": 8530
    },
    {
      "entropy": 5.025771045684815,
      "epoch": 2.4130619168787106,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004486954709978153,
      "loss": 4.7953,
      "mean_token_accuracy": 0.22864664644002913,
      "num_tokens": 6804619.0,
      "step": 8535
    },
    {
      "entropy": 4.876535081863404,
      "epoch": 2.4144755442465367,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00044863006979196555,
      "loss": 4.7827,
      "mean_token_accuracy": 0.2340967759490013,
      "num_tokens": 6808410.0,
      "step": 8540
    },
    {
      "entropy": 4.849480295181275,
      "epoch": 2.4158891716143627,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004485646322985048,
      "loss": 4.7337,
      "mean_token_accuracy": 0.23733790814876557,
      "num_tokens": 6812304.0,
      "step": 8545
    },
    {
      "entropy": 4.756396961212158,
      "epoch": 2.4173027989821882,
      "grad_norm": 1.5,
      "learning_rate": 0.00044849915853110924,
      "loss": 4.7088,
      "mean_token_accuracy": 0.24675475358963012,
      "num_tokens": 6816446.0,
      "step": 8550
    },
    {
      "entropy": 4.825299787521362,
      "epoch": 2.4187164263500143,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00044843364850346245,
      "loss": 4.7909,
      "mean_token_accuracy": 0.23851365596055984,
      "num_tokens": 6820714.0,
      "step": 8555
    },
    {
      "entropy": 4.870847606658936,
      "epoch": 2.42013005371784,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044836810222925554,
      "loss": 4.7464,
      "mean_token_accuracy": 0.24335861653089524,
      "num_tokens": 6824647.0,
      "step": 8560
    },
    {
      "entropy": 4.755172395706177,
      "epoch": 2.421543681085666,
      "grad_norm": 1.515625,
      "learning_rate": 0.00044830251972218746,
      "loss": 4.5909,
      "mean_token_accuracy": 0.2565336927771568,
      "num_tokens": 6828110.0,
      "step": 8565
    },
    {
      "entropy": 4.865521574020386,
      "epoch": 2.4229573084534914,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004482369009959644,
      "loss": 4.8201,
      "mean_token_accuracy": 0.232114377617836,
      "num_tokens": 6832311.0,
      "step": 8570
    },
    {
      "entropy": 4.889907121658325,
      "epoch": 2.4243709358213175,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044817124606430015,
      "loss": 4.7499,
      "mean_token_accuracy": 0.23940238058567048,
      "num_tokens": 6836123.0,
      "step": 8575
    },
    {
      "entropy": 5.053268814086914,
      "epoch": 2.4257845631891435,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004481055549409165,
      "loss": 4.893,
      "mean_token_accuracy": 0.22042559385299682,
      "num_tokens": 6840214.0,
      "step": 8580
    },
    {
      "entropy": 4.900110960006714,
      "epoch": 2.427198190556969,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004480398276395422,
      "loss": 4.6713,
      "mean_token_accuracy": 0.24834479093551637,
      "num_tokens": 6844321.0,
      "step": 8585
    },
    {
      "entropy": 4.9106738567352295,
      "epoch": 2.428611817924795,
      "grad_norm": 1.484375,
      "learning_rate": 0.00044797406417391403,
      "loss": 4.6955,
      "mean_token_accuracy": 0.24145007282495498,
      "num_tokens": 6848365.0,
      "step": 8590
    },
    {
      "entropy": 4.898888015747071,
      "epoch": 2.430025445292621,
      "grad_norm": 1.515625,
      "learning_rate": 0.00044790826455777603,
      "loss": 4.7057,
      "mean_token_accuracy": 0.2386421427130699,
      "num_tokens": 6852179.0,
      "step": 8595
    },
    {
      "entropy": 4.891633415222168,
      "epoch": 2.4314390726604467,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044784242880488006,
      "loss": 4.7802,
      "mean_token_accuracy": 0.23355532437562943,
      "num_tokens": 6856321.0,
      "step": 8600
    },
    {
      "entropy": 4.830768013000489,
      "epoch": 2.4328527000282727,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004477765569289853,
      "loss": 4.836,
      "mean_token_accuracy": 0.2372908979654312,
      "num_tokens": 6860316.0,
      "step": 8605
    },
    {
      "entropy": 4.7828680038452145,
      "epoch": 2.4342663273960983,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044771064894385854,
      "loss": 4.6927,
      "mean_token_accuracy": 0.25352813452482226,
      "num_tokens": 6864119.0,
      "step": 8610
    },
    {
      "entropy": 4.811425971984863,
      "epoch": 2.4356799547639243,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00044764470486327406,
      "loss": 4.7188,
      "mean_token_accuracy": 0.2513998940587044,
      "num_tokens": 6867939.0,
      "step": 8615
    },
    {
      "entropy": 4.873059034347534,
      "epoch": 2.43709358213175,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000447578724701014,
      "loss": 4.7414,
      "mean_token_accuracy": 0.2547308936715126,
      "num_tokens": 6871561.0,
      "step": 8620
    },
    {
      "entropy": 4.916575765609741,
      "epoch": 2.438507209499576,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004475127084708676,
      "loss": 4.798,
      "mean_token_accuracy": 0.24338409900665284,
      "num_tokens": 6875399.0,
      "step": 8625
    },
    {
      "entropy": 4.850693321228027,
      "epoch": 2.439920836867402,
      "grad_norm": 1.578125,
      "learning_rate": 0.000447446656186632,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23682324290275575,
      "num_tokens": 6879298.0,
      "step": 8630
    },
    {
      "entropy": 4.850175666809082,
      "epoch": 2.4413344642352275,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004473805678621115,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23870264291763305,
      "num_tokens": 6883204.0,
      "step": 8635
    },
    {
      "entropy": 4.814730167388916,
      "epoch": 2.4427480916030535,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00044731444351111833,
      "loss": 4.7376,
      "mean_token_accuracy": 0.23845971524715423,
      "num_tokens": 6887469.0,
      "step": 8640
    },
    {
      "entropy": 4.708613538742066,
      "epoch": 2.444161718970879,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00044724828314747213,
      "loss": 4.6559,
      "mean_token_accuracy": 0.24688573479652404,
      "num_tokens": 6891484.0,
      "step": 8645
    },
    {
      "entropy": 4.805187082290649,
      "epoch": 2.445575346338705,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004471820867849998,
      "loss": 4.7901,
      "mean_token_accuracy": 0.2322575107216835,
      "num_tokens": 6895300.0,
      "step": 8650
    },
    {
      "entropy": 4.81672477722168,
      "epoch": 2.446988973706531,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004471158544375361,
      "loss": 4.7691,
      "mean_token_accuracy": 0.24080479592084886,
      "num_tokens": 6899533.0,
      "step": 8655
    },
    {
      "entropy": 4.86491551399231,
      "epoch": 2.4484026010743567,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004470495861189232,
      "loss": 4.8586,
      "mean_token_accuracy": 0.2339324966073036,
      "num_tokens": 6903364.0,
      "step": 8660
    },
    {
      "entropy": 4.866993761062622,
      "epoch": 2.4498162284421827,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004469832818430106,
      "loss": 4.6917,
      "mean_token_accuracy": 0.24559709876775743,
      "num_tokens": 6907500.0,
      "step": 8665
    },
    {
      "entropy": 4.830137777328491,
      "epoch": 2.4512298558100083,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004469169416236557,
      "loss": 4.809,
      "mean_token_accuracy": 0.23158814907073974,
      "num_tokens": 6911372.0,
      "step": 8670
    },
    {
      "entropy": 4.834042882919311,
      "epoch": 2.4526434831778343,
      "grad_norm": 2.03125,
      "learning_rate": 0.0004468505654747231,
      "loss": 4.7524,
      "mean_token_accuracy": 0.2475290596485138,
      "num_tokens": 6915439.0,
      "step": 8675
    },
    {
      "entropy": 4.926545763015747,
      "epoch": 2.4540571105456603,
      "grad_norm": 1.703125,
      "learning_rate": 0.00044678415341008497,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23481882214546204,
      "num_tokens": 6919654.0,
      "step": 8680
    },
    {
      "entropy": 4.878811693191528,
      "epoch": 2.455470737913486,
      "grad_norm": 1.609375,
      "learning_rate": 0.00044671770544362107,
      "loss": 4.744,
      "mean_token_accuracy": 0.24030105173587799,
      "num_tokens": 6923505.0,
      "step": 8685
    },
    {
      "entropy": 4.8963844776153564,
      "epoch": 2.456884365281312,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004466512215892186,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23588506877422333,
      "num_tokens": 6927436.0,
      "step": 8690
    },
    {
      "entropy": 4.823080158233642,
      "epoch": 2.4582979926491375,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00044658470186077233,
      "loss": 4.7886,
      "mean_token_accuracy": 0.24492516368627548,
      "num_tokens": 6931680.0,
      "step": 8695
    },
    {
      "entropy": 4.83220067024231,
      "epoch": 2.4597116200169635,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00044651814627218433,
      "loss": 4.7226,
      "mean_token_accuracy": 0.24518793523311616,
      "num_tokens": 6935855.0,
      "step": 8700
    },
    {
      "entropy": 4.897261381149292,
      "epoch": 2.4611252473847895,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004464515548373645,
      "loss": 4.7303,
      "mean_token_accuracy": 0.23947860300540924,
      "num_tokens": 6939861.0,
      "step": 8705
    },
    {
      "entropy": 4.898551559448242,
      "epoch": 2.462538874752615,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004463849275702299,
      "loss": 4.7531,
      "mean_token_accuracy": 0.24581068605184556,
      "num_tokens": 6944166.0,
      "step": 8710
    },
    {
      "entropy": 4.8112329006195065,
      "epoch": 2.463952502120441,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004463182644847052,
      "loss": 4.7052,
      "mean_token_accuracy": 0.24169591218233108,
      "num_tokens": 6948009.0,
      "step": 8715
    },
    {
      "entropy": 4.862011432647705,
      "epoch": 2.4653661294882667,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004462515655947227,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24529280960559846,
      "num_tokens": 6952269.0,
      "step": 8720
    },
    {
      "entropy": 4.793696737289428,
      "epoch": 2.4667797568560927,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004461848309142219,
      "loss": 4.7706,
      "mean_token_accuracy": 0.23545415997505187,
      "num_tokens": 6956086.0,
      "step": 8725
    },
    {
      "entropy": 4.969640731811523,
      "epoch": 2.4681933842239188,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044611806045715004,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2249186635017395,
      "num_tokens": 6960180.0,
      "step": 8730
    },
    {
      "entropy": 4.934112167358398,
      "epoch": 2.4696070115917443,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004460512542374617,
      "loss": 4.7846,
      "mean_token_accuracy": 0.23657700270414353,
      "num_tokens": 6964352.0,
      "step": 8735
    },
    {
      "entropy": 4.885562133789063,
      "epoch": 2.4710206389595704,
      "grad_norm": 1.765625,
      "learning_rate": 0.00044598441226911903,
      "loss": 4.8619,
      "mean_token_accuracy": 0.2278952866792679,
      "num_tokens": 6968247.0,
      "step": 8740
    },
    {
      "entropy": 4.80047721862793,
      "epoch": 2.472434266327396,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00044591753456609147,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24181403815746308,
      "num_tokens": 6972407.0,
      "step": 8745
    },
    {
      "entropy": 4.804044342041015,
      "epoch": 2.473847893695222,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044585062114235616,
      "loss": 4.763,
      "mean_token_accuracy": 0.23541154116392135,
      "num_tokens": 6976280.0,
      "step": 8750
    },
    {
      "entropy": 4.918060541152954,
      "epoch": 2.4752615210630475,
      "grad_norm": 1.546875,
      "learning_rate": 0.00044578367201189746,
      "loss": 4.791,
      "mean_token_accuracy": 0.24007973670959473,
      "num_tokens": 6980322.0,
      "step": 8755
    },
    {
      "entropy": 5.040391492843628,
      "epoch": 2.4766751484308736,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00044571668718870744,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21947983354330064,
      "num_tokens": 6984653.0,
      "step": 8760
    },
    {
      "entropy": 4.797135496139527,
      "epoch": 2.4780887757986996,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004456496666867855,
      "loss": 4.8338,
      "mean_token_accuracy": 0.24265110343694687,
      "num_tokens": 6988483.0,
      "step": 8765
    },
    {
      "entropy": 4.870652866363526,
      "epoch": 2.479502403166525,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004455826105201385,
      "loss": 4.7619,
      "mean_token_accuracy": 0.24550318121910095,
      "num_tokens": 6992747.0,
      "step": 8770
    },
    {
      "entropy": 4.801453065872193,
      "epoch": 2.480916030534351,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004455155187027807,
      "loss": 4.6066,
      "mean_token_accuracy": 0.24327522665262222,
      "num_tokens": 6996735.0,
      "step": 8775
    },
    {
      "entropy": 4.901353025436402,
      "epoch": 2.482329657902177,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004454483912487339,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24369371980428695,
      "num_tokens": 7000666.0,
      "step": 8780
    },
    {
      "entropy": 4.803237724304199,
      "epoch": 2.4837432852700028,
      "grad_norm": 1.515625,
      "learning_rate": 0.00044538122817202745,
      "loss": 4.6335,
      "mean_token_accuracy": 0.2512310028076172,
      "num_tokens": 7004622.0,
      "step": 8785
    },
    {
      "entropy": 4.7489025592803955,
      "epoch": 2.485156912637829,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004453140294866979,
      "loss": 4.6903,
      "mean_token_accuracy": 0.2521385595202446,
      "num_tokens": 7008685.0,
      "step": 8790
    },
    {
      "entropy": 4.679085159301758,
      "epoch": 2.4865705400056544,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00044524679520678936,
      "loss": 4.582,
      "mean_token_accuracy": 0.24725774079561233,
      "num_tokens": 7012756.0,
      "step": 8795
    },
    {
      "entropy": 4.764791631698609,
      "epoch": 2.4879841673734804,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004451795253463534,
      "loss": 4.7181,
      "mean_token_accuracy": 0.2437589257955551,
      "num_tokens": 7016869.0,
      "step": 8800
    },
    {
      "entropy": 4.855736207962036,
      "epoch": 2.489397794741306,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00044511221991944903,
      "loss": 4.8221,
      "mean_token_accuracy": 0.23986622989177703,
      "num_tokens": 7020594.0,
      "step": 8805
    },
    {
      "entropy": 4.892257833480835,
      "epoch": 2.490811422109132,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044504487894014253,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23632711321115493,
      "num_tokens": 7024674.0,
      "step": 8810
    },
    {
      "entropy": 4.853181028366089,
      "epoch": 2.492225049476958,
      "grad_norm": 1.640625,
      "learning_rate": 0.00044497750242250794,
      "loss": 4.7606,
      "mean_token_accuracy": 0.2432520270347595,
      "num_tokens": 7028302.0,
      "step": 8815
    },
    {
      "entropy": 4.804159164428711,
      "epoch": 2.4936386768447836,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004449100903806264,
      "loss": 4.6896,
      "mean_token_accuracy": 0.2385776549577713,
      "num_tokens": 7032159.0,
      "step": 8820
    },
    {
      "entropy": 4.85439772605896,
      "epoch": 2.4950523042126096,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00044484264282858673,
      "loss": 4.7095,
      "mean_token_accuracy": 0.24124415814876557,
      "num_tokens": 7036309.0,
      "step": 8825
    },
    {
      "entropy": 4.853397369384766,
      "epoch": 2.4964659315804356,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044477515978048497,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24436111152172088,
      "num_tokens": 7040325.0,
      "step": 8830
    },
    {
      "entropy": 4.844921207427978,
      "epoch": 2.497879558948261,
      "grad_norm": 1.640625,
      "learning_rate": 0.00044470764125042464,
      "loss": 4.7069,
      "mean_token_accuracy": 0.24830240607261658,
      "num_tokens": 7044335.0,
      "step": 8835
    },
    {
      "entropy": 4.927169561386108,
      "epoch": 2.499293186316087,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044464008725251673,
      "loss": 4.7462,
      "mean_token_accuracy": 0.24252694994211196,
      "num_tokens": 7048703.0,
      "step": 8840
    },
    {
      "entropy": 4.906860589981079,
      "epoch": 2.500706813683913,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044457249780087963,
      "loss": 4.8216,
      "mean_token_accuracy": 0.23546114563941956,
      "num_tokens": 7052733.0,
      "step": 8845
    },
    {
      "entropy": 4.9245484352111815,
      "epoch": 2.502120441051739,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044450487290963913,
      "loss": 4.8573,
      "mean_token_accuracy": 0.2315475359559059,
      "num_tokens": 7056969.0,
      "step": 8850
    },
    {
      "entropy": 4.974282932281494,
      "epoch": 2.5035340684195644,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004444372125929283,
      "loss": 4.8543,
      "mean_token_accuracy": 0.23569341748952866,
      "num_tokens": 7060769.0,
      "step": 8855
    },
    {
      "entropy": 4.894947862625122,
      "epoch": 2.5049476957873904,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00044436951686488787,
      "loss": 4.6334,
      "mean_token_accuracy": 0.25516571551561357,
      "num_tokens": 7064667.0,
      "step": 8860
    },
    {
      "entropy": 4.88723726272583,
      "epoch": 2.5063613231552164,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004443017857396657,
      "loss": 4.8255,
      "mean_token_accuracy": 0.23390086144208908,
      "num_tokens": 7068675.0,
      "step": 8865
    },
    {
      "entropy": 4.828840827941894,
      "epoch": 2.507774950523042,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044423401923141724,
      "loss": 4.8016,
      "mean_token_accuracy": 0.2369426265358925,
      "num_tokens": 7072628.0,
      "step": 8870
    },
    {
      "entropy": 4.830413770675659,
      "epoch": 2.509188577890868,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004441662173543054,
      "loss": 4.7409,
      "mean_token_accuracy": 0.24998931139707564,
      "num_tokens": 7076879.0,
      "step": 8875
    },
    {
      "entropy": 4.988489246368408,
      "epoch": 2.510602205258694,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004440983801225001,
      "loss": 4.76,
      "mean_token_accuracy": 0.232918643951416,
      "num_tokens": 7080620.0,
      "step": 8880
    },
    {
      "entropy": 4.875664234161377,
      "epoch": 2.5120158326265196,
      "grad_norm": 1.4375,
      "learning_rate": 0.00044403050755017907,
      "loss": 4.8318,
      "mean_token_accuracy": 0.2418210968375206,
      "num_tokens": 7084627.0,
      "step": 8885
    },
    {
      "entropy": 4.753763437271118,
      "epoch": 2.5134294599943456,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004439625996515272,
      "loss": 4.5917,
      "mean_token_accuracy": 0.2601375997066498,
      "num_tokens": 7088427.0,
      "step": 8890
    },
    {
      "entropy": 4.8524558544158936,
      "epoch": 2.514843087362171,
      "grad_norm": 1.703125,
      "learning_rate": 0.00044389465644073685,
      "loss": 4.7411,
      "mean_token_accuracy": 0.24338814467191697,
      "num_tokens": 7092372.0,
      "step": 8895
    },
    {
      "entropy": 4.9182525157928465,
      "epoch": 2.5162567147299972,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00044382667793200776,
      "loss": 4.7894,
      "mean_token_accuracy": 0.22826378792524338,
      "num_tokens": 7096150.0,
      "step": 8900
    },
    {
      "entropy": 4.795201396942138,
      "epoch": 2.517670342097823,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004437586641395469,
      "loss": 4.652,
      "mean_token_accuracy": 0.2425897091627121,
      "num_tokens": 7100209.0,
      "step": 8905
    },
    {
      "entropy": 4.893322324752807,
      "epoch": 2.519083969465649,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004436906150775688,
      "loss": 4.8317,
      "mean_token_accuracy": 0.23420936316251756,
      "num_tokens": 7104278.0,
      "step": 8910
    },
    {
      "entropy": 4.812483406066894,
      "epoch": 2.520497596833475,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00044362253076029543,
      "loss": 4.6602,
      "mean_token_accuracy": 0.2503393366932869,
      "num_tokens": 7108139.0,
      "step": 8915
    },
    {
      "entropy": 4.876856088638306,
      "epoch": 2.5219112242013004,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004435544112019557,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2354557916522026,
      "num_tokens": 7112214.0,
      "step": 8920
    },
    {
      "entropy": 4.905943822860718,
      "epoch": 2.5233248515691264,
      "grad_norm": 1.515625,
      "learning_rate": 0.00044348625641678634,
      "loss": 4.7956,
      "mean_token_accuracy": 0.2358122155070305,
      "num_tokens": 7116545.0,
      "step": 8925
    },
    {
      "entropy": 4.723240089416504,
      "epoch": 2.5247384789369525,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004434180664190314,
      "loss": 4.5672,
      "mean_token_accuracy": 0.26413781344890597,
      "num_tokens": 7120519.0,
      "step": 8930
    },
    {
      "entropy": 4.765473031997681,
      "epoch": 2.526152106304778,
      "grad_norm": 1.59375,
      "learning_rate": 0.000443349841222942,
      "loss": 4.7513,
      "mean_token_accuracy": 0.23854902684688567,
      "num_tokens": 7124567.0,
      "step": 8935
    },
    {
      "entropy": 4.858436393737793,
      "epoch": 2.527565733672604,
      "grad_norm": 1.6875,
      "learning_rate": 0.00044328158084277683,
      "loss": 4.7846,
      "mean_token_accuracy": 0.23841999918222428,
      "num_tokens": 7128670.0,
      "step": 8940
    },
    {
      "entropy": 4.915992403030396,
      "epoch": 2.5289793610404296,
      "grad_norm": 1.6875,
      "learning_rate": 0.00044321328529280183,
      "loss": 4.737,
      "mean_token_accuracy": 0.24155183583498002,
      "num_tokens": 7132504.0,
      "step": 8945
    },
    {
      "entropy": 4.809674549102783,
      "epoch": 2.5303929884082557,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004431449545872904,
      "loss": 4.6837,
      "mean_token_accuracy": 0.24609342068433762,
      "num_tokens": 7136504.0,
      "step": 8950
    },
    {
      "entropy": 4.786500072479248,
      "epoch": 2.5318066157760812,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00044307658874052333,
      "loss": 4.6845,
      "mean_token_accuracy": 0.24187166392803192,
      "num_tokens": 7140253.0,
      "step": 8955
    },
    {
      "entropy": 4.832511901855469,
      "epoch": 2.5332202431439073,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004430081877667886,
      "loss": 4.82,
      "mean_token_accuracy": 0.2334180250763893,
      "num_tokens": 7144699.0,
      "step": 8960
    },
    {
      "entropy": 4.851031637191772,
      "epoch": 2.5346338705117333,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004429397516803814,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24707866162061692,
      "num_tokens": 7148643.0,
      "step": 8965
    },
    {
      "entropy": 4.817906999588013,
      "epoch": 2.536047497879559,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004428712804956047,
      "loss": 4.8535,
      "mean_token_accuracy": 0.22581046670675278,
      "num_tokens": 7152619.0,
      "step": 8970
    },
    {
      "entropy": 4.886417436599731,
      "epoch": 2.537461125247385,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004428027742267685,
      "loss": 4.7967,
      "mean_token_accuracy": 0.23800897598266602,
      "num_tokens": 7156543.0,
      "step": 8975
    },
    {
      "entropy": 4.8807902336120605,
      "epoch": 2.538874752615211,
      "grad_norm": 1.59375,
      "learning_rate": 0.00044273423288819004,
      "loss": 4.7717,
      "mean_token_accuracy": 0.23831001669168472,
      "num_tokens": 7160440.0,
      "step": 8980
    },
    {
      "entropy": 4.9313171863555905,
      "epoch": 2.5402883799830365,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004426656564941942,
      "loss": 4.8559,
      "mean_token_accuracy": 0.24424826204776764,
      "num_tokens": 7164585.0,
      "step": 8985
    },
    {
      "entropy": 4.840216827392578,
      "epoch": 2.541702007350862,
      "grad_norm": 1.609375,
      "learning_rate": 0.000442597045059113,
      "loss": 4.7606,
      "mean_token_accuracy": 0.2503294885158539,
      "num_tokens": 7168527.0,
      "step": 8990
    },
    {
      "entropy": 4.913680219650269,
      "epoch": 2.543115634718688,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004425283985972857,
      "loss": 4.7804,
      "mean_token_accuracy": 0.2326314628124237,
      "num_tokens": 7172494.0,
      "step": 8995
    },
    {
      "entropy": 4.836026668548584,
      "epoch": 2.544529262086514,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004424597171230591,
      "loss": 4.7791,
      "mean_token_accuracy": 0.23710203170776367,
      "num_tokens": 7176627.0,
      "step": 9000
    },
    {
      "epoch": 2.544529262086514,
      "eval_entropy": 4.708660123366823,
      "eval_loss": 5.168454647064209,
      "eval_mean_token_accuracy": 0.23345178927537147,
      "eval_num_tokens": 7176627.0,
      "eval_runtime": 6.7608,
      "eval_samples_per_second": 1669.03,
      "eval_steps_per_second": 208.703,
      "step": 9000
    },
    {
      "entropy": 4.814554500579834,
      "epoch": 2.5459428894543397,
      "grad_norm": 1.5625,
      "learning_rate": 0.00044239100065078714,
      "loss": 4.7061,
      "mean_token_accuracy": 0.2418840855360031,
      "num_tokens": 7180398.0,
      "step": 9005
    },
    {
      "entropy": 4.867837238311767,
      "epoch": 2.5473565168221657,
      "grad_norm": 1.65625,
      "learning_rate": 0.00044232224919483115,
      "loss": 4.7652,
      "mean_token_accuracy": 0.2526453971862793,
      "num_tokens": 7184182.0,
      "step": 9010
    },
    {
      "entropy": 4.846005058288574,
      "epoch": 2.5487701441899917,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004422534627695598,
      "loss": 4.7197,
      "mean_token_accuracy": 0.2320114865899086,
      "num_tokens": 7188139.0,
      "step": 9015
    },
    {
      "entropy": 4.793542098999024,
      "epoch": 2.5501837715578173,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000442184641389349,
      "loss": 4.7948,
      "mean_token_accuracy": 0.24134827852249147,
      "num_tokens": 7191848.0,
      "step": 9020
    },
    {
      "entropy": 4.956062889099121,
      "epoch": 2.5515973989256433,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004421157850685819,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22680881321430207,
      "num_tokens": 7196039.0,
      "step": 9025
    },
    {
      "entropy": 4.837375259399414,
      "epoch": 2.553011026293469,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00044204689382164927,
      "loss": 4.739,
      "mean_token_accuracy": 0.23553290218114853,
      "num_tokens": 7200075.0,
      "step": 9030
    },
    {
      "entropy": 4.8519491195678714,
      "epoch": 2.554424653661295,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004419779676629487,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22773234993219377,
      "num_tokens": 7204236.0,
      "step": 9035
    },
    {
      "entropy": 4.848725605010986,
      "epoch": 2.5558382810291205,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004419090066068856,
      "loss": 4.7354,
      "mean_token_accuracy": 0.2475093796849251,
      "num_tokens": 7208390.0,
      "step": 9040
    },
    {
      "entropy": 4.757932186126709,
      "epoch": 2.5572519083969465,
      "grad_norm": 1.5,
      "learning_rate": 0.0004418400106678721,
      "loss": 4.6166,
      "mean_token_accuracy": 0.2546589091420174,
      "num_tokens": 7212355.0,
      "step": 9045
    },
    {
      "entropy": 4.84939546585083,
      "epoch": 2.5586655357647725,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00044177097986032815,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2400248572230339,
      "num_tokens": 7216336.0,
      "step": 9050
    },
    {
      "entropy": 4.934862232208252,
      "epoch": 2.560079163132598,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00044170191419868064,
      "loss": 4.852,
      "mean_token_accuracy": 0.23075196892023087,
      "num_tokens": 7220115.0,
      "step": 9055
    },
    {
      "entropy": 4.862721014022827,
      "epoch": 2.561492790500424,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000441632813697364,
      "loss": 4.6888,
      "mean_token_accuracy": 0.24001567661762238,
      "num_tokens": 7224230.0,
      "step": 9060
    },
    {
      "entropy": 4.903478288650513,
      "epoch": 2.56290641786825,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00044156367837081966,
      "loss": 4.7676,
      "mean_token_accuracy": 0.23998010158538818,
      "num_tokens": 7228186.0,
      "step": 9065
    },
    {
      "entropy": 4.885419178009033,
      "epoch": 2.5643200452360757,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004414945082334966,
      "loss": 4.8133,
      "mean_token_accuracy": 0.24057676643133163,
      "num_tokens": 7231953.0,
      "step": 9070
    },
    {
      "entropy": 4.8087492942810055,
      "epoch": 2.5657336726039017,
      "grad_norm": 1.78125,
      "learning_rate": 0.0004414253032998508,
      "loss": 4.6534,
      "mean_token_accuracy": 0.25096642673015596,
      "num_tokens": 7235810.0,
      "step": 9075
    },
    {
      "entropy": 4.829167747497559,
      "epoch": 2.5671472999717273,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004413560635843458,
      "loss": 4.7341,
      "mean_token_accuracy": 0.24466728270053864,
      "num_tokens": 7239879.0,
      "step": 9080
    },
    {
      "entropy": 4.935311412811279,
      "epoch": 2.5685609273395533,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044128678910145225,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23399273604154586,
      "num_tokens": 7243948.0,
      "step": 9085
    },
    {
      "entropy": 4.7788300037384035,
      "epoch": 2.569974554707379,
      "grad_norm": 1.59375,
      "learning_rate": 0.000441217479865648,
      "loss": 4.5463,
      "mean_token_accuracy": 0.2636043965816498,
      "num_tokens": 7248024.0,
      "step": 9090
    },
    {
      "entropy": 4.835382986068725,
      "epoch": 2.571388182075205,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004411481358914184,
      "loss": 4.7145,
      "mean_token_accuracy": 0.2367924854159355,
      "num_tokens": 7252130.0,
      "step": 9095
    },
    {
      "entropy": 4.76817307472229,
      "epoch": 2.572801809443031,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004410787571932558,
      "loss": 4.7785,
      "mean_token_accuracy": 0.24601975530385972,
      "num_tokens": 7256326.0,
      "step": 9100
    },
    {
      "entropy": 4.928644514083862,
      "epoch": 2.5742154368108565,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004410093437856599,
      "loss": 4.8377,
      "mean_token_accuracy": 0.22990526258945465,
      "num_tokens": 7260507.0,
      "step": 9105
    },
    {
      "entropy": 5.006984710693359,
      "epoch": 2.5756290641786825,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00044093989568313774,
      "loss": 4.6593,
      "mean_token_accuracy": 0.2568992644548416,
      "num_tokens": 7264805.0,
      "step": 9110
    },
    {
      "entropy": 4.90860915184021,
      "epoch": 2.5770426915465086,
      "grad_norm": 1.9453125,
      "learning_rate": 0.00044087041290020356,
      "loss": 4.6686,
      "mean_token_accuracy": 0.24236927181482315,
      "num_tokens": 7268699.0,
      "step": 9115
    },
    {
      "entropy": 4.77135272026062,
      "epoch": 2.578456318914334,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004408008954513787,
      "loss": 4.6104,
      "mean_token_accuracy": 0.259738627076149,
      "num_tokens": 7272601.0,
      "step": 9120
    },
    {
      "entropy": 4.85414309501648,
      "epoch": 2.57986994628216,
      "grad_norm": 1.625,
      "learning_rate": 0.0004407313433511921,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23150228559970856,
      "num_tokens": 7276570.0,
      "step": 9125
    },
    {
      "entropy": 4.878512811660767,
      "epoch": 2.5812835736499857,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00044066175661417947,
      "loss": 4.753,
      "mean_token_accuracy": 0.23851029723882675,
      "num_tokens": 7280215.0,
      "step": 9130
    },
    {
      "entropy": 4.838480377197266,
      "epoch": 2.5826972010178118,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004405921352548842,
      "loss": 4.647,
      "mean_token_accuracy": 0.2585996761918068,
      "num_tokens": 7284148.0,
      "step": 9135
    },
    {
      "entropy": 4.855335330963134,
      "epoch": 2.5841108283856373,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004405224792878567,
      "loss": 4.776,
      "mean_token_accuracy": 0.23603290021419526,
      "num_tokens": 7288408.0,
      "step": 9140
    },
    {
      "entropy": 4.910908126831055,
      "epoch": 2.5855244557534633,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004404527887276545,
      "loss": 4.8374,
      "mean_token_accuracy": 0.2344924509525299,
      "num_tokens": 7292157.0,
      "step": 9145
    },
    {
      "entropy": 4.917688608169556,
      "epoch": 2.5869380831212894,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004403830635888426,
      "loss": 4.7137,
      "mean_token_accuracy": 0.24108498990535737,
      "num_tokens": 7296193.0,
      "step": 9150
    },
    {
      "entropy": 4.780224275588989,
      "epoch": 2.588351710489115,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00044031330388599315,
      "loss": 4.7127,
      "mean_token_accuracy": 0.24063062965869902,
      "num_tokens": 7300186.0,
      "step": 9155
    },
    {
      "entropy": 4.859019470214844,
      "epoch": 2.589765337856941,
      "grad_norm": 2.015625,
      "learning_rate": 0.0004402435096336854,
      "loss": 4.794,
      "mean_token_accuracy": 0.2335315614938736,
      "num_tokens": 7304559.0,
      "step": 9160
    },
    {
      "entropy": 4.876143121719361,
      "epoch": 2.591178965224767,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00044017368084650605,
      "loss": 4.7138,
      "mean_token_accuracy": 0.2465211644768715,
      "num_tokens": 7308827.0,
      "step": 9165
    },
    {
      "entropy": 4.852374315261841,
      "epoch": 2.5925925925925926,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004401038175390488,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24826837033033372,
      "num_tokens": 7313038.0,
      "step": 9170
    },
    {
      "entropy": 4.8656433582305905,
      "epoch": 2.5940062199604186,
      "grad_norm": 1.625,
      "learning_rate": 0.00044003391972591456,
      "loss": 4.6447,
      "mean_token_accuracy": 0.2521489173173904,
      "num_tokens": 7316913.0,
      "step": 9175
    },
    {
      "entropy": 4.822190809249878,
      "epoch": 2.595419847328244,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004399639874217117,
      "loss": 4.7041,
      "mean_token_accuracy": 0.2504032298922539,
      "num_tokens": 7320634.0,
      "step": 9180
    },
    {
      "entropy": 4.855980587005615,
      "epoch": 2.59683347469607,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043989402064105557,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23528171181678773,
      "num_tokens": 7324663.0,
      "step": 9185
    },
    {
      "entropy": 4.75980544090271,
      "epoch": 2.5982471020638958,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043982401939856875,
      "loss": 4.6013,
      "mean_token_accuracy": 0.2534924611449242,
      "num_tokens": 7328814.0,
      "step": 9190
    },
    {
      "entropy": 4.84786319732666,
      "epoch": 2.5996607294317218,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043975398370888115,
      "loss": 4.643,
      "mean_token_accuracy": 0.24632356613874434,
      "num_tokens": 7332780.0,
      "step": 9195
    },
    {
      "entropy": 4.902121067047119,
      "epoch": 2.601074356799548,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004396839135866298,
      "loss": 4.9097,
      "mean_token_accuracy": 0.22337397038936616,
      "num_tokens": 7336769.0,
      "step": 9200
    },
    {
      "entropy": 4.888132047653198,
      "epoch": 2.6024879841673734,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004396138090464589,
      "loss": 4.7439,
      "mean_token_accuracy": 0.24505586475133895,
      "num_tokens": 7340943.0,
      "step": 9205
    },
    {
      "entropy": 4.8074321269989015,
      "epoch": 2.6039016115351994,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004395436701030198,
      "loss": 4.7591,
      "mean_token_accuracy": 0.23799658715724945,
      "num_tokens": 7344944.0,
      "step": 9210
    },
    {
      "entropy": 4.8759119510650635,
      "epoch": 2.6053152389030254,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043947349677097116,
      "loss": 4.8241,
      "mean_token_accuracy": 0.2355223134160042,
      "num_tokens": 7348778.0,
      "step": 9215
    },
    {
      "entropy": 4.818848943710327,
      "epoch": 2.606728866270851,
      "grad_norm": 1.5,
      "learning_rate": 0.00043940328906497876,
      "loss": 4.7011,
      "mean_token_accuracy": 0.24501485377550125,
      "num_tokens": 7352878.0,
      "step": 9220
    },
    {
      "entropy": 4.9126060962677,
      "epoch": 2.608142493638677,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004393330469997156,
      "loss": 4.7728,
      "mean_token_accuracy": 0.24125273823738097,
      "num_tokens": 7356825.0,
      "step": 9225
    },
    {
      "entropy": 4.707981061935425,
      "epoch": 2.6095561210065026,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004392627705898619,
      "loss": 4.6316,
      "mean_token_accuracy": 0.2630636215209961,
      "num_tokens": 7360603.0,
      "step": 9230
    },
    {
      "entropy": 4.888771104812622,
      "epoch": 2.6109697483743286,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00043919245985010475,
      "loss": 4.8861,
      "mean_token_accuracy": 0.22529222667217255,
      "num_tokens": 7364757.0,
      "step": 9235
    },
    {
      "entropy": 4.8783056259155275,
      "epoch": 2.612383375742154,
      "grad_norm": 1.78125,
      "learning_rate": 0.00043912211479513896,
      "loss": 4.6886,
      "mean_token_accuracy": 0.2461331456899643,
      "num_tokens": 7368581.0,
      "step": 9240
    },
    {
      "entropy": 4.935618114471436,
      "epoch": 2.61379700310998,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000439051735439666,
      "loss": 4.7561,
      "mean_token_accuracy": 0.24252170473337173,
      "num_tokens": 7372448.0,
      "step": 9245
    },
    {
      "entropy": 4.901508617401123,
      "epoch": 2.615210630477806,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004389813217983949,
      "loss": 4.7172,
      "mean_token_accuracy": 0.24929889738559724,
      "num_tokens": 7376464.0,
      "step": 9250
    },
    {
      "entropy": 4.940105199813843,
      "epoch": 2.616624257845632,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004389108738860415,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22001878619194032,
      "num_tokens": 7380445.0,
      "step": 9255
    },
    {
      "entropy": 4.979659557342529,
      "epoch": 2.618037885213458,
      "grad_norm": 1.46875,
      "learning_rate": 0.000438840391717329,
      "loss": 4.8565,
      "mean_token_accuracy": 0.2274993985891342,
      "num_tokens": 7384848.0,
      "step": 9260
    },
    {
      "entropy": 4.901060104370117,
      "epoch": 2.619451512581284,
      "grad_norm": 1.578125,
      "learning_rate": 0.00043876987530698785,
      "loss": 4.7033,
      "mean_token_accuracy": 0.24396551847457887,
      "num_tokens": 7388665.0,
      "step": 9265
    },
    {
      "entropy": 4.841419458389282,
      "epoch": 2.6208651399491094,
      "grad_norm": 1.515625,
      "learning_rate": 0.00043869932466975546,
      "loss": 4.6968,
      "mean_token_accuracy": 0.24063999950885773,
      "num_tokens": 7392807.0,
      "step": 9270
    },
    {
      "entropy": 4.812009572982788,
      "epoch": 2.622278767316935,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004386287398203766,
      "loss": 4.6437,
      "mean_token_accuracy": 0.2569728970527649,
      "num_tokens": 7396592.0,
      "step": 9275
    },
    {
      "entropy": 4.875644636154175,
      "epoch": 2.623692394684761,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004385581207736029,
      "loss": 4.7031,
      "mean_token_accuracy": 0.24194696098566054,
      "num_tokens": 7400540.0,
      "step": 9280
    },
    {
      "entropy": 4.881341028213501,
      "epoch": 2.625106022052587,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00043848746754419335,
      "loss": 4.7037,
      "mean_token_accuracy": 0.24883255511522293,
      "num_tokens": 7404467.0,
      "step": 9285
    },
    {
      "entropy": 4.9059161186218265,
      "epoch": 2.6265196494204126,
      "grad_norm": 1.609375,
      "learning_rate": 0.000438416780146914,
      "loss": 4.6851,
      "mean_token_accuracy": 0.24317810833454132,
      "num_tokens": 7408377.0,
      "step": 9290
    },
    {
      "entropy": 4.85079779624939,
      "epoch": 2.6279332767882386,
      "grad_norm": 1.734375,
      "learning_rate": 0.00043834605859653826,
      "loss": 4.6548,
      "mean_token_accuracy": 0.25417670160531997,
      "num_tokens": 7412375.0,
      "step": 9295
    },
    {
      "entropy": 4.946078634262085,
      "epoch": 2.6293469041560646,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00043827530290784634,
      "loss": 4.8506,
      "mean_token_accuracy": 0.22275288105010987,
      "num_tokens": 7416613.0,
      "step": 9300
    },
    {
      "entropy": 4.94378514289856,
      "epoch": 2.6307605315238902,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043820451309562576,
      "loss": 4.711,
      "mean_token_accuracy": 0.2482324868440628,
      "num_tokens": 7420558.0,
      "step": 9305
    },
    {
      "entropy": 4.833695793151856,
      "epoch": 2.6321741588917162,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004381336891746711,
      "loss": 4.7305,
      "mean_token_accuracy": 0.24658335745334625,
      "num_tokens": 7424394.0,
      "step": 9310
    },
    {
      "entropy": 4.9757139682769775,
      "epoch": 2.633587786259542,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004380628311597843,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21814664155244828,
      "num_tokens": 7428751.0,
      "step": 9315
    },
    {
      "entropy": 4.946564149856568,
      "epoch": 2.635001413627368,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00043799193906577413,
      "loss": 4.9415,
      "mean_token_accuracy": 0.22746722102165223,
      "num_tokens": 7432816.0,
      "step": 9320
    },
    {
      "entropy": 4.862512111663818,
      "epoch": 2.6364150409951934,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00043792101290745654,
      "loss": 4.6679,
      "mean_token_accuracy": 0.25913803428411486,
      "num_tokens": 7436754.0,
      "step": 9325
    },
    {
      "entropy": 4.858878707885742,
      "epoch": 2.6378286683630194,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004378500526996548,
      "loss": 4.7187,
      "mean_token_accuracy": 0.2540144264698029,
      "num_tokens": 7440742.0,
      "step": 9330
    },
    {
      "entropy": 4.818402338027954,
      "epoch": 2.6392422957308455,
      "grad_norm": 1.609375,
      "learning_rate": 0.000437779058457199,
      "loss": 4.6839,
      "mean_token_accuracy": 0.25300066620111467,
      "num_tokens": 7444565.0,
      "step": 9335
    },
    {
      "entropy": 4.890596342086792,
      "epoch": 2.640655923098671,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00043770803019492667,
      "loss": 4.8061,
      "mean_token_accuracy": 0.23124598264694213,
      "num_tokens": 7448510.0,
      "step": 9340
    },
    {
      "entropy": 5.016450309753418,
      "epoch": 2.642069550466497,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004376369679276821,
      "loss": 4.6832,
      "mean_token_accuracy": 0.25214518159627913,
      "num_tokens": 7452536.0,
      "step": 9345
    },
    {
      "entropy": 4.914615249633789,
      "epoch": 2.643483177834323,
      "grad_norm": 1.5625,
      "learning_rate": 0.000437565871670317,
      "loss": 4.7576,
      "mean_token_accuracy": 0.24715423732995986,
      "num_tokens": 7456245.0,
      "step": 9350
    },
    {
      "entropy": 4.871857976913452,
      "epoch": 2.6448968052021486,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004374947414376899,
      "loss": 4.6927,
      "mean_token_accuracy": 0.243508879840374,
      "num_tokens": 7460306.0,
      "step": 9355
    },
    {
      "entropy": 4.890254640579224,
      "epoch": 2.6463104325699747,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004374235772446668,
      "loss": 4.7743,
      "mean_token_accuracy": 0.2396906465291977,
      "num_tokens": 7464103.0,
      "step": 9360
    },
    {
      "entropy": 4.985165023803711,
      "epoch": 2.6477240599378002,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004373523791061204,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23002199679613114,
      "num_tokens": 7468316.0,
      "step": 9365
    },
    {
      "entropy": 5.005202388763427,
      "epoch": 2.6491376873056263,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00043728114703693073,
      "loss": 4.9047,
      "mean_token_accuracy": 0.23780269622802735,
      "num_tokens": 7472254.0,
      "step": 9370
    },
    {
      "entropy": 4.9046228408813475,
      "epoch": 2.650551314673452,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004372098810519849,
      "loss": 4.7258,
      "mean_token_accuracy": 0.25327771306037905,
      "num_tokens": 7476345.0,
      "step": 9375
    },
    {
      "entropy": 4.787554502487183,
      "epoch": 2.651964942041278,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004371385811661769,
      "loss": 4.6731,
      "mean_token_accuracy": 0.2482880935072899,
      "num_tokens": 7480210.0,
      "step": 9380
    },
    {
      "entropy": 4.81629433631897,
      "epoch": 2.653378569409104,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043706724739440815,
      "loss": 4.7158,
      "mean_token_accuracy": 0.2460557624697685,
      "num_tokens": 7483861.0,
      "step": 9385
    },
    {
      "entropy": 4.7258744716644285,
      "epoch": 2.6547921967769295,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004369958797515869,
      "loss": 4.6556,
      "mean_token_accuracy": 0.25460451692342756,
      "num_tokens": 7488041.0,
      "step": 9390
    },
    {
      "entropy": 4.761253881454468,
      "epoch": 2.6562058241447555,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004369244782526284,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23764937967061997,
      "num_tokens": 7492122.0,
      "step": 9395
    },
    {
      "entropy": 4.747553396224975,
      "epoch": 2.6576194515125815,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043685304291245545,
      "loss": 4.6094,
      "mean_token_accuracy": 0.2603394344449043,
      "num_tokens": 7496102.0,
      "step": 9400
    },
    {
      "entropy": 4.830163097381591,
      "epoch": 2.659033078880407,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043678157374599724,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23859505355358124,
      "num_tokens": 7500084.0,
      "step": 9405
    },
    {
      "entropy": 4.803248167037964,
      "epoch": 2.660446706248233,
      "grad_norm": 1.640625,
      "learning_rate": 0.00043671007076819066,
      "loss": 4.6452,
      "mean_token_accuracy": 0.247104711830616,
      "num_tokens": 7504068.0,
      "step": 9410
    },
    {
      "entropy": 4.871783065795898,
      "epoch": 2.6618603336160587,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00043663853399397923,
      "loss": 4.6596,
      "mean_token_accuracy": 0.24927788823843003,
      "num_tokens": 7507903.0,
      "step": 9415
    },
    {
      "entropy": 4.907245779037476,
      "epoch": 2.6632739609838847,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043656696343831377,
      "loss": 4.7177,
      "mean_token_accuracy": 0.2383325695991516,
      "num_tokens": 7511694.0,
      "step": 9420
    },
    {
      "entropy": 4.79429407119751,
      "epoch": 2.6646875883517103,
      "grad_norm": 1.546875,
      "learning_rate": 0.000436495359116152,
      "loss": 4.6365,
      "mean_token_accuracy": 0.2532139614224434,
      "num_tokens": 7516120.0,
      "step": 9425
    },
    {
      "entropy": 4.870751523971558,
      "epoch": 2.6661012157195363,
      "grad_norm": 1.515625,
      "learning_rate": 0.00043642372104245883,
      "loss": 4.6672,
      "mean_token_accuracy": 0.236220683157444,
      "num_tokens": 7520068.0,
      "step": 9430
    },
    {
      "entropy": 4.852788877487183,
      "epoch": 2.6675148430873623,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004363520492322062,
      "loss": 4.6973,
      "mean_token_accuracy": 0.24868786185979844,
      "num_tokens": 7524136.0,
      "step": 9435
    },
    {
      "entropy": 4.891140556335449,
      "epoch": 2.668928470455188,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000436280343700373,
      "loss": 4.6838,
      "mean_token_accuracy": 0.24473499357700348,
      "num_tokens": 7528026.0,
      "step": 9440
    },
    {
      "entropy": 4.7884596824646,
      "epoch": 2.670342097823014,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004362086044619454,
      "loss": 4.651,
      "mean_token_accuracy": 0.25072305649518967,
      "num_tokens": 7532149.0,
      "step": 9445
    },
    {
      "entropy": 4.871756935119629,
      "epoch": 2.67175572519084,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004361368315319162,
      "loss": 4.8039,
      "mean_token_accuracy": 0.22879233360290527,
      "num_tokens": 7536270.0,
      "step": 9450
    },
    {
      "entropy": 4.835328817367554,
      "epoch": 2.6731693525586655,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00043606502492528576,
      "loss": 4.7499,
      "mean_token_accuracy": 0.247857066988945,
      "num_tokens": 7540157.0,
      "step": 9455
    },
    {
      "entropy": 4.88475546836853,
      "epoch": 2.6745829799264915,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004359931846570611,
      "loss": 4.7425,
      "mean_token_accuracy": 0.2472950965166092,
      "num_tokens": 7543861.0,
      "step": 9460
    },
    {
      "entropy": 4.850816106796264,
      "epoch": 2.675996607294317,
      "grad_norm": 1.5,
      "learning_rate": 0.0004359213107422564,
      "loss": 4.6866,
      "mean_token_accuracy": 0.255721877515316,
      "num_tokens": 7547709.0,
      "step": 9465
    },
    {
      "entropy": 4.944008779525757,
      "epoch": 2.677410234662143,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004358494031958929,
      "loss": 4.8085,
      "mean_token_accuracy": 0.24241555333137513,
      "num_tokens": 7551773.0,
      "step": 9470
    },
    {
      "entropy": 4.8565101146698,
      "epoch": 2.6788238620299687,
      "grad_norm": 1.6875,
      "learning_rate": 0.00043577746203299865,
      "loss": 4.7441,
      "mean_token_accuracy": 0.23977454453706742,
      "num_tokens": 7555716.0,
      "step": 9475
    },
    {
      "entropy": 5.058261346817017,
      "epoch": 2.6802374893977947,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004357054872686092,
      "loss": 4.9564,
      "mean_token_accuracy": 0.21716484874486924,
      "num_tokens": 7559583.0,
      "step": 9480
    },
    {
      "entropy": 4.826788711547851,
      "epoch": 2.6816511167656207,
      "grad_norm": 1.625,
      "learning_rate": 0.0004356334789177666,
      "loss": 4.6349,
      "mean_token_accuracy": 0.2560909867286682,
      "num_tokens": 7563484.0,
      "step": 9485
    },
    {
      "entropy": 4.7411651611328125,
      "epoch": 2.6830647441334463,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043556143699552034,
      "loss": 4.7007,
      "mean_token_accuracy": 0.2464948996901512,
      "num_tokens": 7567196.0,
      "step": 9490
    },
    {
      "entropy": 4.838413667678833,
      "epoch": 2.6844783715012723,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004354893615169265,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2351538673043251,
      "num_tokens": 7571161.0,
      "step": 9495
    },
    {
      "entropy": 4.945151281356812,
      "epoch": 2.6858919988690984,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00043541725249704856,
      "loss": 4.8501,
      "mean_token_accuracy": 0.23502508401870728,
      "num_tokens": 7575226.0,
      "step": 9500
    },
    {
      "entropy": 4.978336572647095,
      "epoch": 2.687305626236924,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004353451099509569,
      "loss": 4.7903,
      "mean_token_accuracy": 0.23586281538009643,
      "num_tokens": 7579444.0,
      "step": 9505
    },
    {
      "entropy": 4.9068646907806395,
      "epoch": 2.68871925360475,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004352729338937288,
      "loss": 4.8051,
      "mean_token_accuracy": 0.2435090109705925,
      "num_tokens": 7583298.0,
      "step": 9510
    },
    {
      "entropy": 4.937707185745239,
      "epoch": 2.6901328809725755,
      "grad_norm": 1.5,
      "learning_rate": 0.0004352007243404486,
      "loss": 4.7398,
      "mean_token_accuracy": 0.23945448994636537,
      "num_tokens": 7587292.0,
      "step": 9515
    },
    {
      "entropy": 4.892711877822876,
      "epoch": 2.6915465083404015,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043512848130620763,
      "loss": 4.7233,
      "mean_token_accuracy": 0.24224225878715516,
      "num_tokens": 7591427.0,
      "step": 9520
    },
    {
      "entropy": 4.881094360351563,
      "epoch": 2.692960135708227,
      "grad_norm": 1.671875,
      "learning_rate": 0.00043505620480610433,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23850062042474746,
      "num_tokens": 7595121.0,
      "step": 9525
    },
    {
      "entropy": 4.887492752075195,
      "epoch": 2.694373763076053,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00043498389485524395,
      "loss": 4.7726,
      "mean_token_accuracy": 0.23310711085796357,
      "num_tokens": 7599072.0,
      "step": 9530
    },
    {
      "entropy": 4.947204494476319,
      "epoch": 2.695787390443879,
      "grad_norm": 1.6875,
      "learning_rate": 0.00043491155146873885,
      "loss": 4.7754,
      "mean_token_accuracy": 0.24022504985332488,
      "num_tokens": 7602860.0,
      "step": 9535
    },
    {
      "entropy": 4.917364454269409,
      "epoch": 2.6972010178117047,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0004348391746617083,
      "loss": 4.778,
      "mean_token_accuracy": 0.22759387344121934,
      "num_tokens": 7606856.0,
      "step": 9540
    },
    {
      "entropy": 4.954983711242676,
      "epoch": 2.6986146451795308,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004347667644492788,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22933456897735596,
      "num_tokens": 7611148.0,
      "step": 9545
    },
    {
      "entropy": 4.961693143844604,
      "epoch": 2.700028272547357,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00043469432084658336,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23929639309644699,
      "num_tokens": 7615121.0,
      "step": 9550
    },
    {
      "entropy": 4.868210887908935,
      "epoch": 2.7014418999151824,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00043462184386876246,
      "loss": 4.6906,
      "mean_token_accuracy": 0.2509275197982788,
      "num_tokens": 7619009.0,
      "step": 9555
    },
    {
      "entropy": 4.839197254180908,
      "epoch": 2.702855527283008,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004345493335309632,
      "loss": 4.7112,
      "mean_token_accuracy": 0.24208156317472457,
      "num_tokens": 7623089.0,
      "step": 9560
    },
    {
      "entropy": 4.7921915531158445,
      "epoch": 2.704269154650834,
      "grad_norm": 1.546875,
      "learning_rate": 0.00043447678984833994,
      "loss": 4.6721,
      "mean_token_accuracy": 0.2547679036855698,
      "num_tokens": 7626880.0,
      "step": 9565
    },
    {
      "entropy": 4.851875495910645,
      "epoch": 2.70568278201866,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004344042128360537,
      "loss": 4.7418,
      "mean_token_accuracy": 0.24519296884536743,
      "num_tokens": 7631050.0,
      "step": 9570
    },
    {
      "entropy": 4.81116132736206,
      "epoch": 2.7070964093864855,
      "grad_norm": 1.46875,
      "learning_rate": 0.00043433160250927277,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2472531631588936,
      "num_tokens": 7635306.0,
      "step": 9575
    },
    {
      "entropy": 4.880113697052002,
      "epoch": 2.7085100367543116,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004342589588831721,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24160737544298172,
      "num_tokens": 7639260.0,
      "step": 9580
    },
    {
      "entropy": 4.892963600158692,
      "epoch": 2.7099236641221376,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000434186281972934,
      "loss": 4.8231,
      "mean_token_accuracy": 0.2370707407593727,
      "num_tokens": 7643241.0,
      "step": 9585
    },
    {
      "entropy": 4.8312053203582765,
      "epoch": 2.711337291489963,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004341135717937473,
      "loss": 4.6924,
      "mean_token_accuracy": 0.2426387384533882,
      "num_tokens": 7647204.0,
      "step": 9590
    },
    {
      "entropy": 4.838108015060425,
      "epoch": 2.712750918857789,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00043404082836080796,
      "loss": 4.7397,
      "mean_token_accuracy": 0.24133705347776413,
      "num_tokens": 7651469.0,
      "step": 9595
    },
    {
      "entropy": 4.822805023193359,
      "epoch": 2.714164546225615,
      "grad_norm": 1.609375,
      "learning_rate": 0.000433968051689319,
      "loss": 4.7467,
      "mean_token_accuracy": 0.24389615952968596,
      "num_tokens": 7655291.0,
      "step": 9600
    },
    {
      "entropy": 4.80342435836792,
      "epoch": 2.715578173593441,
      "grad_norm": 1.640625,
      "learning_rate": 0.00043389524179449025,
      "loss": 4.7097,
      "mean_token_accuracy": 0.2528283417224884,
      "num_tokens": 7659381.0,
      "step": 9605
    },
    {
      "entropy": 4.859865474700928,
      "epoch": 2.7169918009612664,
      "grad_norm": 1.546875,
      "learning_rate": 0.00043382239869153855,
      "loss": 4.7621,
      "mean_token_accuracy": 0.2332729294896126,
      "num_tokens": 7663197.0,
      "step": 9610
    },
    {
      "entropy": 4.789547872543335,
      "epoch": 2.7184054283290924,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00043374952239568767,
      "loss": 4.7201,
      "mean_token_accuracy": 0.23138939440250397,
      "num_tokens": 7667445.0,
      "step": 9615
    },
    {
      "entropy": 4.915864515304565,
      "epoch": 2.7198190556969184,
      "grad_norm": 1.75,
      "learning_rate": 0.00043367661292216823,
      "loss": 4.78,
      "mean_token_accuracy": 0.24267341792583466,
      "num_tokens": 7671184.0,
      "step": 9620
    },
    {
      "entropy": 4.886941814422608,
      "epoch": 2.721232683064744,
      "grad_norm": 1.65625,
      "learning_rate": 0.000433603670286218,
      "loss": 4.6862,
      "mean_token_accuracy": 0.246448615193367,
      "num_tokens": 7674911.0,
      "step": 9625
    },
    {
      "entropy": 4.839103269577026,
      "epoch": 2.72264631043257,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00043353069450308146,
      "loss": 4.7529,
      "mean_token_accuracy": 0.24273636788129807,
      "num_tokens": 7678751.0,
      "step": 9630
    },
    {
      "entropy": 4.848040390014648,
      "epoch": 2.724059937800396,
      "grad_norm": 1.734375,
      "learning_rate": 0.00043345768558800997,
      "loss": 4.6692,
      "mean_token_accuracy": 0.2566663473844528,
      "num_tokens": 7682623.0,
      "step": 9635
    },
    {
      "entropy": 4.769739389419556,
      "epoch": 2.7254735651682216,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004333846435562622,
      "loss": 4.6512,
      "mean_token_accuracy": 0.25742220133543015,
      "num_tokens": 7686590.0,
      "step": 9640
    },
    {
      "entropy": 4.814498710632324,
      "epoch": 2.7268871925360476,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00043331156842310326,
      "loss": 4.7036,
      "mean_token_accuracy": 0.24966049641370774,
      "num_tokens": 7690618.0,
      "step": 9645
    },
    {
      "entropy": 4.940662527084351,
      "epoch": 2.728300819903873,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004332384602038055,
      "loss": 4.8264,
      "mean_token_accuracy": 0.22878449112176896,
      "num_tokens": 7694441.0,
      "step": 9650
    },
    {
      "entropy": 4.822365045547485,
      "epoch": 2.729714447271699,
      "grad_norm": 1.671875,
      "learning_rate": 0.00043316531891364815,
      "loss": 4.7192,
      "mean_token_accuracy": 0.24304578006267546,
      "num_tokens": 7698133.0,
      "step": 9655
    },
    {
      "entropy": 4.8477997303009035,
      "epoch": 2.731128074639525,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004330921445679172,
      "loss": 4.6667,
      "mean_token_accuracy": 0.24570106863975524,
      "num_tokens": 7701855.0,
      "step": 9660
    },
    {
      "entropy": 4.874637508392334,
      "epoch": 2.732541702007351,
      "grad_norm": 1.578125,
      "learning_rate": 0.00043301893718190556,
      "loss": 4.7267,
      "mean_token_accuracy": 0.2465088412165642,
      "num_tokens": 7705958.0,
      "step": 9665
    },
    {
      "entropy": 4.886174011230469,
      "epoch": 2.733955329375177,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004329456967709132,
      "loss": 4.7637,
      "mean_token_accuracy": 0.23675761967897416,
      "num_tokens": 7710060.0,
      "step": 9670
    },
    {
      "entropy": 4.728029060363769,
      "epoch": 2.7353689567430024,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00043287242335024706,
      "loss": 4.528,
      "mean_token_accuracy": 0.2564690262079239,
      "num_tokens": 7713712.0,
      "step": 9675
    },
    {
      "entropy": 4.835763645172119,
      "epoch": 2.7367825841108284,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00043279911693522055,
      "loss": 4.7739,
      "mean_token_accuracy": 0.24074805825948714,
      "num_tokens": 7717535.0,
      "step": 9680
    },
    {
      "entropy": 4.879166746139527,
      "epoch": 2.7381962114786544,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004327257775411545,
      "loss": 4.7768,
      "mean_token_accuracy": 0.2320738837122917,
      "num_tokens": 7721150.0,
      "step": 9685
    },
    {
      "entropy": 4.879761457443237,
      "epoch": 2.73960983884648,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00043265240518337617,
      "loss": 4.6729,
      "mean_token_accuracy": 0.2460768550634384,
      "num_tokens": 7724983.0,
      "step": 9690
    },
    {
      "entropy": 4.847979784011841,
      "epoch": 2.741023466214306,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00043257899987722017,
      "loss": 4.7744,
      "mean_token_accuracy": 0.2339828610420227,
      "num_tokens": 7729025.0,
      "step": 9695
    },
    {
      "entropy": 4.771919822692871,
      "epoch": 2.7424370935821316,
      "grad_norm": 1.5,
      "learning_rate": 0.00043250556163802756,
      "loss": 4.6374,
      "mean_token_accuracy": 0.24758270829916001,
      "num_tokens": 7732975.0,
      "step": 9700
    },
    {
      "entropy": 4.807431411743164,
      "epoch": 2.7438507209499576,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004324320904811466,
      "loss": 4.5799,
      "mean_token_accuracy": 0.2641636610031128,
      "num_tokens": 7736843.0,
      "step": 9705
    },
    {
      "entropy": 4.79367470741272,
      "epoch": 2.745264348317783,
      "grad_norm": 1.53125,
      "learning_rate": 0.00043235858642193226,
      "loss": 4.7601,
      "mean_token_accuracy": 0.24112373739480972,
      "num_tokens": 7740476.0,
      "step": 9710
    },
    {
      "entropy": 4.923170757293701,
      "epoch": 2.7466779756856092,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043228504947574644,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22731220722198486,
      "num_tokens": 7744464.0,
      "step": 9715
    },
    {
      "entropy": 4.946039247512817,
      "epoch": 2.7480916030534353,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00043221147965795794,
      "loss": 4.8371,
      "mean_token_accuracy": 0.23082001656293868,
      "num_tokens": 7748325.0,
      "step": 9720
    },
    {
      "entropy": 4.909026002883911,
      "epoch": 2.749505230421261,
      "grad_norm": 1.640625,
      "learning_rate": 0.00043213787698394225,
      "loss": 4.8639,
      "mean_token_accuracy": 0.2374993935227394,
      "num_tokens": 7752106.0,
      "step": 9725
    },
    {
      "entropy": 4.947110271453857,
      "epoch": 2.750918857789087,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000432064241469082,
      "loss": 4.763,
      "mean_token_accuracy": 0.24012924879789352,
      "num_tokens": 7755927.0,
      "step": 9730
    },
    {
      "entropy": 4.851477861404419,
      "epoch": 2.752332485156913,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004319905731287666,
      "loss": 4.6823,
      "mean_token_accuracy": 0.24109867364168167,
      "num_tokens": 7759780.0,
      "step": 9735
    },
    {
      "entropy": 4.892245483398438,
      "epoch": 2.7537461125247384,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004319168719783922,
      "loss": 4.6705,
      "mean_token_accuracy": 0.24606094807386397,
      "num_tokens": 7763854.0,
      "step": 9740
    },
    {
      "entropy": 4.844500875473022,
      "epoch": 2.7551597398925645,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00043184313803336187,
      "loss": 4.7846,
      "mean_token_accuracy": 0.24558024108409882,
      "num_tokens": 7767970.0,
      "step": 9745
    },
    {
      "entropy": 4.87809476852417,
      "epoch": 2.75657336726039,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004317693713090856,
      "loss": 4.722,
      "mean_token_accuracy": 0.24603369385004042,
      "num_tokens": 7771790.0,
      "step": 9750
    },
    {
      "entropy": 4.796171760559082,
      "epoch": 2.757986994628216,
      "grad_norm": 1.5625,
      "learning_rate": 0.00043169557182098017,
      "loss": 4.5469,
      "mean_token_accuracy": 0.26736106872558596,
      "num_tokens": 7775591.0,
      "step": 9755
    },
    {
      "entropy": 4.738289833068848,
      "epoch": 2.7594006219960416,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004316217395844692,
      "loss": 4.6404,
      "mean_token_accuracy": 0.2577574968338013,
      "num_tokens": 7779718.0,
      "step": 9760
    },
    {
      "entropy": 4.892912769317627,
      "epoch": 2.7608142493638677,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004315478746149832,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23552078306674956,
      "num_tokens": 7783734.0,
      "step": 9765
    },
    {
      "entropy": 4.914503479003907,
      "epoch": 2.7622278767316937,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004314739769279594,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2313357561826706,
      "num_tokens": 7787556.0,
      "step": 9770
    },
    {
      "entropy": 4.789224624633789,
      "epoch": 2.7636415040995193,
      "grad_norm": 1.625,
      "learning_rate": 0.00043140004653884214,
      "loss": 4.6644,
      "mean_token_accuracy": 0.2588608071208,
      "num_tokens": 7791304.0,
      "step": 9775
    },
    {
      "entropy": 4.7711591720581055,
      "epoch": 2.7650551314673453,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004313260834630822,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24453600943088533,
      "num_tokens": 7795262.0,
      "step": 9780
    },
    {
      "entropy": 4.808416891098022,
      "epoch": 2.7664687588351713,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004312520877161375,
      "loss": 4.7505,
      "mean_token_accuracy": 0.2452472060918808,
      "num_tokens": 7799618.0,
      "step": 9785
    },
    {
      "entropy": 4.832596015930176,
      "epoch": 2.767882386202997,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004311780593134727,
      "loss": 4.7217,
      "mean_token_accuracy": 0.24693237841129304,
      "num_tokens": 7803543.0,
      "step": 9790
    },
    {
      "entropy": 4.8850208759307865,
      "epoch": 2.769296013570823,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00043110399827055943,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23151310831308364,
      "num_tokens": 7807591.0,
      "step": 9795
    },
    {
      "entropy": 4.797076988220215,
      "epoch": 2.7707096409386485,
      "grad_norm": 1.4609375,
      "learning_rate": 0.00043102990460287566,
      "loss": 4.7555,
      "mean_token_accuracy": 0.2424362540245056,
      "num_tokens": 7811621.0,
      "step": 9800
    },
    {
      "entropy": 4.861184406280517,
      "epoch": 2.7721232683064745,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004309557783259068,
      "loss": 4.7687,
      "mean_token_accuracy": 0.2453055277466774,
      "num_tokens": 7815862.0,
      "step": 9805
    },
    {
      "entropy": 4.893250513076782,
      "epoch": 2.7735368956743,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004308816194551446,
      "loss": 4.7192,
      "mean_token_accuracy": 0.23967200815677642,
      "num_tokens": 7819727.0,
      "step": 9810
    },
    {
      "entropy": 4.847082662582397,
      "epoch": 2.774950523042126,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004308074280060879,
      "loss": 4.771,
      "mean_token_accuracy": 0.24513444900512696,
      "num_tokens": 7823488.0,
      "step": 9815
    },
    {
      "entropy": 4.858628368377685,
      "epoch": 2.776364150409952,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004307332039942422,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23987197875976562,
      "num_tokens": 7827767.0,
      "step": 9820
    },
    {
      "entropy": 4.852741003036499,
      "epoch": 2.7777777777777777,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004306589474351199,
      "loss": 4.849,
      "mean_token_accuracy": 0.23759452253580093,
      "num_tokens": 7831897.0,
      "step": 9825
    },
    {
      "entropy": 4.908924913406372,
      "epoch": 2.7791914051456037,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004305846583442402,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24391320645809172,
      "num_tokens": 7836241.0,
      "step": 9830
    },
    {
      "entropy": 4.7802284240722654,
      "epoch": 2.7806050325134297,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000430510336737129,
      "loss": 4.6949,
      "mean_token_accuracy": 0.24995163083076477,
      "num_tokens": 7840285.0,
      "step": 9835
    },
    {
      "entropy": 4.771403169631958,
      "epoch": 2.7820186598812553,
      "grad_norm": 1.671875,
      "learning_rate": 0.00043043598262931907,
      "loss": 4.7183,
      "mean_token_accuracy": 0.24351109564304352,
      "num_tokens": 7844383.0,
      "step": 9840
    },
    {
      "entropy": 4.7551628112792965,
      "epoch": 2.783432287249081,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043036159603635,
      "loss": 4.6418,
      "mean_token_accuracy": 0.2564341828227043,
      "num_tokens": 7848213.0,
      "step": 9845
    },
    {
      "entropy": 4.797671318054199,
      "epoch": 2.784845914616907,
      "grad_norm": 1.59375,
      "learning_rate": 0.00043028717697376795,
      "loss": 4.7696,
      "mean_token_accuracy": 0.23725852817296983,
      "num_tokens": 7852565.0,
      "step": 9850
    },
    {
      "entropy": 4.913282442092895,
      "epoch": 2.786259541984733,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004302127254571263,
      "loss": 4.9275,
      "mean_token_accuracy": 0.22965581119060516,
      "num_tokens": 7856392.0,
      "step": 9855
    },
    {
      "entropy": 4.906764030456543,
      "epoch": 2.7876731693525585,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004301382415019848,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22927335649728775,
      "num_tokens": 7860156.0,
      "step": 9860
    },
    {
      "entropy": 4.881211519241333,
      "epoch": 2.7890867967203845,
      "grad_norm": 1.625,
      "learning_rate": 0.0004300637251239102,
      "loss": 4.8729,
      "mean_token_accuracy": 0.24116809517145157,
      "num_tokens": 7864189.0,
      "step": 9865
    },
    {
      "entropy": 4.9554526805877686,
      "epoch": 2.7905004240882105,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00042998917633847586,
      "loss": 4.8714,
      "mean_token_accuracy": 0.23136020302772523,
      "num_tokens": 7868265.0,
      "step": 9870
    },
    {
      "entropy": 4.863180017471313,
      "epoch": 2.791914051456036,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00042991459516126217,
      "loss": 4.7845,
      "mean_token_accuracy": 0.24003513008356095,
      "num_tokens": 7872150.0,
      "step": 9875
    },
    {
      "entropy": 4.788481521606445,
      "epoch": 2.793327678823862,
      "grad_norm": 1.5390625,
      "learning_rate": 0.000429839981607856,
      "loss": 4.6067,
      "mean_token_accuracy": 0.24923407137393952,
      "num_tokens": 7876161.0,
      "step": 9880
    },
    {
      "entropy": 4.711701965332031,
      "epoch": 2.794741306191688,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00042976533569385116,
      "loss": 4.7291,
      "mean_token_accuracy": 0.24121407121419908,
      "num_tokens": 7879971.0,
      "step": 9885
    },
    {
      "entropy": 4.875331592559815,
      "epoch": 2.7961549335595137,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004296906574348483,
      "loss": 4.7617,
      "mean_token_accuracy": 0.23963514119386672,
      "num_tokens": 7884201.0,
      "step": 9890
    },
    {
      "entropy": 4.848358821868897,
      "epoch": 2.7975685609273393,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004296159468464545,
      "loss": 4.6951,
      "mean_token_accuracy": 0.2363443747162819,
      "num_tokens": 7888344.0,
      "step": 9895
    },
    {
      "entropy": 4.791094446182251,
      "epoch": 2.7989821882951653,
      "grad_norm": 1.703125,
      "learning_rate": 0.000429541203944284,
      "loss": 4.7074,
      "mean_token_accuracy": 0.24345138818025588,
      "num_tokens": 7892173.0,
      "step": 9900
    },
    {
      "entropy": 4.799661445617676,
      "epoch": 2.8003958156629913,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004294664287439575,
      "loss": 4.6776,
      "mean_token_accuracy": 0.2472405791282654,
      "num_tokens": 7896234.0,
      "step": 9905
    },
    {
      "entropy": 4.790255451202393,
      "epoch": 2.801809443030817,
      "grad_norm": 1.8125,
      "learning_rate": 0.00042939162126110256,
      "loss": 4.7136,
      "mean_token_accuracy": 0.24022591710090638,
      "num_tokens": 7900189.0,
      "step": 9910
    },
    {
      "entropy": 4.845273637771607,
      "epoch": 2.803223070398643,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004293167815113535,
      "loss": 4.7976,
      "mean_token_accuracy": 0.2368995040655136,
      "num_tokens": 7904147.0,
      "step": 9915
    },
    {
      "entropy": 4.807591485977173,
      "epoch": 2.804636697766469,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042924190951035144,
      "loss": 4.7297,
      "mean_token_accuracy": 0.24658870995044707,
      "num_tokens": 7908271.0,
      "step": 9920
    },
    {
      "entropy": 4.846813678741455,
      "epoch": 2.8060503251342945,
      "grad_norm": 1.4375,
      "learning_rate": 0.00042916700527374405,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23912189900875092,
      "num_tokens": 7912338.0,
      "step": 9925
    },
    {
      "entropy": 4.83948335647583,
      "epoch": 2.8074639525021206,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004290920688171859,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24049074798822404,
      "num_tokens": 7916449.0,
      "step": 9930
    },
    {
      "entropy": 4.831209468841553,
      "epoch": 2.808877579869946,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00042901710015633826,
      "loss": 4.8148,
      "mean_token_accuracy": 0.23676508367061616,
      "num_tokens": 7920515.0,
      "step": 9935
    },
    {
      "entropy": 4.846360206604004,
      "epoch": 2.810291207237772,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004289420993068692,
      "loss": 4.7143,
      "mean_token_accuracy": 0.24387696087360383,
      "num_tokens": 7924667.0,
      "step": 9940
    },
    {
      "entropy": 4.805326509475708,
      "epoch": 2.8117048346055977,
      "grad_norm": 1.59375,
      "learning_rate": 0.00042886706628445317,
      "loss": 4.6676,
      "mean_token_accuracy": 0.25103714764118196,
      "num_tokens": 7928834.0,
      "step": 9945
    },
    {
      "entropy": 4.792722225189209,
      "epoch": 2.8131184619734237,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004287920011047718,
      "loss": 4.7784,
      "mean_token_accuracy": 0.2328896328806877,
      "num_tokens": 7932799.0,
      "step": 9950
    },
    {
      "entropy": 4.723158502578736,
      "epoch": 2.8145320893412498,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004287169037835132,
      "loss": 4.6754,
      "mean_token_accuracy": 0.24745884537696838,
      "num_tokens": 7936853.0,
      "step": 9955
    },
    {
      "entropy": 4.853486728668213,
      "epoch": 2.8159457167090753,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042864177433637237,
      "loss": 4.7706,
      "mean_token_accuracy": 0.24394648522138596,
      "num_tokens": 7940797.0,
      "step": 9960
    },
    {
      "entropy": 4.901325416564942,
      "epoch": 2.8173593440769014,
      "grad_norm": 1.59375,
      "learning_rate": 0.00042856661277905066,
      "loss": 4.775,
      "mean_token_accuracy": 0.24447948932647706,
      "num_tokens": 7944829.0,
      "step": 9965
    },
    {
      "entropy": 4.80352668762207,
      "epoch": 2.8187729714447274,
      "grad_norm": 1.484375,
      "learning_rate": 0.00042849141912725664,
      "loss": 4.7224,
      "mean_token_accuracy": 0.24101446717977523,
      "num_tokens": 7948885.0,
      "step": 9970
    },
    {
      "entropy": 4.843134355545044,
      "epoch": 2.820186598812553,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004284161933967051,
      "loss": 4.7048,
      "mean_token_accuracy": 0.2578932121396065,
      "num_tokens": 7952999.0,
      "step": 9975
    },
    {
      "entropy": 4.7688788890838625,
      "epoch": 2.821600226180379,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00042834093560311786,
      "loss": 4.6718,
      "mean_token_accuracy": 0.2496064081788063,
      "num_tokens": 7957111.0,
      "step": 9980
    },
    {
      "entropy": 4.752456426620483,
      "epoch": 2.8230138535482046,
      "grad_norm": 1.6875,
      "learning_rate": 0.00042826564576222336,
      "loss": 4.6736,
      "mean_token_accuracy": 0.24942753165960313,
      "num_tokens": 7960834.0,
      "step": 9985
    },
    {
      "entropy": 4.851052951812744,
      "epoch": 2.8244274809160306,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004281903238897566,
      "loss": 4.7563,
      "mean_token_accuracy": 0.24608222246170045,
      "num_tokens": 7964863.0,
      "step": 9990
    },
    {
      "entropy": 4.979444742202759,
      "epoch": 2.825841108283856,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004281149700014594,
      "loss": 4.7754,
      "mean_token_accuracy": 0.24507151842117308,
      "num_tokens": 7968738.0,
      "step": 9995
    },
    {
      "entropy": 4.88283863067627,
      "epoch": 2.827254735651682,
      "grad_norm": 1.6875,
      "learning_rate": 0.00042803958411308045,
      "loss": 4.6658,
      "mean_token_accuracy": 0.2468786984682083,
      "num_tokens": 7972967.0,
      "step": 10000
    },
    {
      "epoch": 2.827254735651682,
      "eval_entropy": 4.713655055794287,
      "eval_loss": 5.1287360191345215,
      "eval_mean_token_accuracy": 0.23578898683137542,
      "eval_num_tokens": 7972967.0,
      "eval_runtime": 6.6843,
      "eval_samples_per_second": 1688.123,
      "eval_steps_per_second": 211.09,
      "step": 10000
    },
    {
      "entropy": 4.791499376296997,
      "epoch": 2.828668363019508,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004279641662403746,
      "loss": 4.6846,
      "mean_token_accuracy": 0.23586008399724961,
      "num_tokens": 7976806.0,
      "step": 10005
    },
    {
      "entropy": 4.834108829498291,
      "epoch": 2.8300819903873338,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004278887163991042,
      "loss": 4.9083,
      "mean_token_accuracy": 0.23424376547336578,
      "num_tokens": 7980692.0,
      "step": 10010
    },
    {
      "entropy": 4.818598508834839,
      "epoch": 2.83149561775516,
      "grad_norm": 1.625,
      "learning_rate": 0.00042781323460503725,
      "loss": 4.6954,
      "mean_token_accuracy": 0.244588203728199,
      "num_tokens": 7984629.0,
      "step": 10015
    },
    {
      "entropy": 4.881340456008911,
      "epoch": 2.832909245122986,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004277377208739494,
      "loss": 4.7841,
      "mean_token_accuracy": 0.2400134339928627,
      "num_tokens": 7988624.0,
      "step": 10020
    },
    {
      "entropy": 4.8173987865448,
      "epoch": 2.8343228724908114,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042766217522162225,
      "loss": 4.7121,
      "mean_token_accuracy": 0.25045938044786453,
      "num_tokens": 7992157.0,
      "step": 10025
    },
    {
      "entropy": 4.842756938934326,
      "epoch": 2.8357364998586374,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00042758659766384456,
      "loss": 4.735,
      "mean_token_accuracy": 0.2481465756893158,
      "num_tokens": 7996388.0,
      "step": 10030
    },
    {
      "entropy": 4.741224718093872,
      "epoch": 2.837150127226463,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004275109882164115,
      "loss": 4.6726,
      "mean_token_accuracy": 0.24208081215620042,
      "num_tokens": 8000341.0,
      "step": 10035
    },
    {
      "entropy": 4.7802205085754395,
      "epoch": 2.838563754594289,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004274353468951251,
      "loss": 4.7207,
      "mean_token_accuracy": 0.24618931114673615,
      "num_tokens": 8004656.0,
      "step": 10040
    },
    {
      "entropy": 4.84431414604187,
      "epoch": 2.8399773819621146,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004273596737157937,
      "loss": 4.7238,
      "mean_token_accuracy": 0.24430040270090103,
      "num_tokens": 8008548.0,
      "step": 10045
    },
    {
      "entropy": 4.925453567504883,
      "epoch": 2.8413910093299406,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004272839686942327,
      "loss": 4.8128,
      "mean_token_accuracy": 0.24114488065242767,
      "num_tokens": 8012619.0,
      "step": 10050
    },
    {
      "entropy": 4.878169012069702,
      "epoch": 2.8428046366977666,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00042720823184626397,
      "loss": 4.81,
      "mean_token_accuracy": 0.23297590762376785,
      "num_tokens": 8016741.0,
      "step": 10055
    },
    {
      "entropy": 4.842193984985352,
      "epoch": 2.844218264065592,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00042713246318771604,
      "loss": 4.6692,
      "mean_token_accuracy": 0.2592435419559479,
      "num_tokens": 8020790.0,
      "step": 10060
    },
    {
      "entropy": 4.832191753387451,
      "epoch": 2.845631891433418,
      "grad_norm": 1.546875,
      "learning_rate": 0.000427056662734424,
      "loss": 4.6021,
      "mean_token_accuracy": 0.25804333537817004,
      "num_tokens": 8024678.0,
      "step": 10065
    },
    {
      "entropy": 4.7375494003295895,
      "epoch": 2.8470455188012442,
      "grad_norm": 1.8125,
      "learning_rate": 0.00042698083050222974,
      "loss": 4.7929,
      "mean_token_accuracy": 0.235735422372818,
      "num_tokens": 8028986.0,
      "step": 10070
    },
    {
      "entropy": 4.818836355209351,
      "epoch": 2.84845914616907,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00042690496650698174,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2336840346455574,
      "num_tokens": 8033247.0,
      "step": 10075
    },
    {
      "entropy": 4.859717988967896,
      "epoch": 2.849872773536896,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042682907076453506,
      "loss": 4.7757,
      "mean_token_accuracy": 0.24476199597120285,
      "num_tokens": 8036916.0,
      "step": 10080
    },
    {
      "entropy": 4.842935514450073,
      "epoch": 2.8512864009047214,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00042675314329075154,
      "loss": 4.6655,
      "mean_token_accuracy": 0.25364025980234145,
      "num_tokens": 8040934.0,
      "step": 10085
    },
    {
      "entropy": 4.784061765670776,
      "epoch": 2.8527000282725474,
      "grad_norm": 1.625,
      "learning_rate": 0.00042667718410149954,
      "loss": 4.791,
      "mean_token_accuracy": 0.24379255920648574,
      "num_tokens": 8044799.0,
      "step": 10090
    },
    {
      "entropy": 4.805581903457641,
      "epoch": 2.854113655640373,
      "grad_norm": 1.609375,
      "learning_rate": 0.00042660119321265386,
      "loss": 4.7404,
      "mean_token_accuracy": 0.2482749417424202,
      "num_tokens": 8048643.0,
      "step": 10095
    },
    {
      "entropy": 4.809217262268066,
      "epoch": 2.855527283008199,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004265251706400963,
      "loss": 4.6577,
      "mean_token_accuracy": 0.2491634413599968,
      "num_tokens": 8052273.0,
      "step": 10100
    },
    {
      "entropy": 4.815245151519775,
      "epoch": 2.856940910376025,
      "grad_norm": 1.515625,
      "learning_rate": 0.00042644911639971514,
      "loss": 4.7102,
      "mean_token_accuracy": 0.2435444638133049,
      "num_tokens": 8056566.0,
      "step": 10105
    },
    {
      "entropy": 4.837096881866455,
      "epoch": 2.8583545377438506,
      "grad_norm": 1.625,
      "learning_rate": 0.00042637303050740516,
      "loss": 4.8198,
      "mean_token_accuracy": 0.24333268851041795,
      "num_tokens": 8060397.0,
      "step": 10110
    },
    {
      "entropy": 4.844837331771851,
      "epoch": 2.8597681651116766,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004262969129790679,
      "loss": 4.7465,
      "mean_token_accuracy": 0.2388468101620674,
      "num_tokens": 8064258.0,
      "step": 10115
    },
    {
      "entropy": 4.754597091674805,
      "epoch": 2.8611817924795027,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004262207638306114,
      "loss": 4.7455,
      "mean_token_accuracy": 0.25245089530944825,
      "num_tokens": 8068222.0,
      "step": 10120
    },
    {
      "entropy": 4.849435329437256,
      "epoch": 2.8625954198473282,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004261445830779503,
      "loss": 4.7536,
      "mean_token_accuracy": 0.2443015992641449,
      "num_tokens": 8072155.0,
      "step": 10125
    },
    {
      "entropy": 4.841937017440796,
      "epoch": 2.864009047215154,
      "grad_norm": 1.703125,
      "learning_rate": 0.000426068370737006,
      "loss": 4.6912,
      "mean_token_accuracy": 0.24920135885477065,
      "num_tokens": 8076172.0,
      "step": 10130
    },
    {
      "entropy": 4.822752285003662,
      "epoch": 2.86542267458298,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004259921268237065,
      "loss": 4.7564,
      "mean_token_accuracy": 0.24322517812252045,
      "num_tokens": 8080215.0,
      "step": 10135
    },
    {
      "entropy": 4.702842950820923,
      "epoch": 2.866836301950806,
      "grad_norm": 1.5,
      "learning_rate": 0.0004259158513539862,
      "loss": 4.6584,
      "mean_token_accuracy": 0.2641510725021362,
      "num_tokens": 8084313.0,
      "step": 10140
    },
    {
      "entropy": 4.837695503234864,
      "epoch": 2.8682499293186314,
      "grad_norm": 1.609375,
      "learning_rate": 0.00042583954434378615,
      "loss": 4.7853,
      "mean_token_accuracy": 0.24077674746513367,
      "num_tokens": 8088356.0,
      "step": 10145
    },
    {
      "entropy": 4.8694051742553714,
      "epoch": 2.8696635566864575,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004257632058090541,
      "loss": 4.8854,
      "mean_token_accuracy": 0.23105292320251464,
      "num_tokens": 8092569.0,
      "step": 10150
    },
    {
      "entropy": 4.847645664215088,
      "epoch": 2.8710771840542835,
      "grad_norm": 1.5,
      "learning_rate": 0.0004256868357657443,
      "loss": 4.692,
      "mean_token_accuracy": 0.24674782156944275,
      "num_tokens": 8096849.0,
      "step": 10155
    },
    {
      "entropy": 4.868953227996826,
      "epoch": 2.872490811422109,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00042561043422981767,
      "loss": 4.7119,
      "mean_token_accuracy": 0.2560817450284958,
      "num_tokens": 8101000.0,
      "step": 10160
    },
    {
      "entropy": 4.824372720718384,
      "epoch": 2.873904438789935,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004255340012172417,
      "loss": 4.7474,
      "mean_token_accuracy": 0.24388387352228164,
      "num_tokens": 8104844.0,
      "step": 10165
    },
    {
      "entropy": 4.800144815444947,
      "epoch": 2.875318066157761,
      "grad_norm": 1.65625,
      "learning_rate": 0.00042545753674399025,
      "loss": 4.6628,
      "mean_token_accuracy": 0.2501363083720207,
      "num_tokens": 8108788.0,
      "step": 10170
    },
    {
      "entropy": 4.886746788024903,
      "epoch": 2.8767316935255867,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000425381040826044,
      "loss": 4.8685,
      "mean_token_accuracy": 0.23293202370405197,
      "num_tokens": 8112700.0,
      "step": 10175
    },
    {
      "entropy": 4.881789255142212,
      "epoch": 2.8781453208934122,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00042530451347939024,
      "loss": 4.7069,
      "mean_token_accuracy": 0.2472312793135643,
      "num_tokens": 8116647.0,
      "step": 10180
    },
    {
      "entropy": 4.776129055023193,
      "epoch": 2.8795589482612383,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00042522795472002263,
      "loss": 4.697,
      "mean_token_accuracy": 0.25197477638721466,
      "num_tokens": 8120840.0,
      "step": 10185
    },
    {
      "entropy": 4.822296190261841,
      "epoch": 2.8809725756290643,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042515136456394134,
      "loss": 4.6831,
      "mean_token_accuracy": 0.24778514355421066,
      "num_tokens": 8124773.0,
      "step": 10190
    },
    {
      "entropy": 4.894076204299926,
      "epoch": 2.88238620299689,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004250747430271535,
      "loss": 4.7059,
      "mean_token_accuracy": 0.23883723318576813,
      "num_tokens": 8128596.0,
      "step": 10195
    },
    {
      "entropy": 4.781049203872681,
      "epoch": 2.883799830364716,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00042499809012567233,
      "loss": 4.8095,
      "mean_token_accuracy": 0.24575326591730118,
      "num_tokens": 8132596.0,
      "step": 10200
    },
    {
      "entropy": 4.785203266143799,
      "epoch": 2.885213457732542,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00042492140587551785,
      "loss": 4.7325,
      "mean_token_accuracy": 0.23930976688861846,
      "num_tokens": 8136384.0,
      "step": 10205
    },
    {
      "entropy": 4.89449257850647,
      "epoch": 2.8866270851003675,
      "grad_norm": 1.453125,
      "learning_rate": 0.00042484469029271666,
      "loss": 4.7809,
      "mean_token_accuracy": 0.2386889025568962,
      "num_tokens": 8140348.0,
      "step": 10210
    },
    {
      "entropy": 4.879277038574219,
      "epoch": 2.8880407124681935,
      "grad_norm": 1.609375,
      "learning_rate": 0.00042476794339330184,
      "loss": 4.7434,
      "mean_token_accuracy": 0.2437808558344841,
      "num_tokens": 8144248.0,
      "step": 10215
    },
    {
      "entropy": 4.842742872238159,
      "epoch": 2.889454339836019,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00042469116519331296,
      "loss": 4.6779,
      "mean_token_accuracy": 0.25188122540712354,
      "num_tokens": 8148035.0,
      "step": 10220
    },
    {
      "entropy": 4.91721396446228,
      "epoch": 2.890867967203845,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00042461435570879626,
      "loss": 4.8627,
      "mean_token_accuracy": 0.23223422467708588,
      "num_tokens": 8151830.0,
      "step": 10225
    },
    {
      "entropy": 4.837510299682617,
      "epoch": 2.8922815945716707,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00042453751495580435,
      "loss": 4.691,
      "mean_token_accuracy": 0.24130447655916215,
      "num_tokens": 8155974.0,
      "step": 10230
    },
    {
      "entropy": 4.811863613128662,
      "epoch": 2.8936952219394967,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004244606429503965,
      "loss": 4.7287,
      "mean_token_accuracy": 0.23917650580406188,
      "num_tokens": 8159954.0,
      "step": 10235
    },
    {
      "entropy": 4.862403917312622,
      "epoch": 2.8951088493073227,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00042438373970863865,
      "loss": 4.7148,
      "mean_token_accuracy": 0.24001158624887467,
      "num_tokens": 8163736.0,
      "step": 10240
    },
    {
      "entropy": 4.907062816619873,
      "epoch": 2.8965224766751483,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004243068052466028,
      "loss": 4.7334,
      "mean_token_accuracy": 0.2359981805086136,
      "num_tokens": 8167579.0,
      "step": 10245
    },
    {
      "entropy": 4.924914836883545,
      "epoch": 2.8979361040429743,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00042422983958036804,
      "loss": 4.7882,
      "mean_token_accuracy": 0.2316145718097687,
      "num_tokens": 8171521.0,
      "step": 10250
    },
    {
      "entropy": 4.778305768966675,
      "epoch": 2.8993497314108003,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042415284272601957,
      "loss": 4.6939,
      "mean_token_accuracy": 0.24451784342527388,
      "num_tokens": 8175688.0,
      "step": 10255
    },
    {
      "entropy": 4.773069047927857,
      "epoch": 2.900763358778626,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004240758146996493,
      "loss": 4.6612,
      "mean_token_accuracy": 0.24789104014635086,
      "num_tokens": 8179571.0,
      "step": 10260
    },
    {
      "entropy": 4.986332082748413,
      "epoch": 2.902176986146452,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004239987555173557,
      "loss": 4.8446,
      "mean_token_accuracy": 0.22365877777338028,
      "num_tokens": 8183665.0,
      "step": 10265
    },
    {
      "entropy": 4.845275068283081,
      "epoch": 2.9035906135142775,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004239216651952435,
      "loss": 4.746,
      "mean_token_accuracy": 0.23671557605266572,
      "num_tokens": 8187807.0,
      "step": 10270
    },
    {
      "entropy": 4.825612926483155,
      "epoch": 2.9050042408821035,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00042384454374942417,
      "loss": 4.6408,
      "mean_token_accuracy": 0.24168917834758757,
      "num_tokens": 8191908.0,
      "step": 10275
    },
    {
      "entropy": 4.823938655853271,
      "epoch": 2.906417868249929,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042376739119601565,
      "loss": 4.6593,
      "mean_token_accuracy": 0.25081279277801516,
      "num_tokens": 8195909.0,
      "step": 10280
    },
    {
      "entropy": 4.882490158081055,
      "epoch": 2.907831495617755,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004236902075511423,
      "loss": 4.7833,
      "mean_token_accuracy": 0.2404140517115593,
      "num_tokens": 8199443.0,
      "step": 10285
    },
    {
      "entropy": 4.7079273700714115,
      "epoch": 2.909245122985581,
      "grad_norm": 1.640625,
      "learning_rate": 0.000423612992830935,
      "loss": 4.6428,
      "mean_token_accuracy": 0.2514503836631775,
      "num_tokens": 8203448.0,
      "step": 10290
    },
    {
      "entropy": 4.7486096858978275,
      "epoch": 2.9106587503534067,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004235357470515313,
      "loss": 4.7465,
      "mean_token_accuracy": 0.23638797849416732,
      "num_tokens": 8207163.0,
      "step": 10295
    },
    {
      "entropy": 4.842561292648315,
      "epoch": 2.9120723777212327,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00042345847022907495,
      "loss": 4.7035,
      "mean_token_accuracy": 0.24088269025087355,
      "num_tokens": 8210957.0,
      "step": 10300
    },
    {
      "entropy": 4.866099119186401,
      "epoch": 2.9134860050890588,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00042338116237971635,
      "loss": 4.6168,
      "mean_token_accuracy": 0.25678325444459915,
      "num_tokens": 8214555.0,
      "step": 10305
    },
    {
      "entropy": 4.781775903701782,
      "epoch": 2.9148996324568843,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042330382351961235,
      "loss": 4.6205,
      "mean_token_accuracy": 0.2583231136202812,
      "num_tokens": 8218441.0,
      "step": 10310
    },
    {
      "entropy": 4.697985029220581,
      "epoch": 2.9163132598247103,
      "grad_norm": 1.453125,
      "learning_rate": 0.00042322645366492633,
      "loss": 4.5484,
      "mean_token_accuracy": 0.2632806867361069,
      "num_tokens": 8222383.0,
      "step": 10315
    },
    {
      "entropy": 4.873652791976928,
      "epoch": 2.917726887192536,
      "grad_norm": 1.515625,
      "learning_rate": 0.00042314905283182816,
      "loss": 4.7843,
      "mean_token_accuracy": 0.232602159678936,
      "num_tokens": 8226282.0,
      "step": 10320
    },
    {
      "entropy": 4.925016736984253,
      "epoch": 2.919140514560362,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000423071621036494,
      "loss": 4.7029,
      "mean_token_accuracy": 0.25488694608211515,
      "num_tokens": 8230241.0,
      "step": 10325
    },
    {
      "entropy": 4.913437461853027,
      "epoch": 2.9205541419281875,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004229941582951067,
      "loss": 4.731,
      "mean_token_accuracy": 0.24249832779169084,
      "num_tokens": 8233997.0,
      "step": 10330
    },
    {
      "entropy": 4.884800100326538,
      "epoch": 2.9219677692960135,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004229166646238556,
      "loss": 4.7281,
      "mean_token_accuracy": 0.24566900730133057,
      "num_tokens": 8237873.0,
      "step": 10335
    },
    {
      "entropy": 4.912526798248291,
      "epoch": 2.9233813966638396,
      "grad_norm": 1.53125,
      "learning_rate": 0.000422839140038936,
      "loss": 4.7141,
      "mean_token_accuracy": 0.2488257944583893,
      "num_tokens": 8241687.0,
      "step": 10340
    },
    {
      "entropy": 4.804954147338867,
      "epoch": 2.924795024031665,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042276158455655064,
      "loss": 4.7887,
      "mean_token_accuracy": 0.2356165036559105,
      "num_tokens": 8245642.0,
      "step": 10345
    },
    {
      "entropy": 4.817019414901734,
      "epoch": 2.926208651399491,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004226839981929077,
      "loss": 4.6233,
      "mean_token_accuracy": 0.25944034606218336,
      "num_tokens": 8249648.0,
      "step": 10350
    },
    {
      "entropy": 4.821104049682617,
      "epoch": 2.927622278767317,
      "grad_norm": 1.515625,
      "learning_rate": 0.00042260638096422237,
      "loss": 4.7394,
      "mean_token_accuracy": 0.24545091539621353,
      "num_tokens": 8253722.0,
      "step": 10355
    },
    {
      "entropy": 4.89226016998291,
      "epoch": 2.9290359061351428,
      "grad_norm": 1.53125,
      "learning_rate": 0.00042252873288671624,
      "loss": 4.8353,
      "mean_token_accuracy": 0.23633091002702714,
      "num_tokens": 8258161.0,
      "step": 10360
    },
    {
      "entropy": 4.744753217697143,
      "epoch": 2.9304495335029688,
      "grad_norm": 1.5625,
      "learning_rate": 0.00042245105397661724,
      "loss": 4.6256,
      "mean_token_accuracy": 0.24733325093984604,
      "num_tokens": 8262015.0,
      "step": 10365
    },
    {
      "entropy": 4.925482368469238,
      "epoch": 2.9318631608707943,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004223733442501597,
      "loss": 4.8789,
      "mean_token_accuracy": 0.21885365545749663,
      "num_tokens": 8266002.0,
      "step": 10370
    },
    {
      "entropy": 4.8145904541015625,
      "epoch": 2.9332767882386204,
      "grad_norm": 1.625,
      "learning_rate": 0.00042229560372358463,
      "loss": 4.7379,
      "mean_token_accuracy": 0.23693596571683884,
      "num_tokens": 8270125.0,
      "step": 10375
    },
    {
      "entropy": 4.876845598220825,
      "epoch": 2.934690415606446,
      "grad_norm": 1.65625,
      "learning_rate": 0.00042221783241313914,
      "loss": 4.7602,
      "mean_token_accuracy": 0.24229519665241242,
      "num_tokens": 8273686.0,
      "step": 10380
    },
    {
      "entropy": 4.847682189941406,
      "epoch": 2.936104042974272,
      "grad_norm": 1.703125,
      "learning_rate": 0.00042214003033507716,
      "loss": 4.7073,
      "mean_token_accuracy": 0.24081776440143585,
      "num_tokens": 8277499.0,
      "step": 10385
    },
    {
      "entropy": 4.8276176929473875,
      "epoch": 2.937517670342098,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004220621975056586,
      "loss": 4.7609,
      "mean_token_accuracy": 0.24786525070667267,
      "num_tokens": 8281947.0,
      "step": 10390
    },
    {
      "entropy": 4.759427928924561,
      "epoch": 2.9389312977099236,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004219843339411502,
      "loss": 4.7559,
      "mean_token_accuracy": 0.24141957610845566,
      "num_tokens": 8285869.0,
      "step": 10395
    },
    {
      "entropy": 4.8184184551239015,
      "epoch": 2.9403449250777496,
      "grad_norm": 1.609375,
      "learning_rate": 0.00042190643965782495,
      "loss": 4.6956,
      "mean_token_accuracy": 0.24050530344247817,
      "num_tokens": 8289754.0,
      "step": 10400
    },
    {
      "entropy": 4.826197814941406,
      "epoch": 2.9417585524455756,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00042182851467196223,
      "loss": 4.7916,
      "mean_token_accuracy": 0.2393219366669655,
      "num_tokens": 8293424.0,
      "step": 10405
    },
    {
      "entropy": 4.896547460556031,
      "epoch": 2.943172179813401,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00042175055899984796,
      "loss": 4.7438,
      "mean_token_accuracy": 0.24250219762325287,
      "num_tokens": 8297015.0,
      "step": 10410
    },
    {
      "entropy": 4.696416902542114,
      "epoch": 2.9445858071812268,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004216725726577744,
      "loss": 4.6754,
      "mean_token_accuracy": 0.2567340239882469,
      "num_tokens": 8300946.0,
      "step": 10415
    },
    {
      "entropy": 4.803910636901856,
      "epoch": 2.9459994345490528,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004215945556620399,
      "loss": 4.7447,
      "mean_token_accuracy": 0.24302055984735488,
      "num_tokens": 8305373.0,
      "step": 10420
    },
    {
      "entropy": 4.913315963745117,
      "epoch": 2.947413061916879,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042151650802895005,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23596266955137252,
      "num_tokens": 8309336.0,
      "step": 10425
    },
    {
      "entropy": 4.928888893127441,
      "epoch": 2.9488266892847044,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00042143842977481583,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23431365191936493,
      "num_tokens": 8313302.0,
      "step": 10430
    },
    {
      "entropy": 4.675397205352783,
      "epoch": 2.9502403166525304,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004213603209159555,
      "loss": 4.6172,
      "mean_token_accuracy": 0.2598820000886917,
      "num_tokens": 8317193.0,
      "step": 10435
    },
    {
      "entropy": 4.828561496734619,
      "epoch": 2.9516539440203564,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004212821814686931,
      "loss": 4.737,
      "mean_token_accuracy": 0.2430539086461067,
      "num_tokens": 8321139.0,
      "step": 10440
    },
    {
      "entropy": 4.92603759765625,
      "epoch": 2.953067571388182,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004212040114493593,
      "loss": 4.7423,
      "mean_token_accuracy": 0.24312940537929534,
      "num_tokens": 8324862.0,
      "step": 10445
    },
    {
      "entropy": 4.864740610122681,
      "epoch": 2.954481198756008,
      "grad_norm": 1.53125,
      "learning_rate": 0.00042112581087429133,
      "loss": 4.7131,
      "mean_token_accuracy": 0.23532480746507645,
      "num_tokens": 8329219.0,
      "step": 10450
    },
    {
      "entropy": 4.829734754562378,
      "epoch": 2.955894826123834,
      "grad_norm": 1.484375,
      "learning_rate": 0.00042104757975983245,
      "loss": 4.7774,
      "mean_token_accuracy": 0.24235375821590424,
      "num_tokens": 8332969.0,
      "step": 10455
    },
    {
      "entropy": 4.7843772888183596,
      "epoch": 2.9573084534916596,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004209693181223325,
      "loss": 4.7558,
      "mean_token_accuracy": 0.24762057960033418,
      "num_tokens": 8336976.0,
      "step": 10460
    },
    {
      "entropy": 4.863378858566284,
      "epoch": 2.958722080859485,
      "grad_norm": 1.578125,
      "learning_rate": 0.00042089102597814775,
      "loss": 4.6753,
      "mean_token_accuracy": 0.2535947233438492,
      "num_tokens": 8340982.0,
      "step": 10465
    },
    {
      "entropy": 4.90426230430603,
      "epoch": 2.960135708227311,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004208127033436407,
      "loss": 4.7118,
      "mean_token_accuracy": 0.24618061631917953,
      "num_tokens": 8344860.0,
      "step": 10470
    },
    {
      "entropy": 4.866394996643066,
      "epoch": 2.9615493355951372,
      "grad_norm": 1.515625,
      "learning_rate": 0.00042073435023518035,
      "loss": 4.7423,
      "mean_token_accuracy": 0.25223354548215865,
      "num_tokens": 8348926.0,
      "step": 10475
    },
    {
      "entropy": 4.731988716125488,
      "epoch": 2.962962962962963,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00042065596666914195,
      "loss": 4.649,
      "mean_token_accuracy": 0.25639946609735487,
      "num_tokens": 8353083.0,
      "step": 10480
    },
    {
      "entropy": 4.852592134475708,
      "epoch": 2.964376590330789,
      "grad_norm": 1.5,
      "learning_rate": 0.00042057755266190734,
      "loss": 4.8086,
      "mean_token_accuracy": 0.239181487262249,
      "num_tokens": 8357627.0,
      "step": 10485
    },
    {
      "entropy": 4.880895709991455,
      "epoch": 2.965790217698615,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004204991082298643,
      "loss": 4.7436,
      "mean_token_accuracy": 0.23376303613185884,
      "num_tokens": 8361492.0,
      "step": 10490
    },
    {
      "entropy": 4.82518310546875,
      "epoch": 2.9672038450664404,
      "grad_norm": 1.6875,
      "learning_rate": 0.00042042063338940745,
      "loss": 4.6876,
      "mean_token_accuracy": 0.24758541136980056,
      "num_tokens": 8365587.0,
      "step": 10495
    },
    {
      "entropy": 4.773739862442016,
      "epoch": 2.9686174724342664,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004203421281569375,
      "loss": 4.7919,
      "mean_token_accuracy": 0.24491761177778243,
      "num_tokens": 8369460.0,
      "step": 10500
    },
    {
      "entropy": 4.799633741378784,
      "epoch": 2.970031099802092,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00042026359254886147,
      "loss": 4.7061,
      "mean_token_accuracy": 0.2564088016748428,
      "num_tokens": 8373394.0,
      "step": 10505
    },
    {
      "entropy": 4.804653882980347,
      "epoch": 2.971444727169918,
      "grad_norm": 1.703125,
      "learning_rate": 0.00042018502658159294,
      "loss": 4.6556,
      "mean_token_accuracy": 0.247689288854599,
      "num_tokens": 8377278.0,
      "step": 10510
    },
    {
      "entropy": 4.993567276000976,
      "epoch": 2.9728583545377436,
      "grad_norm": 1.4453125,
      "learning_rate": 0.00042010643027155163,
      "loss": 4.883,
      "mean_token_accuracy": 0.22488299161195754,
      "num_tokens": 8381335.0,
      "step": 10515
    },
    {
      "entropy": 4.885659646987915,
      "epoch": 2.9742719819055696,
      "grad_norm": 1.5,
      "learning_rate": 0.0004200278036351638,
      "loss": 4.7509,
      "mean_token_accuracy": 0.24135842323303222,
      "num_tokens": 8385234.0,
      "step": 10520
    },
    {
      "entropy": 4.821668672561645,
      "epoch": 2.9756856092733956,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00041994914668886166,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23865493386983871,
      "num_tokens": 8389656.0,
      "step": 10525
    },
    {
      "entropy": 4.846656465530396,
      "epoch": 2.9770992366412212,
      "grad_norm": 1.609375,
      "learning_rate": 0.00041987045944908433,
      "loss": 4.7669,
      "mean_token_accuracy": 0.2391110360622406,
      "num_tokens": 8393484.0,
      "step": 10530
    },
    {
      "entropy": 4.859679651260376,
      "epoch": 2.9785128640090472,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00041979174193227677,
      "loss": 4.7099,
      "mean_token_accuracy": 0.24288977235555648,
      "num_tokens": 8397672.0,
      "step": 10535
    },
    {
      "entropy": 4.8233551502227785,
      "epoch": 2.9799264913768733,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004197129941548905,
      "loss": 4.7411,
      "mean_token_accuracy": 0.24574666023254393,
      "num_tokens": 8401607.0,
      "step": 10540
    },
    {
      "entropy": 4.862898826599121,
      "epoch": 2.981340118744699,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004196342161333833,
      "loss": 4.8346,
      "mean_token_accuracy": 0.22908495664596557,
      "num_tokens": 8406060.0,
      "step": 10545
    },
    {
      "entropy": 4.927570390701294,
      "epoch": 2.982753746112525,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00041955540788421946,
      "loss": 4.7911,
      "mean_token_accuracy": 0.23549697548151016,
      "num_tokens": 8410098.0,
      "step": 10550
    },
    {
      "entropy": 4.908298587799072,
      "epoch": 2.9841673734803504,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004194765694238691,
      "loss": 4.7508,
      "mean_token_accuracy": 0.24171848744153976,
      "num_tokens": 8414306.0,
      "step": 10555
    },
    {
      "entropy": 4.8168951034545895,
      "epoch": 2.9855810008481765,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041939770076880936,
      "loss": 4.7473,
      "mean_token_accuracy": 0.24014342725276946,
      "num_tokens": 8418204.0,
      "step": 10560
    },
    {
      "entropy": 4.806839513778686,
      "epoch": 2.986994628216002,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004193188019355229,
      "loss": 4.7648,
      "mean_token_accuracy": 0.23602302074432374,
      "num_tokens": 8421946.0,
      "step": 10565
    },
    {
      "entropy": 4.915254974365235,
      "epoch": 2.988408255583828,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004192398729404993,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22657402455806733,
      "num_tokens": 8426087.0,
      "step": 10570
    },
    {
      "entropy": 4.871234655380249,
      "epoch": 2.989821882951654,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041916091380023423,
      "loss": 4.7748,
      "mean_token_accuracy": 0.2371543049812317,
      "num_tokens": 8430145.0,
      "step": 10575
    },
    {
      "entropy": 4.790779495239258,
      "epoch": 2.9912355103194797,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004190819245312297,
      "loss": 4.6584,
      "mean_token_accuracy": 0.24890362322330475,
      "num_tokens": 8434134.0,
      "step": 10580
    },
    {
      "entropy": 4.840692377090454,
      "epoch": 2.9926491376873057,
      "grad_norm": 1.609375,
      "learning_rate": 0.00041900290514999383,
      "loss": 4.753,
      "mean_token_accuracy": 0.2340681731700897,
      "num_tokens": 8438464.0,
      "step": 10585
    },
    {
      "entropy": 4.736099433898926,
      "epoch": 2.9940627650551317,
      "grad_norm": 1.546875,
      "learning_rate": 0.00041892385567304135,
      "loss": 4.7518,
      "mean_token_accuracy": 0.24206361174583435,
      "num_tokens": 8442391.0,
      "step": 10590
    },
    {
      "entropy": 4.828328609466553,
      "epoch": 2.9954763924229573,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041884477611689295,
      "loss": 4.775,
      "mean_token_accuracy": 0.23924427032470702,
      "num_tokens": 8446454.0,
      "step": 10595
    },
    {
      "entropy": 4.760121488571167,
      "epoch": 2.9968900197907833,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004187656664980759,
      "loss": 4.6708,
      "mean_token_accuracy": 0.2564073607325554,
      "num_tokens": 8450444.0,
      "step": 10600
    },
    {
      "entropy": 4.913830232620239,
      "epoch": 2.998303647158609,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004186865268331236,
      "loss": 4.8243,
      "mean_token_accuracy": 0.23767312318086625,
      "num_tokens": 8454278.0,
      "step": 10605
    },
    {
      "entropy": 4.936022472381592,
      "epoch": 2.999717274526435,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004186073571385757,
      "loss": 4.8682,
      "mean_token_accuracy": 0.23375980108976363,
      "num_tokens": 8458573.0,
      "step": 10610
    },
    {
      "entropy": 4.862014675140381,
      "epoch": 3.0011309018942605,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00041852815743097824,
      "loss": 4.6321,
      "mean_token_accuracy": 0.24318708926439286,
      "num_tokens": 8462460.0,
      "step": 10615
    },
    {
      "entropy": 4.832936334609985,
      "epoch": 3.0025445292620865,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004184489277268833,
      "loss": 4.5435,
      "mean_token_accuracy": 0.2505181685090065,
      "num_tokens": 8466363.0,
      "step": 10620
    },
    {
      "entropy": 4.830978679656982,
      "epoch": 3.0039581566299125,
      "grad_norm": 1.515625,
      "learning_rate": 0.00041836966804284964,
      "loss": 4.5371,
      "mean_token_accuracy": 0.25248042345046995,
      "num_tokens": 8470425.0,
      "step": 10625
    },
    {
      "entropy": 4.8899750232696535,
      "epoch": 3.005371783997738,
      "grad_norm": 1.515625,
      "learning_rate": 0.00041829037839544186,
      "loss": 4.6103,
      "mean_token_accuracy": 0.2373036339879036,
      "num_tokens": 8474472.0,
      "step": 10630
    },
    {
      "entropy": 4.869209909439087,
      "epoch": 3.006785411365564,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004182110588012312,
      "loss": 4.4566,
      "mean_token_accuracy": 0.2491020068526268,
      "num_tokens": 8478078.0,
      "step": 10635
    },
    {
      "entropy": 4.740761089324951,
      "epoch": 3.0081990387333897,
      "grad_norm": 1.6875,
      "learning_rate": 0.00041813170927679466,
      "loss": 4.372,
      "mean_token_accuracy": 0.2691865563392639,
      "num_tokens": 8481979.0,
      "step": 10640
    },
    {
      "entropy": 4.76536717414856,
      "epoch": 3.0096126661012157,
      "grad_norm": 1.484375,
      "learning_rate": 0.00041805232983871605,
      "loss": 4.556,
      "mean_token_accuracy": 0.2540141403675079,
      "num_tokens": 8486125.0,
      "step": 10645
    },
    {
      "entropy": 4.759019374847412,
      "epoch": 3.0110262934690417,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004179729205035852,
      "loss": 4.5323,
      "mean_token_accuracy": 0.24392110407352446,
      "num_tokens": 8489837.0,
      "step": 10650
    },
    {
      "entropy": 4.802046775817871,
      "epoch": 3.0124399208368673,
      "grad_norm": 1.703125,
      "learning_rate": 0.000417893481287998,
      "loss": 4.5298,
      "mean_token_accuracy": 0.2462132006883621,
      "num_tokens": 8493931.0,
      "step": 10655
    },
    {
      "entropy": 4.801080417633057,
      "epoch": 3.0138535482046933,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041781401220855683,
      "loss": 4.5851,
      "mean_token_accuracy": 0.25243551582098006,
      "num_tokens": 8497862.0,
      "step": 10660
    },
    {
      "entropy": 4.641027116775513,
      "epoch": 3.015267175572519,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004177345132818703,
      "loss": 4.4845,
      "mean_token_accuracy": 0.2626564458012581,
      "num_tokens": 8501768.0,
      "step": 10665
    },
    {
      "entropy": 4.644099521636963,
      "epoch": 3.016680802940345,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004176549845245531,
      "loss": 4.4538,
      "mean_token_accuracy": 0.2555350184440613,
      "num_tokens": 8505738.0,
      "step": 10670
    },
    {
      "entropy": 4.80027928352356,
      "epoch": 3.018094430308171,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004175754259532263,
      "loss": 4.5642,
      "mean_token_accuracy": 0.24532694965600968,
      "num_tokens": 8509677.0,
      "step": 10675
    },
    {
      "entropy": 4.646010398864746,
      "epoch": 3.0195080576759965,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004174958375845171,
      "loss": 4.4262,
      "mean_token_accuracy": 0.2729600086808205,
      "num_tokens": 8513819.0,
      "step": 10680
    },
    {
      "entropy": 4.747738599777222,
      "epoch": 3.0209216850438225,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004174162194350591,
      "loss": 4.5118,
      "mean_token_accuracy": 0.2524986624717712,
      "num_tokens": 8517987.0,
      "step": 10685
    },
    {
      "entropy": 4.760916709899902,
      "epoch": 3.022335312411648,
      "grad_norm": 1.828125,
      "learning_rate": 0.00041733657152149184,
      "loss": 4.5575,
      "mean_token_accuracy": 0.2574428737163544,
      "num_tokens": 8521650.0,
      "step": 10690
    },
    {
      "entropy": 4.733958196640015,
      "epoch": 3.023748939779474,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00041725689386046136,
      "loss": 4.4324,
      "mean_token_accuracy": 0.26185949295759203,
      "num_tokens": 8525393.0,
      "step": 10695
    },
    {
      "entropy": 4.7712630271911625,
      "epoch": 3.0251625671473,
      "grad_norm": 1.75,
      "learning_rate": 0.00041717718646861975,
      "loss": 4.5475,
      "mean_token_accuracy": 0.26386084109544755,
      "num_tokens": 8529251.0,
      "step": 10700
    },
    {
      "entropy": 4.727570486068726,
      "epoch": 3.0265761945151257,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004170974493626253,
      "loss": 4.4941,
      "mean_token_accuracy": 0.25962809324264524,
      "num_tokens": 8533319.0,
      "step": 10705
    },
    {
      "entropy": 4.755829668045044,
      "epoch": 3.0279898218829517,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004170176825591426,
      "loss": 4.4391,
      "mean_token_accuracy": 0.2646935746073723,
      "num_tokens": 8537210.0,
      "step": 10710
    },
    {
      "entropy": 4.767727088928223,
      "epoch": 3.0294034492507773,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004169378860748426,
      "loss": 4.5382,
      "mean_token_accuracy": 0.2513639941811562,
      "num_tokens": 8541201.0,
      "step": 10715
    },
    {
      "entropy": 4.845394611358643,
      "epoch": 3.0308170766186033,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004168580599264021,
      "loss": 4.5611,
      "mean_token_accuracy": 0.24208363741636277,
      "num_tokens": 8545191.0,
      "step": 10720
    },
    {
      "entropy": 4.7897556781768795,
      "epoch": 3.0322307039864294,
      "grad_norm": 1.625,
      "learning_rate": 0.0004167782041305042,
      "loss": 4.5392,
      "mean_token_accuracy": 0.25295937061309814,
      "num_tokens": 8549269.0,
      "step": 10725
    },
    {
      "entropy": 4.760115385055542,
      "epoch": 3.033644331354255,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041669831870383845,
      "loss": 4.4823,
      "mean_token_accuracy": 0.25992910861968993,
      "num_tokens": 8553252.0,
      "step": 10730
    },
    {
      "entropy": 4.720929670333862,
      "epoch": 3.035057958722081,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004166184036631003,
      "loss": 4.4738,
      "mean_token_accuracy": 0.25361006110906603,
      "num_tokens": 8557120.0,
      "step": 10735
    },
    {
      "entropy": 4.733621263504029,
      "epoch": 3.0364715860899065,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004165384590249915,
      "loss": 4.4237,
      "mean_token_accuracy": 0.2720044836401939,
      "num_tokens": 8561166.0,
      "step": 10740
    },
    {
      "entropy": 4.722866010665894,
      "epoch": 3.0378852134577325,
      "grad_norm": 1.71875,
      "learning_rate": 0.00041645848480622005,
      "loss": 4.4955,
      "mean_token_accuracy": 0.2579579666256905,
      "num_tokens": 8565052.0,
      "step": 10745
    },
    {
      "entropy": 4.780640602111816,
      "epoch": 3.0392988408255586,
      "grad_norm": 1.625,
      "learning_rate": 0.0004163784810235001,
      "loss": 4.5691,
      "mean_token_accuracy": 0.2543944329023361,
      "num_tokens": 8568960.0,
      "step": 10750
    },
    {
      "entropy": 4.8014898777008055,
      "epoch": 3.040712468193384,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004162984476935517,
      "loss": 4.5037,
      "mean_token_accuracy": 0.2534234136343002,
      "num_tokens": 8572961.0,
      "step": 10755
    },
    {
      "entropy": 4.783964490890503,
      "epoch": 3.04212609556121,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00041621838483310163,
      "loss": 4.4609,
      "mean_token_accuracy": 0.2591032192111015,
      "num_tokens": 8576942.0,
      "step": 10760
    },
    {
      "entropy": 4.7480837345123295,
      "epoch": 3.0435397229290357,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004161382924588824,
      "loss": 4.5268,
      "mean_token_accuracy": 0.256145778298378,
      "num_tokens": 8580944.0,
      "step": 10765
    },
    {
      "entropy": 4.709392070770264,
      "epoch": 3.0449533502968618,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004160581705876328,
      "loss": 4.5301,
      "mean_token_accuracy": 0.2557980686426163,
      "num_tokens": 8584685.0,
      "step": 10770
    },
    {
      "entropy": 4.720806312561035,
      "epoch": 3.046366977664688,
      "grad_norm": 1.546875,
      "learning_rate": 0.00041597801923609775,
      "loss": 4.5467,
      "mean_token_accuracy": 0.2537830352783203,
      "num_tokens": 8588733.0,
      "step": 10775
    },
    {
      "entropy": 4.756928253173828,
      "epoch": 3.0477806050325134,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004158978384210285,
      "loss": 4.4417,
      "mean_token_accuracy": 0.25255918502807617,
      "num_tokens": 8592662.0,
      "step": 10780
    },
    {
      "entropy": 4.744703102111816,
      "epoch": 3.0491942324003394,
      "grad_norm": 1.625,
      "learning_rate": 0.00041581762815918235,
      "loss": 4.5754,
      "mean_token_accuracy": 0.253485332429409,
      "num_tokens": 8596458.0,
      "step": 10785
    },
    {
      "entropy": 4.803004455566406,
      "epoch": 3.050607859768165,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041573738846732266,
      "loss": 4.5506,
      "mean_token_accuracy": 0.2507394939661026,
      "num_tokens": 8600413.0,
      "step": 10790
    },
    {
      "entropy": 4.740749311447144,
      "epoch": 3.052021487135991,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004156571193622191,
      "loss": 4.548,
      "mean_token_accuracy": 0.25926205664873125,
      "num_tokens": 8604384.0,
      "step": 10795
    },
    {
      "entropy": 4.7928361892700195,
      "epoch": 3.053435114503817,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004155768208606474,
      "loss": 4.5993,
      "mean_token_accuracy": 0.24098059087991713,
      "num_tokens": 8608295.0,
      "step": 10800
    },
    {
      "entropy": 4.650605583190918,
      "epoch": 3.0548487418716426,
      "grad_norm": 1.625,
      "learning_rate": 0.00041549649297938936,
      "loss": 4.4152,
      "mean_token_accuracy": 0.25775170624256133,
      "num_tokens": 8612290.0,
      "step": 10805
    },
    {
      "entropy": 4.745052480697632,
      "epoch": 3.0562623692394686,
      "grad_norm": 1.6875,
      "learning_rate": 0.00041541613573523316,
      "loss": 4.5297,
      "mean_token_accuracy": 0.25791497975587846,
      "num_tokens": 8615939.0,
      "step": 10810
    },
    {
      "entropy": 4.796959161758423,
      "epoch": 3.057675996607294,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004153357491449729,
      "loss": 4.6443,
      "mean_token_accuracy": 0.2459852546453476,
      "num_tokens": 8619770.0,
      "step": 10815
    },
    {
      "entropy": 4.694482660293579,
      "epoch": 3.05908962397512,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041525533322540886,
      "loss": 4.4286,
      "mean_token_accuracy": 0.25547972321510315,
      "num_tokens": 8623820.0,
      "step": 10820
    },
    {
      "entropy": 4.784135389328003,
      "epoch": 3.060503251342946,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004151748879933475,
      "loss": 4.5287,
      "mean_token_accuracy": 0.24460550397634506,
      "num_tokens": 8627888.0,
      "step": 10825
    },
    {
      "entropy": 4.655793285369873,
      "epoch": 3.061916878710772,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004150944134656014,
      "loss": 4.4455,
      "mean_token_accuracy": 0.2656009167432785,
      "num_tokens": 8631821.0,
      "step": 10830
    },
    {
      "entropy": 4.723963975906372,
      "epoch": 3.063330506078598,
      "grad_norm": 1.640625,
      "learning_rate": 0.00041501390965898913,
      "loss": 4.5735,
      "mean_token_accuracy": 0.24245031774044037,
      "num_tokens": 8635583.0,
      "step": 10835
    },
    {
      "entropy": 4.787011957168579,
      "epoch": 3.0647441334464234,
      "grad_norm": 1.8984375,
      "learning_rate": 0.0004149333765903356,
      "loss": 4.4613,
      "mean_token_accuracy": 0.25617674738168716,
      "num_tokens": 8639455.0,
      "step": 10840
    },
    {
      "entropy": 4.777480220794677,
      "epoch": 3.0661577608142494,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004148528142764717,
      "loss": 4.5648,
      "mean_token_accuracy": 0.2549019604921341,
      "num_tokens": 8643356.0,
      "step": 10845
    },
    {
      "entropy": 4.778955030441284,
      "epoch": 3.067571388182075,
      "grad_norm": 1.640625,
      "learning_rate": 0.00041477222273423437,
      "loss": 4.6035,
      "mean_token_accuracy": 0.24586038738489152,
      "num_tokens": 8647567.0,
      "step": 10850
    },
    {
      "entropy": 4.786934185028076,
      "epoch": 3.068985015549901,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004146916019804669,
      "loss": 4.6208,
      "mean_token_accuracy": 0.2503210499882698,
      "num_tokens": 8651426.0,
      "step": 10855
    },
    {
      "entropy": 4.726963472366333,
      "epoch": 3.070398642917727,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0004146109520320184,
      "loss": 4.5087,
      "mean_token_accuracy": 0.2529487445950508,
      "num_tokens": 8655263.0,
      "step": 10860
    },
    {
      "entropy": 4.716632080078125,
      "epoch": 3.0718122702855526,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004145302729057443,
      "loss": 4.4348,
      "mean_token_accuracy": 0.2575169950723648,
      "num_tokens": 8659230.0,
      "step": 10865
    },
    {
      "entropy": 4.745050144195557,
      "epoch": 3.0732258976533786,
      "grad_norm": 1.625,
      "learning_rate": 0.000414449564618506,
      "loss": 4.4944,
      "mean_token_accuracy": 0.25810589343309404,
      "num_tokens": 8663283.0,
      "step": 10870
    },
    {
      "entropy": 4.7860959529876705,
      "epoch": 3.0746395250212046,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00041436882718717095,
      "loss": 4.6127,
      "mean_token_accuracy": 0.24059853106737136,
      "num_tokens": 8667656.0,
      "step": 10875
    },
    {
      "entropy": 4.755101156234741,
      "epoch": 3.07605315238903,
      "grad_norm": 1.5625,
      "learning_rate": 0.00041428806062861277,
      "loss": 4.523,
      "mean_token_accuracy": 0.25148837864398954,
      "num_tokens": 8671491.0,
      "step": 10880
    },
    {
      "entropy": 4.859960222244263,
      "epoch": 3.0774667797568562,
      "grad_norm": 1.640625,
      "learning_rate": 0.00041420726495971144,
      "loss": 4.5221,
      "mean_token_accuracy": 0.24222787320613862,
      "num_tokens": 8675528.0,
      "step": 10885
    },
    {
      "entropy": 4.744062519073486,
      "epoch": 3.078880407124682,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004141264401973525,
      "loss": 4.5007,
      "mean_token_accuracy": 0.2540679857134819,
      "num_tokens": 8679635.0,
      "step": 10890
    },
    {
      "entropy": 4.750225067138672,
      "epoch": 3.080294034492508,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004140455863584279,
      "loss": 4.508,
      "mean_token_accuracy": 0.25152419209480287,
      "num_tokens": 8683951.0,
      "step": 10895
    },
    {
      "entropy": 4.702732515335083,
      "epoch": 3.0817076618603334,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004139647034598355,
      "loss": 4.4541,
      "mean_token_accuracy": 0.2636131480336189,
      "num_tokens": 8687817.0,
      "step": 10900
    },
    {
      "entropy": 4.7057384014129635,
      "epoch": 3.0831212892281594,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004138837915184795,
      "loss": 4.485,
      "mean_token_accuracy": 0.2470819503068924,
      "num_tokens": 8691748.0,
      "step": 10905
    },
    {
      "entropy": 4.727210235595703,
      "epoch": 3.0845349165959854,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00041380285055126983,
      "loss": 4.5732,
      "mean_token_accuracy": 0.24127229303121567,
      "num_tokens": 8695755.0,
      "step": 10910
    },
    {
      "entropy": 4.796623706817627,
      "epoch": 3.085948543963811,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004137218805751227,
      "loss": 4.5754,
      "mean_token_accuracy": 0.25482684969902036,
      "num_tokens": 8699758.0,
      "step": 10915
    },
    {
      "entropy": 4.794123983383178,
      "epoch": 3.087362171331637,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004136408816069605,
      "loss": 4.6209,
      "mean_token_accuracy": 0.23867883980274202,
      "num_tokens": 8703469.0,
      "step": 10920
    },
    {
      "entropy": 4.616719388961792,
      "epoch": 3.0887757986994626,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004135598536637113,
      "loss": 4.3991,
      "mean_token_accuracy": 0.2679449200630188,
      "num_tokens": 8707283.0,
      "step": 10925
    },
    {
      "entropy": 4.6749773025512695,
      "epoch": 3.0901894260672886,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004134787967623095,
      "loss": 4.5458,
      "mean_token_accuracy": 0.25459216684103014,
      "num_tokens": 8711244.0,
      "step": 10930
    },
    {
      "entropy": 4.779389238357544,
      "epoch": 3.0916030534351147,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041339771091969553,
      "loss": 4.5955,
      "mean_token_accuracy": 0.25156872123479845,
      "num_tokens": 8715230.0,
      "step": 10935
    },
    {
      "entropy": 4.825510549545288,
      "epoch": 3.0930166808029402,
      "grad_norm": 1.625,
      "learning_rate": 0.0004133165961528158,
      "loss": 4.5514,
      "mean_token_accuracy": 0.2573706850409508,
      "num_tokens": 8719071.0,
      "step": 10940
    },
    {
      "entropy": 4.743730115890503,
      "epoch": 3.0944303081707663,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00041323545247862284,
      "loss": 4.541,
      "mean_token_accuracy": 0.250384147465229,
      "num_tokens": 8723068.0,
      "step": 10945
    },
    {
      "entropy": 4.721752166748047,
      "epoch": 3.095843935538592,
      "grad_norm": 1.640625,
      "learning_rate": 0.000413154279914075,
      "loss": 4.5013,
      "mean_token_accuracy": 0.2501935660839081,
      "num_tokens": 8726994.0,
      "step": 10950
    },
    {
      "entropy": 4.735603475570679,
      "epoch": 3.097257562906418,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00041307307847613707,
      "loss": 4.565,
      "mean_token_accuracy": 0.2533019006252289,
      "num_tokens": 8730924.0,
      "step": 10955
    },
    {
      "entropy": 4.626990270614624,
      "epoch": 3.098671190274244,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00041299184818177956,
      "loss": 4.3975,
      "mean_token_accuracy": 0.2635584115982056,
      "num_tokens": 8734936.0,
      "step": 10960
    },
    {
      "entropy": 4.64911904335022,
      "epoch": 3.1000848176420694,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0004129105890479791,
      "loss": 4.4396,
      "mean_token_accuracy": 0.25925976037979126,
      "num_tokens": 8739046.0,
      "step": 10965
    },
    {
      "entropy": 4.73268575668335,
      "epoch": 3.1014984450098955,
      "grad_norm": 1.6875,
      "learning_rate": 0.00041282930109171826,
      "loss": 4.6076,
      "mean_token_accuracy": 0.2519358813762665,
      "num_tokens": 8743183.0,
      "step": 10970
    },
    {
      "entropy": 4.772938919067383,
      "epoch": 3.102912072377721,
      "grad_norm": 1.625,
      "learning_rate": 0.0004127479843299859,
      "loss": 4.5436,
      "mean_token_accuracy": 0.24709134697914123,
      "num_tokens": 8746988.0,
      "step": 10975
    },
    {
      "entropy": 4.744032669067383,
      "epoch": 3.104325699745547,
      "grad_norm": 1.96875,
      "learning_rate": 0.00041266663877977653,
      "loss": 4.4551,
      "mean_token_accuracy": 0.26281827986240386,
      "num_tokens": 8750567.0,
      "step": 10980
    },
    {
      "entropy": 4.707060956954956,
      "epoch": 3.105739327113373,
      "grad_norm": 1.5625,
      "learning_rate": 0.0004125852644580911,
      "loss": 4.4421,
      "mean_token_accuracy": 0.2667118400335312,
      "num_tokens": 8754617.0,
      "step": 10985
    },
    {
      "entropy": 4.642703199386597,
      "epoch": 3.1071529544811987,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00041250386138193615,
      "loss": 4.3824,
      "mean_token_accuracy": 0.2711236000061035,
      "num_tokens": 8758463.0,
      "step": 10990
    },
    {
      "entropy": 4.848560857772827,
      "epoch": 3.1085665818490247,
      "grad_norm": 1.6875,
      "learning_rate": 0.00041242242956832445,
      "loss": 4.5753,
      "mean_token_accuracy": 0.24716860800981522,
      "num_tokens": 8762707.0,
      "step": 10995
    },
    {
      "entropy": 4.722628784179688,
      "epoch": 3.1099802092168503,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004123409690342747,
      "loss": 4.483,
      "mean_token_accuracy": 0.26058695316314695,
      "num_tokens": 8766618.0,
      "step": 11000
    },
    {
      "epoch": 3.1099802092168503,
      "eval_entropy": 4.609517450623272,
      "eval_loss": 5.115386962890625,
      "eval_mean_token_accuracy": 0.2380724423515467,
      "eval_num_tokens": 8766618.0,
      "eval_runtime": 6.4925,
      "eval_samples_per_second": 1738.015,
      "eval_steps_per_second": 217.329,
      "step": 11000
    },
    {
      "entropy": 4.685725116729737,
      "epoch": 3.1113938365846763,
      "grad_norm": 1.71875,
      "learning_rate": 0.00041225947979681176,
      "loss": 4.4787,
      "mean_token_accuracy": 0.25468558818101883,
      "num_tokens": 8770312.0,
      "step": 11005
    },
    {
      "entropy": 4.6683775901794435,
      "epoch": 3.1128074639525023,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004121779618729664,
      "loss": 4.4203,
      "mean_token_accuracy": 0.257630418241024,
      "num_tokens": 8774518.0,
      "step": 11010
    },
    {
      "entropy": 4.688323545455932,
      "epoch": 3.114221091320328,
      "grad_norm": 1.6875,
      "learning_rate": 0.00041209641527977523,
      "loss": 4.515,
      "mean_token_accuracy": 0.26217578947544096,
      "num_tokens": 8778465.0,
      "step": 11015
    },
    {
      "entropy": 4.741280603408813,
      "epoch": 3.115634718688154,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00041201484003428095,
      "loss": 4.524,
      "mean_token_accuracy": 0.2526798203587532,
      "num_tokens": 8782508.0,
      "step": 11020
    },
    {
      "entropy": 4.781152582168579,
      "epoch": 3.1170483460559795,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004119332361535325,
      "loss": 4.5729,
      "mean_token_accuracy": 0.2508168309926987,
      "num_tokens": 8786416.0,
      "step": 11025
    },
    {
      "entropy": 4.733751583099365,
      "epoch": 3.1184619734238055,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004118516036545844,
      "loss": 4.5751,
      "mean_token_accuracy": 0.2506732687354088,
      "num_tokens": 8790189.0,
      "step": 11030
    },
    {
      "entropy": 4.757095813751221,
      "epoch": 3.1198756007916315,
      "grad_norm": 1.65625,
      "learning_rate": 0.00041176994255449743,
      "loss": 4.4855,
      "mean_token_accuracy": 0.2527921125292778,
      "num_tokens": 8794234.0,
      "step": 11035
    },
    {
      "entropy": 4.819914436340332,
      "epoch": 3.121289228159457,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004116882528703383,
      "loss": 4.5231,
      "mean_token_accuracy": 0.2604486018419266,
      "num_tokens": 8797949.0,
      "step": 11040
    },
    {
      "entropy": 4.739858102798462,
      "epoch": 3.122702855527283,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00041160653461917945,
      "loss": 4.4634,
      "mean_token_accuracy": 0.2639915391802788,
      "num_tokens": 8801856.0,
      "step": 11045
    },
    {
      "entropy": 4.606939125061035,
      "epoch": 3.1241164828951087,
      "grad_norm": 1.578125,
      "learning_rate": 0.00041152478781809963,
      "loss": 4.3942,
      "mean_token_accuracy": 0.2718272179365158,
      "num_tokens": 8805677.0,
      "step": 11050
    },
    {
      "entropy": 4.697218179702759,
      "epoch": 3.1255301102629347,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00041144301248418355,
      "loss": 4.5341,
      "mean_token_accuracy": 0.2526373267173767,
      "num_tokens": 8809777.0,
      "step": 11055
    },
    {
      "entropy": 4.704259157180786,
      "epoch": 3.1269437376307607,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004113612086345214,
      "loss": 4.4672,
      "mean_token_accuracy": 0.25652903914451597,
      "num_tokens": 8813635.0,
      "step": 11060
    },
    {
      "entropy": 4.819505882263184,
      "epoch": 3.1283573649985863,
      "grad_norm": 1.703125,
      "learning_rate": 0.00041127937628621,
      "loss": 4.5476,
      "mean_token_accuracy": 0.24700187593698503,
      "num_tokens": 8817529.0,
      "step": 11065
    },
    {
      "entropy": 4.791625356674194,
      "epoch": 3.1297709923664123,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004111975154563517,
      "loss": 4.5584,
      "mean_token_accuracy": 0.2531870245933533,
      "num_tokens": 8821366.0,
      "step": 11070
    },
    {
      "entropy": 4.674828052520752,
      "epoch": 3.131184619734238,
      "grad_norm": 1.609375,
      "learning_rate": 0.00041111562616205486,
      "loss": 4.4846,
      "mean_token_accuracy": 0.25750513672828673,
      "num_tokens": 8825246.0,
      "step": 11075
    },
    {
      "entropy": 4.700824022293091,
      "epoch": 3.132598247102064,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004110337084204338,
      "loss": 4.5221,
      "mean_token_accuracy": 0.24988095611333846,
      "num_tokens": 8829182.0,
      "step": 11080
    },
    {
      "entropy": 4.678703308105469,
      "epoch": 3.13401187446989,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004109517622486089,
      "loss": 4.5101,
      "mean_token_accuracy": 0.26422027200460435,
      "num_tokens": 8833322.0,
      "step": 11085
    },
    {
      "entropy": 4.786971139907837,
      "epoch": 3.1354255018377155,
      "grad_norm": 1.609375,
      "learning_rate": 0.00041086978766370646,
      "loss": 4.5366,
      "mean_token_accuracy": 0.24888875782489778,
      "num_tokens": 8837277.0,
      "step": 11090
    },
    {
      "entropy": 4.7792792320251465,
      "epoch": 3.1368391292055415,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004107877846828586,
      "loss": 4.4946,
      "mean_token_accuracy": 0.2589211493730545,
      "num_tokens": 8840917.0,
      "step": 11095
    },
    {
      "entropy": 4.686439895629883,
      "epoch": 3.138252756573367,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004107057533232034,
      "loss": 4.4349,
      "mean_token_accuracy": 0.2685119479894638,
      "num_tokens": 8844967.0,
      "step": 11100
    },
    {
      "entropy": 4.876041650772095,
      "epoch": 3.139666383941193,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00041062369360188485,
      "loss": 4.6159,
      "mean_token_accuracy": 0.23074452131986617,
      "num_tokens": 8848843.0,
      "step": 11105
    },
    {
      "entropy": 4.79096474647522,
      "epoch": 3.141080011309019,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004105416055360531,
      "loss": 4.605,
      "mean_token_accuracy": 0.23728083968162536,
      "num_tokens": 8852893.0,
      "step": 11110
    },
    {
      "entropy": 4.758256816864014,
      "epoch": 3.1424936386768447,
      "grad_norm": 1.65625,
      "learning_rate": 0.00041045948914286397,
      "loss": 4.6179,
      "mean_token_accuracy": 0.251055009663105,
      "num_tokens": 8856842.0,
      "step": 11115
    },
    {
      "entropy": 4.831765413284302,
      "epoch": 3.1439072660446707,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004103773444394792,
      "loss": 4.5647,
      "mean_token_accuracy": 0.2495760515332222,
      "num_tokens": 8860585.0,
      "step": 11120
    },
    {
      "entropy": 4.744802474975586,
      "epoch": 3.1453208934124963,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004102951714430666,
      "loss": 4.4697,
      "mean_token_accuracy": 0.25734851956367494,
      "num_tokens": 8864515.0,
      "step": 11125
    },
    {
      "entropy": 4.815416717529297,
      "epoch": 3.1467345207803223,
      "grad_norm": 1.625,
      "learning_rate": 0.00041021297017079984,
      "loss": 4.5793,
      "mean_token_accuracy": 0.25660873055458067,
      "num_tokens": 8868376.0,
      "step": 11130
    },
    {
      "entropy": 4.77088098526001,
      "epoch": 3.148148148148148,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00041013074063985845,
      "loss": 4.5044,
      "mean_token_accuracy": 0.2544386267662048,
      "num_tokens": 8872051.0,
      "step": 11135
    },
    {
      "entropy": 4.725830602645874,
      "epoch": 3.149561775515974,
      "grad_norm": 1.59375,
      "learning_rate": 0.00041004848286742783,
      "loss": 4.5449,
      "mean_token_accuracy": 0.255491603910923,
      "num_tokens": 8876260.0,
      "step": 11140
    },
    {
      "entropy": 4.658672618865967,
      "epoch": 3.1509754028838,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004099661968706995,
      "loss": 4.4413,
      "mean_token_accuracy": 0.24995138198137284,
      "num_tokens": 8880587.0,
      "step": 11145
    },
    {
      "entropy": 4.717408275604248,
      "epoch": 3.1523890302516255,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004098838826668705,
      "loss": 4.5402,
      "mean_token_accuracy": 0.2541167065501213,
      "num_tokens": 8884640.0,
      "step": 11150
    },
    {
      "entropy": 4.82821044921875,
      "epoch": 3.1538026576194516,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00040980154027314426,
      "loss": 4.6266,
      "mean_token_accuracy": 0.23774757832288743,
      "num_tokens": 8888381.0,
      "step": 11155
    },
    {
      "entropy": 4.78118896484375,
      "epoch": 3.1552162849872776,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00040971916970672964,
      "loss": 4.4329,
      "mean_token_accuracy": 0.2653129309415817,
      "num_tokens": 8892104.0,
      "step": 11160
    },
    {
      "entropy": 4.6843360424041744,
      "epoch": 3.156629912355103,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004096367709848415,
      "loss": 4.466,
      "mean_token_accuracy": 0.2566357120871544,
      "num_tokens": 8896017.0,
      "step": 11165
    },
    {
      "entropy": 4.715541076660156,
      "epoch": 3.158043539722929,
      "grad_norm": 1.6875,
      "learning_rate": 0.000409554344124701,
      "loss": 4.5275,
      "mean_token_accuracy": 0.25675762742757796,
      "num_tokens": 8899778.0,
      "step": 11170
    },
    {
      "entropy": 4.664355039596558,
      "epoch": 3.1594571670907547,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0004094718891435346,
      "loss": 4.564,
      "mean_token_accuracy": 0.2520316556096077,
      "num_tokens": 8903794.0,
      "step": 11175
    },
    {
      "entropy": 4.781720781326294,
      "epoch": 3.1608707944585808,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00040938940605857475,
      "loss": 4.6015,
      "mean_token_accuracy": 0.2538514629006386,
      "num_tokens": 8907503.0,
      "step": 11180
    },
    {
      "entropy": 4.820354843139649,
      "epoch": 3.1622844218264063,
      "grad_norm": 1.625,
      "learning_rate": 0.00040930689488706025,
      "loss": 4.5174,
      "mean_token_accuracy": 0.24963568896055222,
      "num_tokens": 8911388.0,
      "step": 11185
    },
    {
      "entropy": 4.813544464111328,
      "epoch": 3.1636980491942324,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00040922435564623525,
      "loss": 4.488,
      "mean_token_accuracy": 0.26188787519931794,
      "num_tokens": 8915093.0,
      "step": 11190
    },
    {
      "entropy": 4.82562665939331,
      "epoch": 3.1651116765620584,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004091417883533498,
      "loss": 4.5327,
      "mean_token_accuracy": 0.2494380295276642,
      "num_tokens": 8919568.0,
      "step": 11195
    },
    {
      "entropy": 4.82492356300354,
      "epoch": 3.166525303929884,
      "grad_norm": 1.671875,
      "learning_rate": 0.00040905919302566015,
      "loss": 4.666,
      "mean_token_accuracy": 0.23381507992744446,
      "num_tokens": 8924059.0,
      "step": 11200
    },
    {
      "entropy": 4.652480840682983,
      "epoch": 3.16793893129771,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004089765696804281,
      "loss": 4.4441,
      "mean_token_accuracy": 0.26573537439107897,
      "num_tokens": 8927937.0,
      "step": 11205
    },
    {
      "entropy": 4.709825468063355,
      "epoch": 3.1693525586655356,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040889391833492163,
      "loss": 4.4878,
      "mean_token_accuracy": 0.2522228941321373,
      "num_tokens": 8932226.0,
      "step": 11210
    },
    {
      "entropy": 4.732637882232666,
      "epoch": 3.1707661860333616,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004088112390064141,
      "loss": 4.5486,
      "mean_token_accuracy": 0.2534849911928177,
      "num_tokens": 8936086.0,
      "step": 11215
    },
    {
      "entropy": 4.819317197799682,
      "epoch": 3.1721798134011876,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004087285317121851,
      "loss": 4.6833,
      "mean_token_accuracy": 0.24085136353969575,
      "num_tokens": 8940338.0,
      "step": 11220
    },
    {
      "entropy": 4.631814956665039,
      "epoch": 3.173593440769013,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00040864579646951996,
      "loss": 4.4443,
      "mean_token_accuracy": 0.2596859812736511,
      "num_tokens": 8944590.0,
      "step": 11225
    },
    {
      "entropy": 4.801629304885864,
      "epoch": 3.175007068136839,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00040856303329570975,
      "loss": 4.5048,
      "mean_token_accuracy": 0.2583625689148903,
      "num_tokens": 8948321.0,
      "step": 11230
    },
    {
      "entropy": 4.754375839233399,
      "epoch": 3.1764206955046648,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004084802422080515,
      "loss": 4.4925,
      "mean_token_accuracy": 0.26951687335968016,
      "num_tokens": 8952344.0,
      "step": 11235
    },
    {
      "entropy": 4.812505578994751,
      "epoch": 3.177834322872491,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040839742322384807,
      "loss": 4.5815,
      "mean_token_accuracy": 0.2535375952720642,
      "num_tokens": 8956353.0,
      "step": 11240
    },
    {
      "entropy": 4.812648725509644,
      "epoch": 3.179247950240317,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004083145763604081,
      "loss": 4.613,
      "mean_token_accuracy": 0.25010265558958056,
      "num_tokens": 8960643.0,
      "step": 11245
    },
    {
      "entropy": 4.814487171173096,
      "epoch": 3.1806615776081424,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000408231701635046,
      "loss": 4.4997,
      "mean_token_accuracy": 0.26594152450561526,
      "num_tokens": 8964661.0,
      "step": 11250
    },
    {
      "entropy": 4.805761432647705,
      "epoch": 3.1820752049759684,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004081487990650822,
      "loss": 4.5762,
      "mean_token_accuracy": 0.25760549306869507,
      "num_tokens": 8968631.0,
      "step": 11255
    },
    {
      "entropy": 4.784621334075927,
      "epoch": 3.183488832343794,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004080658686678428,
      "loss": 4.4512,
      "mean_token_accuracy": 0.2569512963294983,
      "num_tokens": 8972955.0,
      "step": 11260
    },
    {
      "entropy": 4.829685211181641,
      "epoch": 3.18490245971162,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004079829104606596,
      "loss": 4.5864,
      "mean_token_accuracy": 0.2514196127653122,
      "num_tokens": 8977393.0,
      "step": 11265
    },
    {
      "entropy": 4.743310213088989,
      "epoch": 3.186316087079446,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004078999244608705,
      "loss": 4.3961,
      "mean_token_accuracy": 0.25886539965867994,
      "num_tokens": 8981206.0,
      "step": 11270
    },
    {
      "entropy": 4.748126745223999,
      "epoch": 3.1877297144472716,
      "grad_norm": 1.640625,
      "learning_rate": 0.00040781691068581905,
      "loss": 4.5539,
      "mean_token_accuracy": 0.2477567419409752,
      "num_tokens": 8985567.0,
      "step": 11275
    },
    {
      "entropy": 4.754375743865967,
      "epoch": 3.1891433418150976,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004077338691528545,
      "loss": 4.6015,
      "mean_token_accuracy": 0.2384279862046242,
      "num_tokens": 8989599.0,
      "step": 11280
    },
    {
      "entropy": 4.765879201889038,
      "epoch": 3.190556969182923,
      "grad_norm": 1.5,
      "learning_rate": 0.0004076507998793322,
      "loss": 4.4565,
      "mean_token_accuracy": 0.2553873032331467,
      "num_tokens": 8993662.0,
      "step": 11285
    },
    {
      "entropy": 4.717434072494507,
      "epoch": 3.191970596550749,
      "grad_norm": 1.671875,
      "learning_rate": 0.00040756770288261296,
      "loss": 4.5473,
      "mean_token_accuracy": 0.2563343748450279,
      "num_tokens": 8997598.0,
      "step": 11290
    },
    {
      "entropy": 4.706716060638428,
      "epoch": 3.1933842239185752,
      "grad_norm": 1.6875,
      "learning_rate": 0.00040748457818006367,
      "loss": 4.5432,
      "mean_token_accuracy": 0.2513645887374878,
      "num_tokens": 9001709.0,
      "step": 11295
    },
    {
      "entropy": 4.785051584243774,
      "epoch": 3.194797851286401,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0004074014257890567,
      "loss": 4.5634,
      "mean_token_accuracy": 0.25241773277521135,
      "num_tokens": 9005861.0,
      "step": 11300
    },
    {
      "entropy": 4.645834302902221,
      "epoch": 3.196211478654227,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004073182457269705,
      "loss": 4.4479,
      "mean_token_accuracy": 0.2648533329367638,
      "num_tokens": 9009897.0,
      "step": 11305
    },
    {
      "entropy": 4.782556438446045,
      "epoch": 3.1976251060220524,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00040723503801118923,
      "loss": 4.5441,
      "mean_token_accuracy": 0.2540422111749649,
      "num_tokens": 9014021.0,
      "step": 11310
    },
    {
      "entropy": 4.776630783081055,
      "epoch": 3.1990387333898784,
      "grad_norm": 1.59375,
      "learning_rate": 0.00040715180265910267,
      "loss": 4.5143,
      "mean_token_accuracy": 0.25995569974184035,
      "num_tokens": 9018062.0,
      "step": 11315
    },
    {
      "entropy": 4.71671986579895,
      "epoch": 3.2004523607577045,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004070685396881066,
      "loss": 4.6352,
      "mean_token_accuracy": 0.2359670251607895,
      "num_tokens": 9022276.0,
      "step": 11320
    },
    {
      "entropy": 4.723827648162842,
      "epoch": 3.20186598812553,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0004069852491156024,
      "loss": 4.4951,
      "mean_token_accuracy": 0.2542228132486343,
      "num_tokens": 9026479.0,
      "step": 11325
    },
    {
      "entropy": 4.892261028289795,
      "epoch": 3.203279615493356,
      "grad_norm": 1.59375,
      "learning_rate": 0.0004069019309589973,
      "loss": 4.7007,
      "mean_token_accuracy": 0.23851681351661683,
      "num_tokens": 9030828.0,
      "step": 11330
    },
    {
      "entropy": 4.729081821441651,
      "epoch": 3.2046932428611816,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00040681858523570433,
      "loss": 4.4097,
      "mean_token_accuracy": 0.2714554235339165,
      "num_tokens": 9034761.0,
      "step": 11335
    },
    {
      "entropy": 4.6872556686401365,
      "epoch": 3.2061068702290076,
      "grad_norm": 1.640625,
      "learning_rate": 0.00040673521196314216,
      "loss": 4.4502,
      "mean_token_accuracy": 0.26070596575737,
      "num_tokens": 9038778.0,
      "step": 11340
    },
    {
      "entropy": 4.75812873840332,
      "epoch": 3.2075204975968337,
      "grad_norm": 1.515625,
      "learning_rate": 0.00040665181115873524,
      "loss": 4.5152,
      "mean_token_accuracy": 0.2502188667654991,
      "num_tokens": 9042922.0,
      "step": 11345
    },
    {
      "entropy": 4.822296905517578,
      "epoch": 3.2089341249646592,
      "grad_norm": 1.734375,
      "learning_rate": 0.00040656838283991396,
      "loss": 4.662,
      "mean_token_accuracy": 0.23720723390579224,
      "num_tokens": 9047323.0,
      "step": 11350
    },
    {
      "entropy": 4.8889084815979,
      "epoch": 3.2103477523324853,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00040648492702411424,
      "loss": 4.6291,
      "mean_token_accuracy": 0.2551712289452553,
      "num_tokens": 9051144.0,
      "step": 11355
    },
    {
      "entropy": 4.746240234375,
      "epoch": 3.211761379700311,
      "grad_norm": 1.53125,
      "learning_rate": 0.000406401443728778,
      "loss": 4.604,
      "mean_token_accuracy": 0.2504086077213287,
      "num_tokens": 9054919.0,
      "step": 11360
    },
    {
      "entropy": 4.726230955123901,
      "epoch": 3.213175007068137,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040631793297135247,
      "loss": 4.5002,
      "mean_token_accuracy": 0.26156305223703386,
      "num_tokens": 9058804.0,
      "step": 11365
    },
    {
      "entropy": 4.674736881256104,
      "epoch": 3.214588634435963,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00040623439476929093,
      "loss": 4.4185,
      "mean_token_accuracy": 0.26230787932872773,
      "num_tokens": 9062893.0,
      "step": 11370
    },
    {
      "entropy": 4.800751972198486,
      "epoch": 3.2160022618037885,
      "grad_norm": 1.625,
      "learning_rate": 0.00040615082914005255,
      "loss": 4.6222,
      "mean_token_accuracy": 0.2481490671634674,
      "num_tokens": 9066861.0,
      "step": 11375
    },
    {
      "entropy": 4.699177217483521,
      "epoch": 3.2174158891716145,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00040606723610110175,
      "loss": 4.4945,
      "mean_token_accuracy": 0.2530717045068741,
      "num_tokens": 9070783.0,
      "step": 11380
    },
    {
      "entropy": 4.786669969558716,
      "epoch": 3.21882951653944,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00040598361566990916,
      "loss": 4.4955,
      "mean_token_accuracy": 0.2564746022224426,
      "num_tokens": 9074778.0,
      "step": 11385
    },
    {
      "entropy": 4.726076984405518,
      "epoch": 3.220243143907266,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004058999678639509,
      "loss": 4.4968,
      "mean_token_accuracy": 0.2552886322140694,
      "num_tokens": 9078438.0,
      "step": 11390
    },
    {
      "entropy": 4.798279285430908,
      "epoch": 3.221656771275092,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040581629270070876,
      "loss": 4.4788,
      "mean_token_accuracy": 0.2546743765473366,
      "num_tokens": 9082403.0,
      "step": 11395
    },
    {
      "entropy": 4.706258487701416,
      "epoch": 3.2230703986429177,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0004057325901976704,
      "loss": 4.5089,
      "mean_token_accuracy": 0.2503048896789551,
      "num_tokens": 9086342.0,
      "step": 11400
    },
    {
      "entropy": 4.740490102767945,
      "epoch": 3.2244840260107437,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004056488603723292,
      "loss": 4.5666,
      "mean_token_accuracy": 0.25425730347633363,
      "num_tokens": 9090508.0,
      "step": 11405
    },
    {
      "entropy": 4.651725816726684,
      "epoch": 3.2258976533785693,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000405565103242184,
      "loss": 4.4439,
      "mean_token_accuracy": 0.24690255671739578,
      "num_tokens": 9094688.0,
      "step": 11410
    },
    {
      "entropy": 4.616346788406372,
      "epoch": 3.2273112807463953,
      "grad_norm": 1.59375,
      "learning_rate": 0.00040548131882473963,
      "loss": 4.5055,
      "mean_token_accuracy": 0.25910854041576387,
      "num_tokens": 9098885.0,
      "step": 11415
    },
    {
      "entropy": 4.70210108757019,
      "epoch": 3.228724908114221,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00040539750713750657,
      "loss": 4.4918,
      "mean_token_accuracy": 0.25806377083063126,
      "num_tokens": 9103286.0,
      "step": 11420
    },
    {
      "entropy": 4.715872049331665,
      "epoch": 3.230138535482047,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040531366819800076,
      "loss": 4.4834,
      "mean_token_accuracy": 0.25907076746225355,
      "num_tokens": 9107406.0,
      "step": 11425
    },
    {
      "entropy": 4.6433555603027346,
      "epoch": 3.231552162849873,
      "grad_norm": 1.53125,
      "learning_rate": 0.00040522980202374414,
      "loss": 4.4278,
      "mean_token_accuracy": 0.2520529180765152,
      "num_tokens": 9111630.0,
      "step": 11430
    },
    {
      "entropy": 4.628477096557617,
      "epoch": 3.2329657902176985,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004051459086322643,
      "loss": 4.4434,
      "mean_token_accuracy": 0.2585014745593071,
      "num_tokens": 9115784.0,
      "step": 11435
    },
    {
      "entropy": 4.6911359310150145,
      "epoch": 3.2343794175855245,
      "grad_norm": 1.609375,
      "learning_rate": 0.00040506198804109427,
      "loss": 4.5341,
      "mean_token_accuracy": 0.23990325033664703,
      "num_tokens": 9119715.0,
      "step": 11440
    },
    {
      "entropy": 4.709607648849487,
      "epoch": 3.2357930449533505,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004049780402677731,
      "loss": 4.4144,
      "mean_token_accuracy": 0.2721054330468178,
      "num_tokens": 9123574.0,
      "step": 11445
    },
    {
      "entropy": 4.6553911685943605,
      "epoch": 3.237206672321176,
      "grad_norm": 1.71875,
      "learning_rate": 0.0004048940653298452,
      "loss": 4.4807,
      "mean_token_accuracy": 0.2653071269392967,
      "num_tokens": 9127329.0,
      "step": 11450
    },
    {
      "entropy": 4.700263595581054,
      "epoch": 3.238620299689002,
      "grad_norm": 1.6875,
      "learning_rate": 0.000404810063244861,
      "loss": 4.4648,
      "mean_token_accuracy": 0.2528074949979782,
      "num_tokens": 9130974.0,
      "step": 11455
    },
    {
      "entropy": 4.687939405441284,
      "epoch": 3.2400339270568277,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0004047260340303762,
      "loss": 4.5057,
      "mean_token_accuracy": 0.2599417433142662,
      "num_tokens": 9135042.0,
      "step": 11460
    },
    {
      "entropy": 4.697766971588135,
      "epoch": 3.2414475544246537,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004046419777039525,
      "loss": 4.5454,
      "mean_token_accuracy": 0.2559804439544678,
      "num_tokens": 9138921.0,
      "step": 11465
    },
    {
      "entropy": 4.665698099136352,
      "epoch": 3.2428611817924793,
      "grad_norm": 1.609375,
      "learning_rate": 0.00040455789428315707,
      "loss": 4.4668,
      "mean_token_accuracy": 0.25632801055908205,
      "num_tokens": 9142994.0,
      "step": 11470
    },
    {
      "entropy": 4.67761116027832,
      "epoch": 3.2442748091603053,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00040447378378556307,
      "loss": 4.6675,
      "mean_token_accuracy": 0.2362545520067215,
      "num_tokens": 9146911.0,
      "step": 11475
    },
    {
      "entropy": 4.704593276977539,
      "epoch": 3.2456884365281313,
      "grad_norm": 1.5625,
      "learning_rate": 0.00040438964622874874,
      "loss": 4.4374,
      "mean_token_accuracy": 0.2586700186133385,
      "num_tokens": 9151026.0,
      "step": 11480
    },
    {
      "entropy": 4.693751621246338,
      "epoch": 3.247102063895957,
      "grad_norm": 1.609375,
      "learning_rate": 0.0004043054816302986,
      "loss": 4.4976,
      "mean_token_accuracy": 0.25725201368331907,
      "num_tokens": 9154999.0,
      "step": 11485
    },
    {
      "entropy": 4.648928642272949,
      "epoch": 3.248515691263783,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00040422129000780235,
      "loss": 4.5246,
      "mean_token_accuracy": 0.260522848367691,
      "num_tokens": 9159099.0,
      "step": 11490
    },
    {
      "entropy": 4.681011533737182,
      "epoch": 3.249929318631609,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004041370713788555,
      "loss": 4.4551,
      "mean_token_accuracy": 0.25827138125896454,
      "num_tokens": 9162752.0,
      "step": 11495
    },
    {
      "entropy": 4.7788911819458,
      "epoch": 3.2513429459994345,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004040528257610593,
      "loss": 4.5664,
      "mean_token_accuracy": 0.2609078511595726,
      "num_tokens": 9166537.0,
      "step": 11500
    },
    {
      "entropy": 4.837022304534912,
      "epoch": 3.2527565733672605,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004039685531720207,
      "loss": 4.6241,
      "mean_token_accuracy": 0.2446375772356987,
      "num_tokens": 9170684.0,
      "step": 11505
    },
    {
      "entropy": 4.820692491531372,
      "epoch": 3.254170200735086,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0004038842536293518,
      "loss": 4.5578,
      "mean_token_accuracy": 0.24526094645261765,
      "num_tokens": 9174642.0,
      "step": 11510
    },
    {
      "entropy": 4.8394030094146725,
      "epoch": 3.255583828102912,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0004037999271506711,
      "loss": 4.6108,
      "mean_token_accuracy": 0.24286877363920212,
      "num_tokens": 9178653.0,
      "step": 11515
    },
    {
      "entropy": 4.773522233963012,
      "epoch": 3.2569974554707377,
      "grad_norm": 1.71875,
      "learning_rate": 0.00040371557375360194,
      "loss": 4.5434,
      "mean_token_accuracy": 0.25553553700447085,
      "num_tokens": 9182509.0,
      "step": 11520
    },
    {
      "entropy": 4.662529945373535,
      "epoch": 3.2584110828385637,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004036311934557739,
      "loss": 4.4451,
      "mean_token_accuracy": 0.2674607992172241,
      "num_tokens": 9186605.0,
      "step": 11525
    },
    {
      "entropy": 4.715557336807251,
      "epoch": 3.2598247102063898,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004035467862748219,
      "loss": 4.4377,
      "mean_token_accuracy": 0.26141873747110367,
      "num_tokens": 9190437.0,
      "step": 11530
    },
    {
      "entropy": 4.653559923171997,
      "epoch": 3.2612383375742153,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004034623522283863,
      "loss": 4.4235,
      "mean_token_accuracy": 0.2605094239115715,
      "num_tokens": 9194757.0,
      "step": 11535
    },
    {
      "entropy": 4.746738004684448,
      "epoch": 3.2626519649420413,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0004033778913341136,
      "loss": 4.5889,
      "mean_token_accuracy": 0.24819060415029526,
      "num_tokens": 9199165.0,
      "step": 11540
    },
    {
      "entropy": 4.71032681465149,
      "epoch": 3.264065592309867,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00040329340360965556,
      "loss": 4.4906,
      "mean_token_accuracy": 0.26231479793787005,
      "num_tokens": 9203283.0,
      "step": 11545
    },
    {
      "entropy": 4.706463050842285,
      "epoch": 3.265479219677693,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0004032088890726694,
      "loss": 4.5005,
      "mean_token_accuracy": 0.25031210035085677,
      "num_tokens": 9207038.0,
      "step": 11550
    },
    {
      "entropy": 4.71755576133728,
      "epoch": 3.266892847045519,
      "grad_norm": 1.5625,
      "learning_rate": 0.00040312434774081817,
      "loss": 4.5395,
      "mean_token_accuracy": 0.25039356052875517,
      "num_tokens": 9211035.0,
      "step": 11555
    },
    {
      "entropy": 4.766290760040283,
      "epoch": 3.2683064744133445,
      "grad_norm": 1.625,
      "learning_rate": 0.0004030397796317707,
      "loss": 4.5296,
      "mean_token_accuracy": 0.24798959642648696,
      "num_tokens": 9215226.0,
      "step": 11560
    },
    {
      "entropy": 4.821901512145996,
      "epoch": 3.2697201017811706,
      "grad_norm": 1.6171875,
      "learning_rate": 0.000402955184763201,
      "loss": 4.5384,
      "mean_token_accuracy": 0.24522289037704467,
      "num_tokens": 9219461.0,
      "step": 11565
    },
    {
      "entropy": 4.706340265274048,
      "epoch": 3.271133729148996,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004028705631527889,
      "loss": 4.5587,
      "mean_token_accuracy": 0.2603594437241554,
      "num_tokens": 9223501.0,
      "step": 11570
    },
    {
      "entropy": 4.620186233520508,
      "epoch": 3.272547356516822,
      "grad_norm": 1.625,
      "learning_rate": 0.0004027859148182199,
      "loss": 4.5442,
      "mean_token_accuracy": 0.2642295897006989,
      "num_tokens": 9227649.0,
      "step": 11575
    },
    {
      "entropy": 4.7570868015289305,
      "epoch": 3.273960983884648,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00040270123977718486,
      "loss": 4.5932,
      "mean_token_accuracy": 0.2513409420847893,
      "num_tokens": 9231512.0,
      "step": 11580
    },
    {
      "entropy": 4.844477272033691,
      "epoch": 3.2753746112524738,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004026165380473804,
      "loss": 4.5959,
      "mean_token_accuracy": 0.2502386450767517,
      "num_tokens": 9235413.0,
      "step": 11585
    },
    {
      "entropy": 4.832235956192017,
      "epoch": 3.2767882386202998,
      "grad_norm": 1.703125,
      "learning_rate": 0.0004025318096465087,
      "loss": 4.5088,
      "mean_token_accuracy": 0.24797432869672775,
      "num_tokens": 9239256.0,
      "step": 11590
    },
    {
      "entropy": 4.719915914535522,
      "epoch": 3.2782018659881254,
      "grad_norm": 1.671875,
      "learning_rate": 0.0004024470545922773,
      "loss": 4.4874,
      "mean_token_accuracy": 0.2520067095756531,
      "num_tokens": 9243193.0,
      "step": 11595
    },
    {
      "entropy": 4.738100814819336,
      "epoch": 3.2796154933559514,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004023622729023997,
      "loss": 4.5898,
      "mean_token_accuracy": 0.2474185883998871,
      "num_tokens": 9247151.0,
      "step": 11600
    },
    {
      "entropy": 4.6696852684021,
      "epoch": 3.2810291207237774,
      "grad_norm": 1.578125,
      "learning_rate": 0.00040227746459459455,
      "loss": 4.5078,
      "mean_token_accuracy": 0.25256339758634566,
      "num_tokens": 9250959.0,
      "step": 11605
    },
    {
      "entropy": 4.702102613449097,
      "epoch": 3.282442748091603,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00040219262968658646,
      "loss": 4.5087,
      "mean_token_accuracy": 0.25967957079410553,
      "num_tokens": 9254998.0,
      "step": 11610
    },
    {
      "entropy": 4.750143241882324,
      "epoch": 3.283856375459429,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00040210776819610527,
      "loss": 4.5558,
      "mean_token_accuracy": 0.2528282627463341,
      "num_tokens": 9259023.0,
      "step": 11615
    },
    {
      "entropy": 4.9519267082214355,
      "epoch": 3.2852700028272546,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004020228801408866,
      "loss": 4.7761,
      "mean_token_accuracy": 0.23289839625358583,
      "num_tokens": 9263230.0,
      "step": 11620
    },
    {
      "entropy": 4.787384510040283,
      "epoch": 3.2866836301950806,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00040193796553867137,
      "loss": 4.5117,
      "mean_token_accuracy": 0.26138696223497393,
      "num_tokens": 9267296.0,
      "step": 11625
    },
    {
      "entropy": 4.797619438171386,
      "epoch": 3.2880972575629066,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00040185302440720627,
      "loss": 4.5425,
      "mean_token_accuracy": 0.2522766560316086,
      "num_tokens": 9271535.0,
      "step": 11630
    },
    {
      "entropy": 4.874114084243774,
      "epoch": 3.289510884930732,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0004017680567642435,
      "loss": 4.6572,
      "mean_token_accuracy": 0.24154347628355027,
      "num_tokens": 9275543.0,
      "step": 11635
    },
    {
      "entropy": 4.703148984909058,
      "epoch": 3.290924512298558,
      "grad_norm": 1.765625,
      "learning_rate": 0.0004016830626275409,
      "loss": 4.4591,
      "mean_token_accuracy": 0.2648521259427071,
      "num_tokens": 9279438.0,
      "step": 11640
    },
    {
      "entropy": 4.752873420715332,
      "epoch": 3.292338139666384,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00040159804201486156,
      "loss": 4.4827,
      "mean_token_accuracy": 0.25875334441661835,
      "num_tokens": 9283615.0,
      "step": 11645
    },
    {
      "entropy": 4.868975448608398,
      "epoch": 3.29375176703421,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004015129949439743,
      "loss": 4.6081,
      "mean_token_accuracy": 0.24504052549600602,
      "num_tokens": 9287593.0,
      "step": 11650
    },
    {
      "entropy": 4.792555809020996,
      "epoch": 3.2951653944020354,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00040142792143265337,
      "loss": 4.5291,
      "mean_token_accuracy": 0.25770835280418397,
      "num_tokens": 9291398.0,
      "step": 11655
    },
    {
      "entropy": 4.750729274749756,
      "epoch": 3.2965790217698614,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004013428214986787,
      "loss": 4.5361,
      "mean_token_accuracy": 0.25210361629724504,
      "num_tokens": 9295340.0,
      "step": 11660
    },
    {
      "entropy": 4.8025869846344,
      "epoch": 3.2979926491376874,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00040125769515983564,
      "loss": 4.5186,
      "mean_token_accuracy": 0.2434157520532608,
      "num_tokens": 9299462.0,
      "step": 11665
    },
    {
      "entropy": 4.770634365081787,
      "epoch": 3.299406276505513,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00040117254243391505,
      "loss": 4.565,
      "mean_token_accuracy": 0.2406589150428772,
      "num_tokens": 9303330.0,
      "step": 11670
    },
    {
      "entropy": 4.802989435195923,
      "epoch": 3.300819903873339,
      "grad_norm": 1.6875,
      "learning_rate": 0.0004010873633387133,
      "loss": 4.642,
      "mean_token_accuracy": 0.2477644070982933,
      "num_tokens": 9307247.0,
      "step": 11675
    },
    {
      "entropy": 4.801580619812012,
      "epoch": 3.302233531241165,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00040100215789203226,
      "loss": 4.5549,
      "mean_token_accuracy": 0.24997903704643248,
      "num_tokens": 9311340.0,
      "step": 11680
    },
    {
      "entropy": 4.756694889068603,
      "epoch": 3.3036471586089906,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004009169261116794,
      "loss": 4.486,
      "mean_token_accuracy": 0.2655807867646217,
      "num_tokens": 9315451.0,
      "step": 11685
    },
    {
      "entropy": 4.799614191055298,
      "epoch": 3.3050607859768166,
      "grad_norm": 1.53125,
      "learning_rate": 0.0004008316680154678,
      "loss": 4.6054,
      "mean_token_accuracy": 0.24398130029439927,
      "num_tokens": 9319657.0,
      "step": 11690
    },
    {
      "entropy": 4.786456489562989,
      "epoch": 3.306474413344642,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00040074638362121547,
      "loss": 4.4713,
      "mean_token_accuracy": 0.25596644282341,
      "num_tokens": 9323702.0,
      "step": 11695
    },
    {
      "entropy": 4.725726366043091,
      "epoch": 3.3078880407124682,
      "grad_norm": 1.734375,
      "learning_rate": 0.0004006610729467467,
      "loss": 4.5458,
      "mean_token_accuracy": 0.26209097653627395,
      "num_tokens": 9327701.0,
      "step": 11700
    },
    {
      "entropy": 4.684254264831543,
      "epoch": 3.309301668080294,
      "grad_norm": 1.65625,
      "learning_rate": 0.0004005757360098907,
      "loss": 4.5621,
      "mean_token_accuracy": 0.2534474804997444,
      "num_tokens": 9331328.0,
      "step": 11705
    },
    {
      "entropy": 4.732868385314942,
      "epoch": 3.31071529544812,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00040049037282848237,
      "loss": 4.5088,
      "mean_token_accuracy": 0.255293470621109,
      "num_tokens": 9335187.0,
      "step": 11710
    },
    {
      "entropy": 4.770110034942627,
      "epoch": 3.312128922815946,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00040040498342036216,
      "loss": 4.6149,
      "mean_token_accuracy": 0.24745299071073532,
      "num_tokens": 9339167.0,
      "step": 11715
    },
    {
      "entropy": 4.700202894210816,
      "epoch": 3.3135425501837714,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000400319567803376,
      "loss": 4.52,
      "mean_token_accuracy": 0.24982665181159974,
      "num_tokens": 9343172.0,
      "step": 11720
    },
    {
      "entropy": 4.655005311965942,
      "epoch": 3.3149561775515974,
      "grad_norm": 1.609375,
      "learning_rate": 0.00040023412599537496,
      "loss": 4.6022,
      "mean_token_accuracy": 0.25487871170043946,
      "num_tokens": 9347296.0,
      "step": 11725
    },
    {
      "entropy": 4.665102434158325,
      "epoch": 3.3163698049194235,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0004001486580142161,
      "loss": 4.4894,
      "mean_token_accuracy": 0.26578675359487536,
      "num_tokens": 9350919.0,
      "step": 11730
    },
    {
      "entropy": 4.750127506256104,
      "epoch": 3.317783432287249,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00040006316387776166,
      "loss": 4.4894,
      "mean_token_accuracy": 0.25779502391815184,
      "num_tokens": 9354748.0,
      "step": 11735
    },
    {
      "entropy": 4.5782506465911865,
      "epoch": 3.319197059655075,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003999776436038792,
      "loss": 4.4228,
      "mean_token_accuracy": 0.261199477314949,
      "num_tokens": 9358794.0,
      "step": 11740
    },
    {
      "entropy": 4.682069730758667,
      "epoch": 3.3206106870229006,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039989209721044217,
      "loss": 4.416,
      "mean_token_accuracy": 0.26730038672685624,
      "num_tokens": 9362882.0,
      "step": 11745
    },
    {
      "entropy": 4.825874376296997,
      "epoch": 3.3220243143907267,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0003998065247153291,
      "loss": 4.653,
      "mean_token_accuracy": 0.24836213290691375,
      "num_tokens": 9367004.0,
      "step": 11750
    },
    {
      "entropy": 4.787203311920166,
      "epoch": 3.3234379417585522,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003997209261364241,
      "loss": 4.5873,
      "mean_token_accuracy": 0.24769493341445922,
      "num_tokens": 9371259.0,
      "step": 11755
    },
    {
      "entropy": 4.744393730163575,
      "epoch": 3.3248515691263782,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00039963530149161685,
      "loss": 4.5295,
      "mean_token_accuracy": 0.2525615617632866,
      "num_tokens": 9375296.0,
      "step": 11760
    },
    {
      "entropy": 4.6531554698944095,
      "epoch": 3.3262651964942043,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003995496507988023,
      "loss": 4.4906,
      "mean_token_accuracy": 0.26066738814115525,
      "num_tokens": 9379505.0,
      "step": 11765
    },
    {
      "entropy": 4.661748552322388,
      "epoch": 3.32767882386203,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003994639740758809,
      "loss": 4.5905,
      "mean_token_accuracy": 0.24727074503898622,
      "num_tokens": 9383450.0,
      "step": 11770
    },
    {
      "entropy": 4.8557604312896725,
      "epoch": 3.329092451229856,
      "grad_norm": 1.828125,
      "learning_rate": 0.00039937827134075866,
      "loss": 4.66,
      "mean_token_accuracy": 0.23838064819574356,
      "num_tokens": 9387632.0,
      "step": 11775
    },
    {
      "entropy": 4.771892118453979,
      "epoch": 3.330506078597682,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003992925426113468,
      "loss": 4.5129,
      "mean_token_accuracy": 0.24799626022577287,
      "num_tokens": 9391768.0,
      "step": 11780
    },
    {
      "entropy": 4.723316431045532,
      "epoch": 3.3319197059655075,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000399206787905562,
      "loss": 4.5193,
      "mean_token_accuracy": 0.2449290096759796,
      "num_tokens": 9395606.0,
      "step": 11785
    },
    {
      "entropy": 4.707998037338257,
      "epoch": 3.3333333333333335,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003991210072413268,
      "loss": 4.6498,
      "mean_token_accuracy": 0.24798137098550796,
      "num_tokens": 9399624.0,
      "step": 11790
    },
    {
      "entropy": 4.873126077651977,
      "epoch": 3.334746960701159,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003990352006365685,
      "loss": 4.6507,
      "mean_token_accuracy": 0.24394975155591964,
      "num_tokens": 9403537.0,
      "step": 11795
    },
    {
      "entropy": 4.814060306549072,
      "epoch": 3.336160588068985,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003989493681092204,
      "loss": 4.5306,
      "mean_token_accuracy": 0.2703120693564415,
      "num_tokens": 9407496.0,
      "step": 11800
    },
    {
      "entropy": 4.76416916847229,
      "epoch": 3.3375742154368107,
      "grad_norm": 1.65625,
      "learning_rate": 0.00039886350967722074,
      "loss": 4.5901,
      "mean_token_accuracy": 0.24035219699144364,
      "num_tokens": 9411397.0,
      "step": 11805
    },
    {
      "entropy": 4.572795391082764,
      "epoch": 3.3389878428046367,
      "grad_norm": 2.0,
      "learning_rate": 0.0003987776253585135,
      "loss": 4.4349,
      "mean_token_accuracy": 0.2682676583528519,
      "num_tokens": 9415404.0,
      "step": 11810
    },
    {
      "entropy": 4.686169862747192,
      "epoch": 3.3404014701724627,
      "grad_norm": 1.703125,
      "learning_rate": 0.00039869171517104804,
      "loss": 4.4761,
      "mean_token_accuracy": 0.25725776255130767,
      "num_tokens": 9419509.0,
      "step": 11815
    },
    {
      "entropy": 4.828030300140381,
      "epoch": 3.3418150975402883,
      "grad_norm": 1.703125,
      "learning_rate": 0.000398605779132779,
      "loss": 4.5663,
      "mean_token_accuracy": 0.24422105699777602,
      "num_tokens": 9423516.0,
      "step": 11820
    },
    {
      "entropy": 4.785270023345947,
      "epoch": 3.3432287249081143,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0003985198172616664,
      "loss": 4.5813,
      "mean_token_accuracy": 0.24609220921993255,
      "num_tokens": 9427337.0,
      "step": 11825
    },
    {
      "entropy": 4.773994302749633,
      "epoch": 3.3446423522759403,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039843382957567585,
      "loss": 4.4805,
      "mean_token_accuracy": 0.26341297179460527,
      "num_tokens": 9431188.0,
      "step": 11830
    },
    {
      "entropy": 4.693052768707275,
      "epoch": 3.346055979643766,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003983478160927782,
      "loss": 4.4766,
      "mean_token_accuracy": 0.27057758122682574,
      "num_tokens": 9434877.0,
      "step": 11835
    },
    {
      "entropy": 4.686493349075318,
      "epoch": 3.347469607011592,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00039826177683094984,
      "loss": 4.5126,
      "mean_token_accuracy": 0.2532248005270958,
      "num_tokens": 9438937.0,
      "step": 11840
    },
    {
      "entropy": 4.630051374435425,
      "epoch": 3.3488832343794175,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003981757118081722,
      "loss": 4.5423,
      "mean_token_accuracy": 0.25401895493268967,
      "num_tokens": 9443167.0,
      "step": 11845
    },
    {
      "entropy": 4.723645734786987,
      "epoch": 3.3502968617472435,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003980896210424326,
      "loss": 4.5966,
      "mean_token_accuracy": 0.24817835688591003,
      "num_tokens": 9446946.0,
      "step": 11850
    },
    {
      "entropy": 4.648865556716919,
      "epoch": 3.351710489115069,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003980035045517233,
      "loss": 4.4755,
      "mean_token_accuracy": 0.270558999478817,
      "num_tokens": 9450886.0,
      "step": 11855
    },
    {
      "entropy": 4.663324213027954,
      "epoch": 3.353124116482895,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003979173623540423,
      "loss": 4.5749,
      "mean_token_accuracy": 0.2618050917983055,
      "num_tokens": 9454896.0,
      "step": 11860
    },
    {
      "entropy": 4.7491294860839846,
      "epoch": 3.354537743850721,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003978311944673925,
      "loss": 4.6111,
      "mean_token_accuracy": 0.24897866249084472,
      "num_tokens": 9458730.0,
      "step": 11865
    },
    {
      "entropy": 4.879478979110718,
      "epoch": 3.3559513712185467,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00039774500090978274,
      "loss": 4.6004,
      "mean_token_accuracy": 0.24495669901371003,
      "num_tokens": 9462909.0,
      "step": 11870
    },
    {
      "entropy": 4.824182796478271,
      "epoch": 3.3573649985863727,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039765878169922673,
      "loss": 4.6649,
      "mean_token_accuracy": 0.24313800036907196,
      "num_tokens": 9467145.0,
      "step": 11875
    },
    {
      "entropy": 4.679928970336914,
      "epoch": 3.3587786259541983,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00039757253685374387,
      "loss": 4.443,
      "mean_token_accuracy": 0.2634525939822197,
      "num_tokens": 9471095.0,
      "step": 11880
    },
    {
      "entropy": 4.7136725902557375,
      "epoch": 3.3601922533220243,
      "grad_norm": 1.578125,
      "learning_rate": 0.00039748626639135885,
      "loss": 4.4322,
      "mean_token_accuracy": 0.2648313954472542,
      "num_tokens": 9474857.0,
      "step": 11885
    },
    {
      "entropy": 4.779872465133667,
      "epoch": 3.3616058806898503,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00039739997033010153,
      "loss": 4.5235,
      "mean_token_accuracy": 0.254133303463459,
      "num_tokens": 9478676.0,
      "step": 11890
    },
    {
      "entropy": 4.744399118423462,
      "epoch": 3.363019508057676,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003973136486880073,
      "loss": 4.4872,
      "mean_token_accuracy": 0.2647264748811722,
      "num_tokens": 9482774.0,
      "step": 11895
    },
    {
      "entropy": 4.765792322158814,
      "epoch": 3.364433135425502,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0003972273014831169,
      "loss": 4.5141,
      "mean_token_accuracy": 0.25604102909564974,
      "num_tokens": 9486640.0,
      "step": 11900
    },
    {
      "entropy": 4.666509914398193,
      "epoch": 3.3658467627933275,
      "grad_norm": 1.71875,
      "learning_rate": 0.00039714092873347623,
      "loss": 4.6101,
      "mean_token_accuracy": 0.24821129739284514,
      "num_tokens": 9490425.0,
      "step": 11905
    },
    {
      "entropy": 4.752802753448487,
      "epoch": 3.3672603901611535,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00039705453045713686,
      "loss": 4.6563,
      "mean_token_accuracy": 0.2416580006480217,
      "num_tokens": 9494514.0,
      "step": 11910
    },
    {
      "entropy": 4.729694414138794,
      "epoch": 3.3686740175289795,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003969681066721553,
      "loss": 4.579,
      "mean_token_accuracy": 0.2529765531420708,
      "num_tokens": 9498485.0,
      "step": 11915
    },
    {
      "entropy": 4.792001390457154,
      "epoch": 3.370087644896805,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003968816573965938,
      "loss": 4.5194,
      "mean_token_accuracy": 0.25520717948675153,
      "num_tokens": 9502540.0,
      "step": 11920
    },
    {
      "entropy": 4.738965463638306,
      "epoch": 3.371501272264631,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00039679518264851944,
      "loss": 4.5218,
      "mean_token_accuracy": 0.2574215650558472,
      "num_tokens": 9506645.0,
      "step": 11925
    },
    {
      "entropy": 4.769187259674072,
      "epoch": 3.3729148996324567,
      "grad_norm": 1.6875,
      "learning_rate": 0.00039670868244600515,
      "loss": 4.4819,
      "mean_token_accuracy": 0.262229922413826,
      "num_tokens": 9510894.0,
      "step": 11930
    },
    {
      "entropy": 4.716852188110352,
      "epoch": 3.3743285270002827,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003966221568071288,
      "loss": 4.5896,
      "mean_token_accuracy": 0.2610588014125824,
      "num_tokens": 9514990.0,
      "step": 11935
    },
    {
      "entropy": 4.7713415145874025,
      "epoch": 3.3757421543681088,
      "grad_norm": 1.8125,
      "learning_rate": 0.00039653560574997383,
      "loss": 4.5596,
      "mean_token_accuracy": 0.2479489579796791,
      "num_tokens": 9518815.0,
      "step": 11940
    },
    {
      "entropy": 4.769192552566528,
      "epoch": 3.3771557817359343,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003964490292926288,
      "loss": 4.5346,
      "mean_token_accuracy": 0.25109580159187317,
      "num_tokens": 9522736.0,
      "step": 11945
    },
    {
      "entropy": 4.731974744796753,
      "epoch": 3.3785694091037604,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003963624274531877,
      "loss": 4.562,
      "mean_token_accuracy": 0.25398544073104856,
      "num_tokens": 9526995.0,
      "step": 11950
    },
    {
      "entropy": 4.681202220916748,
      "epoch": 3.379983036471586,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003962758002497497,
      "loss": 4.4954,
      "mean_token_accuracy": 0.2581408962607384,
      "num_tokens": 9531271.0,
      "step": 11955
    },
    {
      "entropy": 4.700651025772094,
      "epoch": 3.381396663839412,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00039618914770041936,
      "loss": 4.5308,
      "mean_token_accuracy": 0.2609393820166588,
      "num_tokens": 9535337.0,
      "step": 11960
    },
    {
      "entropy": 4.704057025909424,
      "epoch": 3.382810291207238,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003961024698233067,
      "loss": 4.4556,
      "mean_token_accuracy": 0.25774316787719725,
      "num_tokens": 9539526.0,
      "step": 11965
    },
    {
      "entropy": 4.653316974639893,
      "epoch": 3.3842239185750635,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003960157666365265,
      "loss": 4.4848,
      "mean_token_accuracy": 0.24914825111627578,
      "num_tokens": 9543563.0,
      "step": 11970
    },
    {
      "entropy": 4.66391921043396,
      "epoch": 3.3856375459428896,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003959290381581996,
      "loss": 4.5161,
      "mean_token_accuracy": 0.25648744255304334,
      "num_tokens": 9547389.0,
      "step": 11975
    },
    {
      "entropy": 4.794810485839844,
      "epoch": 3.387051173310715,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003958422844064514,
      "loss": 4.6429,
      "mean_token_accuracy": 0.24404185265302658,
      "num_tokens": 9551721.0,
      "step": 11980
    },
    {
      "entropy": 4.797810363769531,
      "epoch": 3.388464800678541,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00039575550539941315,
      "loss": 4.6472,
      "mean_token_accuracy": 0.2539676994085312,
      "num_tokens": 9555610.0,
      "step": 11985
    },
    {
      "entropy": 4.8122124671936035,
      "epoch": 3.3898784280463667,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003956687011552209,
      "loss": 4.5577,
      "mean_token_accuracy": 0.24337861686944962,
      "num_tokens": 9559618.0,
      "step": 11990
    },
    {
      "entropy": 4.726791524887085,
      "epoch": 3.3912920554141928,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00039558187169201636,
      "loss": 4.5482,
      "mean_token_accuracy": 0.25508700907230375,
      "num_tokens": 9563717.0,
      "step": 11995
    },
    {
      "entropy": 4.723432779312134,
      "epoch": 3.392705682782019,
      "grad_norm": 1.765625,
      "learning_rate": 0.00039549501702794616,
      "loss": 4.4795,
      "mean_token_accuracy": 0.2645320162177086,
      "num_tokens": 9567550.0,
      "step": 12000
    },
    {
      "epoch": 3.392705682782019,
      "eval_entropy": 4.617690959101784,
      "eval_loss": 5.092161655426025,
      "eval_mean_token_accuracy": 0.2396124096615022,
      "eval_num_tokens": 9567550.0,
      "eval_runtime": 6.6692,
      "eval_samples_per_second": 1691.958,
      "eval_steps_per_second": 211.57,
      "step": 12000
    },
    {
      "entropy": 4.7671009540557865,
      "epoch": 3.3941193101498444,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003954081371811627,
      "loss": 4.607,
      "mean_token_accuracy": 0.24825267344713212,
      "num_tokens": 9571316.0,
      "step": 12005
    },
    {
      "entropy": 4.66207709312439,
      "epoch": 3.3955329375176704,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003953212321698231,
      "loss": 4.4848,
      "mean_token_accuracy": 0.26309034526348113,
      "num_tokens": 9575231.0,
      "step": 12010
    },
    {
      "entropy": 4.689404535293579,
      "epoch": 3.3969465648854964,
      "grad_norm": 1.6875,
      "learning_rate": 0.00039523430201209,
      "loss": 4.4926,
      "mean_token_accuracy": 0.2501275524497032,
      "num_tokens": 9579180.0,
      "step": 12015
    },
    {
      "entropy": 4.750282573699951,
      "epoch": 3.398360192253322,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00039514734672613125,
      "loss": 4.5346,
      "mean_token_accuracy": 0.2474864214658737,
      "num_tokens": 9583171.0,
      "step": 12020
    },
    {
      "entropy": 4.635117626190185,
      "epoch": 3.399773819621148,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00039506036633012,
      "loss": 4.4643,
      "mean_token_accuracy": 0.260587203502655,
      "num_tokens": 9586780.0,
      "step": 12025
    },
    {
      "entropy": 4.768659782409668,
      "epoch": 3.4011874469889736,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003949733608422347,
      "loss": 4.5541,
      "mean_token_accuracy": 0.24761455357074738,
      "num_tokens": 9591225.0,
      "step": 12030
    },
    {
      "entropy": 4.770403003692627,
      "epoch": 3.4026010743567996,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00039488633028065883,
      "loss": 4.5722,
      "mean_token_accuracy": 0.2507979333400726,
      "num_tokens": 9595179.0,
      "step": 12035
    },
    {
      "entropy": 4.736874675750732,
      "epoch": 3.404014701724625,
      "grad_norm": 1.609375,
      "learning_rate": 0.00039479927466358125,
      "loss": 4.6494,
      "mean_token_accuracy": 0.2534463584423065,
      "num_tokens": 9599325.0,
      "step": 12040
    },
    {
      "entropy": 4.79804253578186,
      "epoch": 3.405428329092451,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003947121940091962,
      "loss": 4.6186,
      "mean_token_accuracy": 0.24191030263900756,
      "num_tokens": 9603760.0,
      "step": 12045
    },
    {
      "entropy": 4.712527084350586,
      "epoch": 3.406841956460277,
      "grad_norm": 1.578125,
      "learning_rate": 0.00039462508833570275,
      "loss": 4.5042,
      "mean_token_accuracy": 0.2526901736855507,
      "num_tokens": 9607826.0,
      "step": 12050
    },
    {
      "entropy": 4.775414657592774,
      "epoch": 3.408255583828103,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00039453795766130565,
      "loss": 4.6037,
      "mean_token_accuracy": 0.2320679321885109,
      "num_tokens": 9611859.0,
      "step": 12055
    },
    {
      "entropy": 4.826092386245728,
      "epoch": 3.409669211195929,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003944508020042145,
      "loss": 4.6269,
      "mean_token_accuracy": 0.24139921367168427,
      "num_tokens": 9615938.0,
      "step": 12060
    },
    {
      "entropy": 4.836429786682129,
      "epoch": 3.411082838563755,
      "grad_norm": 1.8125,
      "learning_rate": 0.0003943636213826445,
      "loss": 4.6022,
      "mean_token_accuracy": 0.2541595220565796,
      "num_tokens": 9619722.0,
      "step": 12065
    },
    {
      "entropy": 4.64605941772461,
      "epoch": 3.4124964659315804,
      "grad_norm": 1.765625,
      "learning_rate": 0.00039427641581481573,
      "loss": 4.5525,
      "mean_token_accuracy": 0.255465205013752,
      "num_tokens": 9623623.0,
      "step": 12070
    },
    {
      "entropy": 4.766227340698242,
      "epoch": 3.4139100932994064,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003941891853189536,
      "loss": 4.6688,
      "mean_token_accuracy": 0.24336500763893126,
      "num_tokens": 9627515.0,
      "step": 12075
    },
    {
      "entropy": 4.749230766296387,
      "epoch": 3.415323720667232,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039410192991328884,
      "loss": 4.4954,
      "mean_token_accuracy": 0.2546343207359314,
      "num_tokens": 9631513.0,
      "step": 12080
    },
    {
      "entropy": 4.782394456863403,
      "epoch": 3.416737348035058,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003940146496160573,
      "loss": 4.5588,
      "mean_token_accuracy": 0.2489120528101921,
      "num_tokens": 9635425.0,
      "step": 12085
    },
    {
      "entropy": 4.7441911697387695,
      "epoch": 3.4181509754028836,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039392734444549976,
      "loss": 4.4869,
      "mean_token_accuracy": 0.25672404319047926,
      "num_tokens": 9639144.0,
      "step": 12090
    },
    {
      "entropy": 4.791217279434204,
      "epoch": 3.4195646027707096,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003938400144198628,
      "loss": 4.5629,
      "mean_token_accuracy": 0.2587124854326248,
      "num_tokens": 9643492.0,
      "step": 12095
    },
    {
      "entropy": 4.671883249282837,
      "epoch": 3.4209782301385356,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003937526595573976,
      "loss": 4.3622,
      "mean_token_accuracy": 0.2827158570289612,
      "num_tokens": 9647510.0,
      "step": 12100
    },
    {
      "entropy": 4.747403621673584,
      "epoch": 3.422391857506361,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00039366527987636104,
      "loss": 4.6292,
      "mean_token_accuracy": 0.2552821084856987,
      "num_tokens": 9651586.0,
      "step": 12105
    },
    {
      "entropy": 4.715870761871338,
      "epoch": 3.4238054848741872,
      "grad_norm": 1.6875,
      "learning_rate": 0.00039357787539501466,
      "loss": 4.668,
      "mean_token_accuracy": 0.24073173701763154,
      "num_tokens": 9655567.0,
      "step": 12110
    },
    {
      "entropy": 4.860812759399414,
      "epoch": 3.4252191122420133,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00039349044613162565,
      "loss": 4.7146,
      "mean_token_accuracy": 0.2340009406208992,
      "num_tokens": 9659271.0,
      "step": 12115
    },
    {
      "entropy": 4.818365478515625,
      "epoch": 3.426632739609839,
      "grad_norm": 1.5625,
      "learning_rate": 0.00039340299210446607,
      "loss": 4.5618,
      "mean_token_accuracy": 0.25082509517669677,
      "num_tokens": 9663145.0,
      "step": 12120
    },
    {
      "entropy": 4.828974294662475,
      "epoch": 3.428046366977665,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00039331551333181335,
      "loss": 4.5849,
      "mean_token_accuracy": 0.24261448234319688,
      "num_tokens": 9666806.0,
      "step": 12125
    },
    {
      "entropy": 4.701805496215821,
      "epoch": 3.4294599943454904,
      "grad_norm": 1.734375,
      "learning_rate": 0.00039322800983195,
      "loss": 4.4541,
      "mean_token_accuracy": 0.2554605662822723,
      "num_tokens": 9670619.0,
      "step": 12130
    },
    {
      "entropy": 4.6122880458831785,
      "epoch": 3.4308736217133164,
      "grad_norm": 1.6875,
      "learning_rate": 0.00039314048162316376,
      "loss": 4.4826,
      "mean_token_accuracy": 0.2643768638372421,
      "num_tokens": 9674600.0,
      "step": 12135
    },
    {
      "entropy": 4.672778511047364,
      "epoch": 3.432287249081142,
      "grad_norm": 1.546875,
      "learning_rate": 0.00039305292872374734,
      "loss": 4.5527,
      "mean_token_accuracy": 0.2467150643467903,
      "num_tokens": 9678535.0,
      "step": 12140
    },
    {
      "entropy": 4.68914828300476,
      "epoch": 3.433700876448968,
      "grad_norm": 1.59375,
      "learning_rate": 0.00039296535115199887,
      "loss": 4.52,
      "mean_token_accuracy": 0.2683263197541237,
      "num_tokens": 9682428.0,
      "step": 12145
    },
    {
      "entropy": 4.803245735168457,
      "epoch": 3.435114503816794,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003928777489262214,
      "loss": 4.6751,
      "mean_token_accuracy": 0.24007664918899535,
      "num_tokens": 9686497.0,
      "step": 12150
    },
    {
      "entropy": 4.711838245391846,
      "epoch": 3.4365281311846196,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003927901220647234,
      "loss": 4.5529,
      "mean_token_accuracy": 0.26818049103021624,
      "num_tokens": 9690468.0,
      "step": 12155
    },
    {
      "entropy": 4.724407482147217,
      "epoch": 3.4379417585524457,
      "grad_norm": 1.796875,
      "learning_rate": 0.0003927024705858183,
      "loss": 4.4343,
      "mean_token_accuracy": 0.26411320716142656,
      "num_tokens": 9694041.0,
      "step": 12160
    },
    {
      "entropy": 4.625420093536377,
      "epoch": 3.4393553859202712,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00039261479450782466,
      "loss": 4.4004,
      "mean_token_accuracy": 0.2657060816884041,
      "num_tokens": 9698094.0,
      "step": 12165
    },
    {
      "entropy": 4.763422966003418,
      "epoch": 3.4407690132880973,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003925270938490663,
      "loss": 4.4933,
      "mean_token_accuracy": 0.2662922278046608,
      "num_tokens": 9701868.0,
      "step": 12170
    },
    {
      "entropy": 4.62923674583435,
      "epoch": 3.4421826406559233,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003924393686278721,
      "loss": 4.486,
      "mean_token_accuracy": 0.26520304530858996,
      "num_tokens": 9705912.0,
      "step": 12175
    },
    {
      "entropy": 4.693800544738769,
      "epoch": 3.443596268023749,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000392351618862576,
      "loss": 4.4811,
      "mean_token_accuracy": 0.2566176652908325,
      "num_tokens": 9709893.0,
      "step": 12180
    },
    {
      "entropy": 4.766989850997925,
      "epoch": 3.445009895391575,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003922638445715174,
      "loss": 4.5776,
      "mean_token_accuracy": 0.2475569173693657,
      "num_tokens": 9713786.0,
      "step": 12185
    },
    {
      "entropy": 4.730065536499024,
      "epoch": 3.4464235227594004,
      "grad_norm": 1.796875,
      "learning_rate": 0.0003921760457730403,
      "loss": 4.4769,
      "mean_token_accuracy": 0.2581873133778572,
      "num_tokens": 9717914.0,
      "step": 12190
    },
    {
      "entropy": 4.791991138458252,
      "epoch": 3.4478371501272265,
      "grad_norm": 1.625,
      "learning_rate": 0.0003920882224854942,
      "loss": 4.6014,
      "mean_token_accuracy": 0.23804907351732255,
      "num_tokens": 9721901.0,
      "step": 12195
    },
    {
      "entropy": 4.757010746002197,
      "epoch": 3.4492507774950525,
      "grad_norm": 1.5625,
      "learning_rate": 0.00039200037472723377,
      "loss": 4.5986,
      "mean_token_accuracy": 0.24626206159591674,
      "num_tokens": 9726371.0,
      "step": 12200
    },
    {
      "entropy": 4.7238757610321045,
      "epoch": 3.450664404862878,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003919125025166185,
      "loss": 4.5166,
      "mean_token_accuracy": 0.25165210068225863,
      "num_tokens": 9730501.0,
      "step": 12205
    },
    {
      "entropy": 4.704705619812012,
      "epoch": 3.452078032230704,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003918246058720131,
      "loss": 4.4382,
      "mean_token_accuracy": 0.25268374383449554,
      "num_tokens": 9734500.0,
      "step": 12210
    },
    {
      "entropy": 4.701063442230224,
      "epoch": 3.4534916595985297,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003917366848117875,
      "loss": 4.5779,
      "mean_token_accuracy": 0.25046837627887725,
      "num_tokens": 9738258.0,
      "step": 12215
    },
    {
      "entropy": 4.805082321166992,
      "epoch": 3.4549052869663557,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003916487393543167,
      "loss": 4.6896,
      "mean_token_accuracy": 0.23905782997608185,
      "num_tokens": 9742373.0,
      "step": 12220
    },
    {
      "entropy": 4.705698919296265,
      "epoch": 3.4563189143341817,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00039156076951798064,
      "loss": 4.5456,
      "mean_token_accuracy": 0.2550085410475731,
      "num_tokens": 9746296.0,
      "step": 12225
    },
    {
      "entropy": 4.721251630783081,
      "epoch": 3.4577325417020073,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003914727753211647,
      "loss": 4.6348,
      "mean_token_accuracy": 0.24618425220251083,
      "num_tokens": 9750545.0,
      "step": 12230
    },
    {
      "entropy": 4.744362497329712,
      "epoch": 3.4591461690698333,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003913847567822588,
      "loss": 4.5166,
      "mean_token_accuracy": 0.25057742893695834,
      "num_tokens": 9754504.0,
      "step": 12235
    },
    {
      "entropy": 4.751663398742676,
      "epoch": 3.460559796437659,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003912967139196585,
      "loss": 4.5872,
      "mean_token_accuracy": 0.24895195364952089,
      "num_tokens": 9758878.0,
      "step": 12240
    },
    {
      "entropy": 4.684665012359619,
      "epoch": 3.461973423805485,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00039120864675176416,
      "loss": 4.4683,
      "mean_token_accuracy": 0.2631945997476578,
      "num_tokens": 9762818.0,
      "step": 12245
    },
    {
      "entropy": 4.672690439224243,
      "epoch": 3.463387051173311,
      "grad_norm": 1.640625,
      "learning_rate": 0.00039112055529698126,
      "loss": 4.4171,
      "mean_token_accuracy": 0.261438949406147,
      "num_tokens": 9766789.0,
      "step": 12250
    },
    {
      "entropy": 4.80817551612854,
      "epoch": 3.4648006785411365,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003910324395737203,
      "loss": 4.6682,
      "mean_token_accuracy": 0.24912968128919602,
      "num_tokens": 9770721.0,
      "step": 12255
    },
    {
      "entropy": 4.663032817840576,
      "epoch": 3.4662143059089625,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00039094429960039707,
      "loss": 4.446,
      "mean_token_accuracy": 0.26136722713708876,
      "num_tokens": 9774568.0,
      "step": 12260
    },
    {
      "entropy": 4.723871278762817,
      "epoch": 3.467627933276788,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003908561353954322,
      "loss": 4.4311,
      "mean_token_accuracy": 0.2617762118577957,
      "num_tokens": 9778694.0,
      "step": 12265
    },
    {
      "entropy": 4.72531099319458,
      "epoch": 3.469041560644614,
      "grad_norm": 1.828125,
      "learning_rate": 0.0003907679469772514,
      "loss": 4.5407,
      "mean_token_accuracy": 0.24882714003324508,
      "num_tokens": 9782501.0,
      "step": 12270
    },
    {
      "entropy": 4.687029600143433,
      "epoch": 3.4704551880124397,
      "grad_norm": 1.75,
      "learning_rate": 0.00039067973436428553,
      "loss": 4.4902,
      "mean_token_accuracy": 0.2584337621927261,
      "num_tokens": 9786410.0,
      "step": 12275
    },
    {
      "entropy": 4.731009244918823,
      "epoch": 3.4718688153802657,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003905914975749707,
      "loss": 4.6618,
      "mean_token_accuracy": 0.23771025240421295,
      "num_tokens": 9790575.0,
      "step": 12280
    },
    {
      "entropy": 4.793943595886231,
      "epoch": 3.4732824427480917,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003905032366277475,
      "loss": 4.5176,
      "mean_token_accuracy": 0.25024998337030413,
      "num_tokens": 9794440.0,
      "step": 12285
    },
    {
      "entropy": 4.914901447296143,
      "epoch": 3.4746960701159173,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003904149515410622,
      "loss": 4.7379,
      "mean_token_accuracy": 0.24119525998830796,
      "num_tokens": 9798287.0,
      "step": 12290
    },
    {
      "entropy": 4.78026819229126,
      "epoch": 3.4761096974837433,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003903266423333657,
      "loss": 4.5112,
      "mean_token_accuracy": 0.2564535677433014,
      "num_tokens": 9802177.0,
      "step": 12295
    },
    {
      "entropy": 4.770306348800659,
      "epoch": 3.4775233248515693,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003902383090231141,
      "loss": 4.5848,
      "mean_token_accuracy": 0.24870537966489792,
      "num_tokens": 9805906.0,
      "step": 12300
    },
    {
      "entropy": 4.750717210769653,
      "epoch": 3.478936952219395,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003901499516287685,
      "loss": 4.597,
      "mean_token_accuracy": 0.23993682563304902,
      "num_tokens": 9809599.0,
      "step": 12305
    },
    {
      "entropy": 4.757948589324951,
      "epoch": 3.480350579587221,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003900615701687952,
      "loss": 4.4135,
      "mean_token_accuracy": 0.2608377367258072,
      "num_tokens": 9813439.0,
      "step": 12310
    },
    {
      "entropy": 4.658061933517456,
      "epoch": 3.4817642069550465,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003899731646616651,
      "loss": 4.5079,
      "mean_token_accuracy": 0.25779037475585936,
      "num_tokens": 9817386.0,
      "step": 12315
    },
    {
      "entropy": 4.779843425750732,
      "epoch": 3.4831778343228725,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003898847351258547,
      "loss": 4.6327,
      "mean_token_accuracy": 0.23872901797294616,
      "num_tokens": 9821589.0,
      "step": 12320
    },
    {
      "entropy": 4.752112245559692,
      "epoch": 3.484591461690698,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003897962815798449,
      "loss": 4.5344,
      "mean_token_accuracy": 0.2554395243525505,
      "num_tokens": 9825671.0,
      "step": 12325
    },
    {
      "entropy": 4.754328107833862,
      "epoch": 3.486005089058524,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003897078040421223,
      "loss": 4.5779,
      "mean_token_accuracy": 0.24378155022859574,
      "num_tokens": 9830123.0,
      "step": 12330
    },
    {
      "entropy": 4.683629512786865,
      "epoch": 3.48741871642635,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003896193025311779,
      "loss": 4.504,
      "mean_token_accuracy": 0.26248418241739274,
      "num_tokens": 9834281.0,
      "step": 12335
    },
    {
      "entropy": 4.722589302062988,
      "epoch": 3.4888323437941757,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003895307770655081,
      "loss": 4.5298,
      "mean_token_accuracy": 0.2503043204545975,
      "num_tokens": 9837707.0,
      "step": 12340
    },
    {
      "entropy": 4.7183225631713865,
      "epoch": 3.4902459711620017,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003894422276636141,
      "loss": 4.5144,
      "mean_token_accuracy": 0.25981279611587527,
      "num_tokens": 9841837.0,
      "step": 12345
    },
    {
      "entropy": 4.702732086181641,
      "epoch": 3.4916595985298278,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003893536543440022,
      "loss": 4.4361,
      "mean_token_accuracy": 0.26140045672655104,
      "num_tokens": 9846199.0,
      "step": 12350
    },
    {
      "entropy": 4.6832848072052,
      "epoch": 3.4930732258976533,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003892650571251838,
      "loss": 4.4714,
      "mean_token_accuracy": 0.2590473771095276,
      "num_tokens": 9850477.0,
      "step": 12355
    },
    {
      "entropy": 4.675268888473511,
      "epoch": 3.4944868532654794,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003891764360256749,
      "loss": 4.4981,
      "mean_token_accuracy": 0.2585751831531525,
      "num_tokens": 9854639.0,
      "step": 12360
    },
    {
      "entropy": 4.803682518005371,
      "epoch": 3.495900480633305,
      "grad_norm": 1.75,
      "learning_rate": 0.00038908779106399695,
      "loss": 4.7051,
      "mean_token_accuracy": 0.24189767092466355,
      "num_tokens": 9858265.0,
      "step": 12365
    },
    {
      "entropy": 4.6827044010162355,
      "epoch": 3.497314108001131,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003889991222586762,
      "loss": 4.4387,
      "mean_token_accuracy": 0.27309798896312715,
      "num_tokens": 9862176.0,
      "step": 12370
    },
    {
      "entropy": 4.707692241668701,
      "epoch": 3.4987277353689565,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00038891042962824386,
      "loss": 4.5275,
      "mean_token_accuracy": 0.24576974362134935,
      "num_tokens": 9866058.0,
      "step": 12375
    },
    {
      "entropy": 4.758933877944946,
      "epoch": 3.5001413627367826,
      "grad_norm": 1.6875,
      "learning_rate": 0.00038882171319123607,
      "loss": 4.5505,
      "mean_token_accuracy": 0.2576568901538849,
      "num_tokens": 9870000.0,
      "step": 12380
    },
    {
      "entropy": 4.693151187896729,
      "epoch": 3.5015549901046086,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003887329729661941,
      "loss": 4.5229,
      "mean_token_accuracy": 0.26097756773233416,
      "num_tokens": 9873797.0,
      "step": 12385
    },
    {
      "entropy": 4.676103830337524,
      "epoch": 3.502968617472434,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00038864420897166414,
      "loss": 4.3908,
      "mean_token_accuracy": 0.2630652472376823,
      "num_tokens": 9877808.0,
      "step": 12390
    },
    {
      "entropy": 4.7025140762329105,
      "epoch": 3.50438224484026,
      "grad_norm": 1.765625,
      "learning_rate": 0.00038855542122619716,
      "loss": 4.5425,
      "mean_token_accuracy": 0.2565742269158363,
      "num_tokens": 9881640.0,
      "step": 12395
    },
    {
      "entropy": 4.7160101413726805,
      "epoch": 3.505795872208086,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003884666097483495,
      "loss": 4.5481,
      "mean_token_accuracy": 0.2620694264769554,
      "num_tokens": 9885834.0,
      "step": 12400
    },
    {
      "entropy": 4.758160066604614,
      "epoch": 3.5072094995759118,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00038837777455668184,
      "loss": 4.5777,
      "mean_token_accuracy": 0.2427653893828392,
      "num_tokens": 9890240.0,
      "step": 12405
    },
    {
      "entropy": 4.711787462234497,
      "epoch": 3.508623126943738,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003882889156697606,
      "loss": 4.506,
      "mean_token_accuracy": 0.26320191025733947,
      "num_tokens": 9893862.0,
      "step": 12410
    },
    {
      "entropy": 4.650814580917358,
      "epoch": 3.5100367543115634,
      "grad_norm": 1.671875,
      "learning_rate": 0.00038820003310615644,
      "loss": 4.4405,
      "mean_token_accuracy": 0.2548428371548653,
      "num_tokens": 9897552.0,
      "step": 12415
    },
    {
      "entropy": 4.792955827713013,
      "epoch": 3.5114503816793894,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003881111268844454,
      "loss": 4.6466,
      "mean_token_accuracy": 0.24090384542942048,
      "num_tokens": 9901634.0,
      "step": 12420
    },
    {
      "entropy": 4.747173166275024,
      "epoch": 3.512864009047215,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003880221970232083,
      "loss": 4.5755,
      "mean_token_accuracy": 0.2401556193828583,
      "num_tokens": 9905741.0,
      "step": 12425
    },
    {
      "entropy": 4.671509695053101,
      "epoch": 3.514277636415041,
      "grad_norm": 1.625,
      "learning_rate": 0.0003879332435410309,
      "loss": 4.4466,
      "mean_token_accuracy": 0.25544962882995603,
      "num_tokens": 9909563.0,
      "step": 12430
    },
    {
      "entropy": 4.760368871688843,
      "epoch": 3.515691263782867,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003878442664565039,
      "loss": 4.5179,
      "mean_token_accuracy": 0.2566922456026077,
      "num_tokens": 9913750.0,
      "step": 12435
    },
    {
      "entropy": 4.79218168258667,
      "epoch": 3.5171048911506926,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00038775526578822323,
      "loss": 4.4815,
      "mean_token_accuracy": 0.2580343335866928,
      "num_tokens": 9917653.0,
      "step": 12440
    },
    {
      "entropy": 4.710677003860473,
      "epoch": 3.5185185185185186,
      "grad_norm": 1.53125,
      "learning_rate": 0.00038766624155478914,
      "loss": 4.5829,
      "mean_token_accuracy": 0.2522234350442886,
      "num_tokens": 9921579.0,
      "step": 12445
    },
    {
      "entropy": 4.784183073043823,
      "epoch": 3.5199321458863446,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003875771937748073,
      "loss": 4.6245,
      "mean_token_accuracy": 0.23820348381996154,
      "num_tokens": 9925545.0,
      "step": 12450
    },
    {
      "entropy": 4.716958951950073,
      "epoch": 3.52134577325417,
      "grad_norm": 1.640625,
      "learning_rate": 0.00038748812246688804,
      "loss": 4.5285,
      "mean_token_accuracy": 0.2574568331241608,
      "num_tokens": 9929682.0,
      "step": 12455
    },
    {
      "entropy": 4.7001604557037355,
      "epoch": 3.5227594006219958,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003873990276496468,
      "loss": 4.5991,
      "mean_token_accuracy": 0.2527391642332077,
      "num_tokens": 9933488.0,
      "step": 12460
    },
    {
      "entropy": 4.842548608779907,
      "epoch": 3.524173027989822,
      "grad_norm": 1.75,
      "learning_rate": 0.0003873099093417039,
      "loss": 4.605,
      "mean_token_accuracy": 0.2517306759953499,
      "num_tokens": 9937470.0,
      "step": 12465
    },
    {
      "entropy": 4.685521793365479,
      "epoch": 3.525586655357648,
      "grad_norm": 1.640625,
      "learning_rate": 0.00038722076756168443,
      "loss": 4.462,
      "mean_token_accuracy": 0.26418753564357755,
      "num_tokens": 9941139.0,
      "step": 12470
    },
    {
      "entropy": 4.714101791381836,
      "epoch": 3.5270002827254734,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003871316023282185,
      "loss": 4.5808,
      "mean_token_accuracy": 0.2552934303879738,
      "num_tokens": 9944992.0,
      "step": 12475
    },
    {
      "entropy": 4.751831769943237,
      "epoch": 3.5284139100932994,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003870424136599411,
      "loss": 4.6156,
      "mean_token_accuracy": 0.2501934438943863,
      "num_tokens": 9949132.0,
      "step": 12480
    },
    {
      "entropy": 4.754928255081177,
      "epoch": 3.5298275374611254,
      "grad_norm": 1.609375,
      "learning_rate": 0.000386953201575492,
      "loss": 4.5988,
      "mean_token_accuracy": 0.2370246022939682,
      "num_tokens": 9952928.0,
      "step": 12485
    },
    {
      "entropy": 4.818630409240723,
      "epoch": 3.531241164828951,
      "grad_norm": 1.65625,
      "learning_rate": 0.00038686396609351615,
      "loss": 4.5912,
      "mean_token_accuracy": 0.24843545109033585,
      "num_tokens": 9957000.0,
      "step": 12490
    },
    {
      "entropy": 4.764599990844727,
      "epoch": 3.532654792196777,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003867747072326631,
      "loss": 4.5883,
      "mean_token_accuracy": 0.2547958046197891,
      "num_tokens": 9960738.0,
      "step": 12495
    },
    {
      "entropy": 4.678762626647949,
      "epoch": 3.534068419564603,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003866854250115875,
      "loss": 4.4973,
      "mean_token_accuracy": 0.25824659019708635,
      "num_tokens": 9964756.0,
      "step": 12500
    },
    {
      "entropy": 4.655408334732056,
      "epoch": 3.5354820469324286,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003865961194489487,
      "loss": 4.4305,
      "mean_token_accuracy": 0.2644162356853485,
      "num_tokens": 9968835.0,
      "step": 12505
    },
    {
      "entropy": 4.678800249099732,
      "epoch": 3.536895674300254,
      "grad_norm": 1.65625,
      "learning_rate": 0.00038650679056341096,
      "loss": 4.5615,
      "mean_token_accuracy": 0.24491141885519027,
      "num_tokens": 9972721.0,
      "step": 12510
    },
    {
      "entropy": 4.811769676208496,
      "epoch": 3.53830930166808,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00038641743837364355,
      "loss": 4.6498,
      "mean_token_accuracy": 0.2420573890209198,
      "num_tokens": 9976758.0,
      "step": 12515
    },
    {
      "entropy": 4.721271276473999,
      "epoch": 3.5397229290359062,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003863280628983205,
      "loss": 4.5334,
      "mean_token_accuracy": 0.26362267732620237,
      "num_tokens": 9980506.0,
      "step": 12520
    },
    {
      "entropy": 4.7006213665008545,
      "epoch": 3.541136556403732,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003862386641561208,
      "loss": 4.5114,
      "mean_token_accuracy": 0.25782876312732694,
      "num_tokens": 9984338.0,
      "step": 12525
    },
    {
      "entropy": 4.781050014495849,
      "epoch": 3.542550183771558,
      "grad_norm": 1.5625,
      "learning_rate": 0.00038614924216572803,
      "loss": 4.5568,
      "mean_token_accuracy": 0.25693943053483964,
      "num_tokens": 9988196.0,
      "step": 12530
    },
    {
      "entropy": 4.70241379737854,
      "epoch": 3.543963811139384,
      "grad_norm": 1.6875,
      "learning_rate": 0.00038605979694583117,
      "loss": 4.4693,
      "mean_token_accuracy": 0.26669895350933076,
      "num_tokens": 9992045.0,
      "step": 12535
    },
    {
      "entropy": 4.7372771263122555,
      "epoch": 3.5453774385072094,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003859703285151234,
      "loss": 4.6088,
      "mean_token_accuracy": 0.23730184137821198,
      "num_tokens": 9996026.0,
      "step": 12540
    },
    {
      "entropy": 4.693777799606323,
      "epoch": 3.5467910658750355,
      "grad_norm": 1.625,
      "learning_rate": 0.0003858808368923032,
      "loss": 4.5885,
      "mean_token_accuracy": 0.24868002086877822,
      "num_tokens": 9999778.0,
      "step": 12545
    },
    {
      "entropy": 4.701164245605469,
      "epoch": 3.548204693242861,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003857913220960737,
      "loss": 4.4865,
      "mean_token_accuracy": 0.2617734372615814,
      "num_tokens": 10003978.0,
      "step": 12550
    },
    {
      "entropy": 4.819319391250611,
      "epoch": 3.549618320610687,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00038570178414514314,
      "loss": 4.5265,
      "mean_token_accuracy": 0.2668521165847778,
      "num_tokens": 10007870.0,
      "step": 12555
    },
    {
      "entropy": 4.801152563095092,
      "epoch": 3.5510319479785126,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00038561222305822415,
      "loss": 4.6564,
      "mean_token_accuracy": 0.2332952633500099,
      "num_tokens": 10011865.0,
      "step": 12560
    },
    {
      "entropy": 4.591542434692383,
      "epoch": 3.5524455753463386,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003855226388540345,
      "loss": 4.4043,
      "mean_token_accuracy": 0.26575944721698763,
      "num_tokens": 10015840.0,
      "step": 12565
    },
    {
      "entropy": 4.710676860809326,
      "epoch": 3.5538592027141647,
      "grad_norm": 1.625,
      "learning_rate": 0.0003854330315512969,
      "loss": 4.5542,
      "mean_token_accuracy": 0.2604998230934143,
      "num_tokens": 10020110.0,
      "step": 12570
    },
    {
      "entropy": 4.765235471725464,
      "epoch": 3.5552728300819902,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003853434011687386,
      "loss": 4.5442,
      "mean_token_accuracy": 0.2585788980126381,
      "num_tokens": 10024246.0,
      "step": 12575
    },
    {
      "entropy": 4.791489028930664,
      "epoch": 3.5566864574498163,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003852537477250917,
      "loss": 4.5889,
      "mean_token_accuracy": 0.25171988904476167,
      "num_tokens": 10028499.0,
      "step": 12580
    },
    {
      "entropy": 4.597690391540527,
      "epoch": 3.5581000848176423,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00038516407123909336,
      "loss": 4.4512,
      "mean_token_accuracy": 0.2698229044675827,
      "num_tokens": 10032423.0,
      "step": 12585
    },
    {
      "entropy": 4.670028591156006,
      "epoch": 3.559513712185468,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00038507437172948536,
      "loss": 4.5715,
      "mean_token_accuracy": 0.25499300807714465,
      "num_tokens": 10036201.0,
      "step": 12590
    },
    {
      "entropy": 4.796184444427491,
      "epoch": 3.560927339553294,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003849846492150144,
      "loss": 4.686,
      "mean_token_accuracy": 0.24565829187631608,
      "num_tokens": 10040123.0,
      "step": 12595
    },
    {
      "entropy": 4.814067602157593,
      "epoch": 3.5623409669211195,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003848949037144318,
      "loss": 4.6518,
      "mean_token_accuracy": 0.2465318039059639,
      "num_tokens": 10044074.0,
      "step": 12600
    },
    {
      "entropy": 4.867226505279541,
      "epoch": 3.5637545942889455,
      "grad_norm": 1.5625,
      "learning_rate": 0.00038480513524649396,
      "loss": 4.6392,
      "mean_token_accuracy": 0.24264366626739503,
      "num_tokens": 10048114.0,
      "step": 12605
    },
    {
      "entropy": 4.83232421875,
      "epoch": 3.565168221656771,
      "grad_norm": 1.71875,
      "learning_rate": 0.00038471534382996184,
      "loss": 4.6996,
      "mean_token_accuracy": 0.24229566752910614,
      "num_tokens": 10052145.0,
      "step": 12610
    },
    {
      "entropy": 4.646299219131469,
      "epoch": 3.566581849024597,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003846255294836013,
      "loss": 4.4741,
      "mean_token_accuracy": 0.25151547491550447,
      "num_tokens": 10056033.0,
      "step": 12615
    },
    {
      "entropy": 4.6217326641082765,
      "epoch": 3.567995476392423,
      "grad_norm": 1.625,
      "learning_rate": 0.00038453569222618304,
      "loss": 4.4536,
      "mean_token_accuracy": 0.25374812334775926,
      "num_tokens": 10059757.0,
      "step": 12620
    },
    {
      "entropy": 4.68637523651123,
      "epoch": 3.5694091037602487,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0003844458320764824,
      "loss": 4.4837,
      "mean_token_accuracy": 0.2634815827012062,
      "num_tokens": 10063995.0,
      "step": 12625
    },
    {
      "entropy": 4.746660327911377,
      "epoch": 3.5708227311280747,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00038435594905327966,
      "loss": 4.5302,
      "mean_token_accuracy": 0.2476650059223175,
      "num_tokens": 10068014.0,
      "step": 12630
    },
    {
      "entropy": 4.87226037979126,
      "epoch": 3.5722363584959007,
      "grad_norm": 1.71875,
      "learning_rate": 0.00038426604317535985,
      "loss": 4.6403,
      "mean_token_accuracy": 0.23977744430303574,
      "num_tokens": 10072186.0,
      "step": 12635
    },
    {
      "entropy": 4.832816028594971,
      "epoch": 3.5736499858637263,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003841761144615126,
      "loss": 4.5965,
      "mean_token_accuracy": 0.24705232828855514,
      "num_tokens": 10076152.0,
      "step": 12640
    },
    {
      "entropy": 4.755417680740356,
      "epoch": 3.5750636132315523,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003840861629305327,
      "loss": 4.5605,
      "mean_token_accuracy": 0.25313246101140974,
      "num_tokens": 10079844.0,
      "step": 12645
    },
    {
      "entropy": 4.675730800628662,
      "epoch": 3.576477240599378,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00038399618860121917,
      "loss": 4.5119,
      "mean_token_accuracy": 0.26151220202445985,
      "num_tokens": 10083674.0,
      "step": 12650
    },
    {
      "entropy": 4.6807678699493405,
      "epoch": 3.577890867967204,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003839061914923763,
      "loss": 4.5669,
      "mean_token_accuracy": 0.23931439518928527,
      "num_tokens": 10087641.0,
      "step": 12655
    },
    {
      "entropy": 4.767383193969726,
      "epoch": 3.5793044953350295,
      "grad_norm": 1.625,
      "learning_rate": 0.0003838161716228129,
      "loss": 4.525,
      "mean_token_accuracy": 0.24922583252191544,
      "num_tokens": 10091507.0,
      "step": 12660
    },
    {
      "entropy": 4.812088680267334,
      "epoch": 3.5807181227028555,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003837261290113425,
      "loss": 4.5541,
      "mean_token_accuracy": 0.24006054550409317,
      "num_tokens": 10095269.0,
      "step": 12665
    },
    {
      "entropy": 4.653850126266479,
      "epoch": 3.5821317500706815,
      "grad_norm": 1.5625,
      "learning_rate": 0.00038363606367678343,
      "loss": 4.4576,
      "mean_token_accuracy": 0.260239177942276,
      "num_tokens": 10099098.0,
      "step": 12670
    },
    {
      "entropy": 4.697702360153198,
      "epoch": 3.583545377438507,
      "grad_norm": 1.640625,
      "learning_rate": 0.00038354597563795893,
      "loss": 4.607,
      "mean_token_accuracy": 0.24732397496700287,
      "num_tokens": 10103040.0,
      "step": 12675
    },
    {
      "entropy": 4.647365856170654,
      "epoch": 3.584959004806333,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003834558649136967,
      "loss": 4.5021,
      "mean_token_accuracy": 0.26386127918958663,
      "num_tokens": 10106862.0,
      "step": 12680
    },
    {
      "entropy": 4.683796072006226,
      "epoch": 3.586372632174159,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0003833657315228294,
      "loss": 4.3811,
      "mean_token_accuracy": 0.2633207604289055,
      "num_tokens": 10110980.0,
      "step": 12685
    },
    {
      "entropy": 4.777696990966797,
      "epoch": 3.5877862595419847,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003832755754841943,
      "loss": 4.6571,
      "mean_token_accuracy": 0.23991871476173401,
      "num_tokens": 10114913.0,
      "step": 12690
    },
    {
      "entropy": 4.771435642242432,
      "epoch": 3.5891998869098107,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003831853968166335,
      "loss": 4.614,
      "mean_token_accuracy": 0.24734264314174653,
      "num_tokens": 10119046.0,
      "step": 12695
    },
    {
      "entropy": 4.754781007766724,
      "epoch": 3.5906135142776363,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00038309519553899376,
      "loss": 4.636,
      "mean_token_accuracy": 0.24306143969297409,
      "num_tokens": 10123055.0,
      "step": 12700
    },
    {
      "entropy": 4.801106119155884,
      "epoch": 3.5920271416454623,
      "grad_norm": 1.65625,
      "learning_rate": 0.00038300497167012657,
      "loss": 4.621,
      "mean_token_accuracy": 0.24374054968357087,
      "num_tokens": 10127006.0,
      "step": 12705
    },
    {
      "entropy": 4.751120853424072,
      "epoch": 3.593440769013288,
      "grad_norm": 1.78125,
      "learning_rate": 0.00038291472522888816,
      "loss": 4.5889,
      "mean_token_accuracy": 0.2455888032913208,
      "num_tokens": 10130601.0,
      "step": 12710
    },
    {
      "entropy": 4.703146743774414,
      "epoch": 3.594854396381114,
      "grad_norm": 1.65625,
      "learning_rate": 0.00038282445623413953,
      "loss": 4.5895,
      "mean_token_accuracy": 0.26082912385463713,
      "num_tokens": 10134657.0,
      "step": 12715
    },
    {
      "entropy": 4.740464448928833,
      "epoch": 3.59626802374894,
      "grad_norm": 1.9375,
      "learning_rate": 0.0003827341647047463,
      "loss": 4.5382,
      "mean_token_accuracy": 0.25531924813985823,
      "num_tokens": 10138147.0,
      "step": 12720
    },
    {
      "entropy": 4.855399179458618,
      "epoch": 3.5976816511167655,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00038264385065957897,
      "loss": 4.7372,
      "mean_token_accuracy": 0.23535808324813842,
      "num_tokens": 10142101.0,
      "step": 12725
    },
    {
      "entropy": 4.747475004196167,
      "epoch": 3.5990952784845915,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00038255351411751234,
      "loss": 4.4318,
      "mean_token_accuracy": 0.2637227430939674,
      "num_tokens": 10146204.0,
      "step": 12730
    },
    {
      "entropy": 4.72800645828247,
      "epoch": 3.6005089058524176,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003824631550974264,
      "loss": 4.5491,
      "mean_token_accuracy": 0.2522769570350647,
      "num_tokens": 10149996.0,
      "step": 12735
    },
    {
      "entropy": 4.790418577194214,
      "epoch": 3.601922533220243,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003823727736182056,
      "loss": 4.6289,
      "mean_token_accuracy": 0.25051296055316924,
      "num_tokens": 10154021.0,
      "step": 12740
    },
    {
      "entropy": 4.709406995773316,
      "epoch": 3.6033361605880687,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003822823696987391,
      "loss": 4.5237,
      "mean_token_accuracy": 0.24612504988908768,
      "num_tokens": 10158070.0,
      "step": 12745
    },
    {
      "entropy": 4.711632537841797,
      "epoch": 3.6047497879558947,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003821919433579207,
      "loss": 4.518,
      "mean_token_accuracy": 0.2642649859189987,
      "num_tokens": 10162122.0,
      "step": 12750
    },
    {
      "entropy": 4.809568738937378,
      "epoch": 3.6061634153237208,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000382101494614649,
      "loss": 4.6023,
      "mean_token_accuracy": 0.25133455246686937,
      "num_tokens": 10165892.0,
      "step": 12755
    },
    {
      "entropy": 4.647157859802246,
      "epoch": 3.6075770426915463,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003820110234878272,
      "loss": 4.3749,
      "mean_token_accuracy": 0.27708097994327546,
      "num_tokens": 10169693.0,
      "step": 12760
    },
    {
      "entropy": 4.598480606079102,
      "epoch": 3.6089906700593724,
      "grad_norm": 1.6875,
      "learning_rate": 0.00038192052999636337,
      "loss": 4.435,
      "mean_token_accuracy": 0.26503183990716933,
      "num_tokens": 10173612.0,
      "step": 12765
    },
    {
      "entropy": 4.650153255462646,
      "epoch": 3.6104042974271984,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003818300141591698,
      "loss": 4.4324,
      "mean_token_accuracy": 0.26044865548610685,
      "num_tokens": 10177824.0,
      "step": 12770
    },
    {
      "entropy": 4.755365514755249,
      "epoch": 3.611817924795024,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003817394759951639,
      "loss": 4.5523,
      "mean_token_accuracy": 0.25777827501296996,
      "num_tokens": 10181712.0,
      "step": 12775
    },
    {
      "entropy": 4.747201061248779,
      "epoch": 3.61323155216285,
      "grad_norm": 1.640625,
      "learning_rate": 0.00038164891552326766,
      "loss": 4.5238,
      "mean_token_accuracy": 0.2583910435438156,
      "num_tokens": 10185942.0,
      "step": 12780
    },
    {
      "entropy": 4.7715253829956055,
      "epoch": 3.614645179530676,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003815583327624075,
      "loss": 4.5959,
      "mean_token_accuracy": 0.251844023168087,
      "num_tokens": 10190148.0,
      "step": 12785
    },
    {
      "entropy": 4.657932281494141,
      "epoch": 3.6160588068985016,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003814677277315147,
      "loss": 4.4966,
      "mean_token_accuracy": 0.26179673075675963,
      "num_tokens": 10194283.0,
      "step": 12790
    },
    {
      "entropy": 4.855694675445557,
      "epoch": 3.617472434266327,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003813771004495253,
      "loss": 4.6635,
      "mean_token_accuracy": 0.23786077350378038,
      "num_tokens": 10198202.0,
      "step": 12795
    },
    {
      "entropy": 4.704127264022827,
      "epoch": 3.618886061634153,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003812864509353797,
      "loss": 4.4907,
      "mean_token_accuracy": 0.2600321665406227,
      "num_tokens": 10201995.0,
      "step": 12800
    },
    {
      "entropy": 4.669480466842652,
      "epoch": 3.620299689001979,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00038119577920802295,
      "loss": 4.5448,
      "mean_token_accuracy": 0.2574765011668205,
      "num_tokens": 10205938.0,
      "step": 12805
    },
    {
      "entropy": 4.620076990127563,
      "epoch": 3.6217133163698048,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003811050852864051,
      "loss": 4.5102,
      "mean_token_accuracy": 0.2648425206542015,
      "num_tokens": 10210226.0,
      "step": 12810
    },
    {
      "entropy": 4.7150860786437985,
      "epoch": 3.623126943737631,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00038101436918948074,
      "loss": 4.5266,
      "mean_token_accuracy": 0.2521339640021324,
      "num_tokens": 10213986.0,
      "step": 12815
    },
    {
      "entropy": 4.7100037097930905,
      "epoch": 3.624540571105457,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003809236309362087,
      "loss": 4.5233,
      "mean_token_accuracy": 0.26093397885560987,
      "num_tokens": 10218053.0,
      "step": 12820
    },
    {
      "entropy": 4.762806749343872,
      "epoch": 3.6259541984732824,
      "grad_norm": 1.671875,
      "learning_rate": 0.00038083287054555264,
      "loss": 4.5917,
      "mean_token_accuracy": 0.25313022881746294,
      "num_tokens": 10221997.0,
      "step": 12825
    },
    {
      "entropy": 4.7969776630401615,
      "epoch": 3.6273678258411084,
      "grad_norm": 1.875,
      "learning_rate": 0.0003807420880364813,
      "loss": 4.6243,
      "mean_token_accuracy": 0.2492845192551613,
      "num_tokens": 10226270.0,
      "step": 12830
    },
    {
      "entropy": 4.8270673751831055,
      "epoch": 3.628781453208934,
      "grad_norm": 1.796875,
      "learning_rate": 0.00038065128342796726,
      "loss": 4.616,
      "mean_token_accuracy": 0.24698579758405687,
      "num_tokens": 10230243.0,
      "step": 12835
    },
    {
      "entropy": 4.716474533081055,
      "epoch": 3.63019508057676,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003805604567389885,
      "loss": 4.5253,
      "mean_token_accuracy": 0.258763912320137,
      "num_tokens": 10234388.0,
      "step": 12840
    },
    {
      "entropy": 4.68907151222229,
      "epoch": 3.6316087079445856,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00038046960798852696,
      "loss": 4.5339,
      "mean_token_accuracy": 0.2624957889318466,
      "num_tokens": 10238411.0,
      "step": 12845
    },
    {
      "entropy": 4.679296827316284,
      "epoch": 3.6330223353124116,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00038037873719556957,
      "loss": 4.5061,
      "mean_token_accuracy": 0.25261744260787966,
      "num_tokens": 10242272.0,
      "step": 12850
    },
    {
      "entropy": 4.70474681854248,
      "epoch": 3.6344359626802376,
      "grad_norm": 1.59375,
      "learning_rate": 0.00038028784437910774,
      "loss": 4.5494,
      "mean_token_accuracy": 0.2514804646372795,
      "num_tokens": 10246165.0,
      "step": 12855
    },
    {
      "entropy": 4.665062999725341,
      "epoch": 3.635849590048063,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003801969295581375,
      "loss": 4.5411,
      "mean_token_accuracy": 0.2576843500137329,
      "num_tokens": 10249909.0,
      "step": 12860
    },
    {
      "entropy": 4.736473941802979,
      "epoch": 3.637263217415889,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003801059927516595,
      "loss": 4.5565,
      "mean_token_accuracy": 0.253127321600914,
      "num_tokens": 10254296.0,
      "step": 12865
    },
    {
      "entropy": 4.739093494415283,
      "epoch": 3.6386768447837152,
      "grad_norm": 1.65625,
      "learning_rate": 0.00038001503397867896,
      "loss": 4.5541,
      "mean_token_accuracy": 0.2533505782485008,
      "num_tokens": 10258394.0,
      "step": 12870
    },
    {
      "entropy": 4.672406768798828,
      "epoch": 3.640090472151541,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003799240532582058,
      "loss": 4.4479,
      "mean_token_accuracy": 0.26773455590009687,
      "num_tokens": 10262460.0,
      "step": 12875
    },
    {
      "entropy": 4.75471043586731,
      "epoch": 3.641504099519367,
      "grad_norm": 1.53125,
      "learning_rate": 0.00037983305060925425,
      "loss": 4.6538,
      "mean_token_accuracy": 0.24986883550882338,
      "num_tokens": 10266707.0,
      "step": 12880
    },
    {
      "entropy": 4.755679178237915,
      "epoch": 3.6429177268871924,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003797420260508434,
      "loss": 4.5971,
      "mean_token_accuracy": 0.2531541034579277,
      "num_tokens": 10271110.0,
      "step": 12885
    },
    {
      "entropy": 4.702559614181519,
      "epoch": 3.6443313542550184,
      "grad_norm": 1.734375,
      "learning_rate": 0.00037965097960199684,
      "loss": 4.4982,
      "mean_token_accuracy": 0.2526899576187134,
      "num_tokens": 10275255.0,
      "step": 12890
    },
    {
      "entropy": 4.718622875213623,
      "epoch": 3.645744981622844,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003795599112817427,
      "loss": 4.6227,
      "mean_token_accuracy": 0.25288901627063753,
      "num_tokens": 10279372.0,
      "step": 12895
    },
    {
      "entropy": 4.6861693382263185,
      "epoch": 3.64715860899067,
      "grad_norm": 1.578125,
      "learning_rate": 0.00037946882110911364,
      "loss": 4.4967,
      "mean_token_accuracy": 0.24874725043773652,
      "num_tokens": 10283277.0,
      "step": 12900
    },
    {
      "entropy": 4.80086669921875,
      "epoch": 3.648572236358496,
      "grad_norm": 1.625,
      "learning_rate": 0.000379377709103147,
      "loss": 4.7006,
      "mean_token_accuracy": 0.23966424465179442,
      "num_tokens": 10287323.0,
      "step": 12905
    },
    {
      "entropy": 4.663335609436035,
      "epoch": 3.6499858637263216,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00037928657528288464,
      "loss": 4.4301,
      "mean_token_accuracy": 0.2685204952955246,
      "num_tokens": 10291191.0,
      "step": 12910
    },
    {
      "entropy": 4.694800806045532,
      "epoch": 3.6513994910941476,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003791954196673729,
      "loss": 4.5064,
      "mean_token_accuracy": 0.259264187514782,
      "num_tokens": 10295111.0,
      "step": 12915
    },
    {
      "entropy": 4.759398365020752,
      "epoch": 3.6528131184619737,
      "grad_norm": 1.59375,
      "learning_rate": 0.00037910424227566276,
      "loss": 4.5994,
      "mean_token_accuracy": 0.2428250029683113,
      "num_tokens": 10299418.0,
      "step": 12920
    },
    {
      "entropy": 4.730670213699341,
      "epoch": 3.6542267458297992,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003790130431268097,
      "loss": 4.5381,
      "mean_token_accuracy": 0.26092619597911837,
      "num_tokens": 10303556.0,
      "step": 12925
    },
    {
      "entropy": 4.726977109909058,
      "epoch": 3.6556403731976252,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00037892182223987395,
      "loss": 4.587,
      "mean_token_accuracy": 0.26530673652887343,
      "num_tokens": 10307373.0,
      "step": 12930
    },
    {
      "entropy": 4.692935943603516,
      "epoch": 3.657054000565451,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037883057963391994,
      "loss": 4.4731,
      "mean_token_accuracy": 0.26542289555072784,
      "num_tokens": 10311604.0,
      "step": 12935
    },
    {
      "entropy": 4.725019454956055,
      "epoch": 3.658467627933277,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00037873931532801685,
      "loss": 4.5583,
      "mean_token_accuracy": 0.2567869171500206,
      "num_tokens": 10315628.0,
      "step": 12940
    },
    {
      "entropy": 4.767247343063355,
      "epoch": 3.6598812553011024,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003786480293412383,
      "loss": 4.6174,
      "mean_token_accuracy": 0.24424019008874892,
      "num_tokens": 10319970.0,
      "step": 12945
    },
    {
      "entropy": 4.696600914001465,
      "epoch": 3.6612948826689284,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003785567216926626,
      "loss": 4.4375,
      "mean_token_accuracy": 0.25566643476486206,
      "num_tokens": 10323696.0,
      "step": 12950
    },
    {
      "entropy": 4.793924140930176,
      "epoch": 3.6627085100367545,
      "grad_norm": 1.5,
      "learning_rate": 0.00037846539240137244,
      "loss": 4.5899,
      "mean_token_accuracy": 0.24605975300073624,
      "num_tokens": 10327502.0,
      "step": 12955
    },
    {
      "entropy": 4.691178703308106,
      "epoch": 3.66412213740458,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003783740414864551,
      "loss": 4.5334,
      "mean_token_accuracy": 0.25874372869729995,
      "num_tokens": 10331520.0,
      "step": 12960
    },
    {
      "entropy": 4.709694242477417,
      "epoch": 3.665535764772406,
      "grad_norm": 1.65625,
      "learning_rate": 0.00037828266896700237,
      "loss": 4.5172,
      "mean_token_accuracy": 0.25487100034952165,
      "num_tokens": 10335504.0,
      "step": 12965
    },
    {
      "entropy": 4.6805657863616945,
      "epoch": 3.666949392140232,
      "grad_norm": 1.546875,
      "learning_rate": 0.00037819127486211045,
      "loss": 4.4365,
      "mean_token_accuracy": 0.26190069764852525,
      "num_tokens": 10339955.0,
      "step": 12970
    },
    {
      "entropy": 4.7753034114837645,
      "epoch": 3.6683630195080577,
      "grad_norm": 1.5625,
      "learning_rate": 0.00037809985919088013,
      "loss": 4.5511,
      "mean_token_accuracy": 0.25094142705202105,
      "num_tokens": 10344047.0,
      "step": 12975
    },
    {
      "entropy": 4.787731313705445,
      "epoch": 3.6697766468758837,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003780084219724169,
      "loss": 4.6273,
      "mean_token_accuracy": 0.25052869617938994,
      "num_tokens": 10348020.0,
      "step": 12980
    },
    {
      "entropy": 4.697200727462769,
      "epoch": 3.6711902742437093,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00037791696322583043,
      "loss": 4.4237,
      "mean_token_accuracy": 0.2640311747789383,
      "num_tokens": 10351910.0,
      "step": 12985
    },
    {
      "entropy": 4.7134308338165285,
      "epoch": 3.6726039016115353,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003778254829702351,
      "loss": 4.527,
      "mean_token_accuracy": 0.2707617789506912,
      "num_tokens": 10355937.0,
      "step": 12990
    },
    {
      "entropy": 4.708440637588501,
      "epoch": 3.674017528979361,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003777339812247496,
      "loss": 4.5342,
      "mean_token_accuracy": 0.25271027386188505,
      "num_tokens": 10359744.0,
      "step": 12995
    },
    {
      "entropy": 4.680083703994751,
      "epoch": 3.675431156347187,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037764245800849737,
      "loss": 4.4979,
      "mean_token_accuracy": 0.2586936458945274,
      "num_tokens": 10363628.0,
      "step": 13000
    },
    {
      "epoch": 3.675431156347187,
      "eval_entropy": 4.590178150897496,
      "eval_loss": 5.05978536605835,
      "eval_mean_token_accuracy": 0.24213871367217632,
      "eval_num_tokens": 10363628.0,
      "eval_runtime": 6.6464,
      "eval_samples_per_second": 1697.767,
      "eval_steps_per_second": 212.296,
      "step": 13000
    },
    {
      "entropy": 4.8483256816864015,
      "epoch": 3.676844783715013,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003775509133406061,
      "loss": 4.6175,
      "mean_token_accuracy": 0.24079225957393646,
      "num_tokens": 10367258.0,
      "step": 13005
    },
    {
      "entropy": 4.853681325912476,
      "epoch": 3.6782584110828385,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003774593472402081,
      "loss": 4.6084,
      "mean_token_accuracy": 0.25941989570856094,
      "num_tokens": 10371383.0,
      "step": 13010
    },
    {
      "entropy": 4.828049659729004,
      "epoch": 3.6796720384506645,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003773677597264402,
      "loss": 4.5989,
      "mean_token_accuracy": 0.24006949812173844,
      "num_tokens": 10375198.0,
      "step": 13015
    },
    {
      "entropy": 4.8391022205352785,
      "epoch": 3.6810856658184905,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003772761508184435,
      "loss": 4.6088,
      "mean_token_accuracy": 0.2383268177509308,
      "num_tokens": 10379265.0,
      "step": 13020
    },
    {
      "entropy": 4.757513809204101,
      "epoch": 3.682499293186316,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00037718452053536375,
      "loss": 4.4938,
      "mean_token_accuracy": 0.25974127650260925,
      "num_tokens": 10383449.0,
      "step": 13025
    },
    {
      "entropy": 4.658124494552612,
      "epoch": 3.6839129205541417,
      "grad_norm": 1.53125,
      "learning_rate": 0.000377092868896351,
      "loss": 4.4416,
      "mean_token_accuracy": 0.2698409095406532,
      "num_tokens": 10387282.0,
      "step": 13030
    },
    {
      "entropy": 4.685115003585816,
      "epoch": 3.6853265479219677,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00037700119592056,
      "loss": 4.5213,
      "mean_token_accuracy": 0.2569262072443962,
      "num_tokens": 10391351.0,
      "step": 13035
    },
    {
      "entropy": 4.647048234939575,
      "epoch": 3.6867401752897937,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037690950162714995,
      "loss": 4.3983,
      "mean_token_accuracy": 0.26536639779806137,
      "num_tokens": 10395225.0,
      "step": 13040
    },
    {
      "entropy": 4.725645542144775,
      "epoch": 3.6881538026576193,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003768177860352841,
      "loss": 4.5809,
      "mean_token_accuracy": 0.2538579314947128,
      "num_tokens": 10399340.0,
      "step": 13045
    },
    {
      "entropy": 4.725207757949829,
      "epoch": 3.6895674300254453,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037672604916413064,
      "loss": 4.5886,
      "mean_token_accuracy": 0.25287462025880814,
      "num_tokens": 10403608.0,
      "step": 13050
    },
    {
      "entropy": 4.6872235298156735,
      "epoch": 3.6909810573932713,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00037663429103286197,
      "loss": 4.4653,
      "mean_token_accuracy": 0.2559305801987648,
      "num_tokens": 10407812.0,
      "step": 13055
    },
    {
      "entropy": 4.7078453540802006,
      "epoch": 3.692394684761097,
      "grad_norm": 1.59375,
      "learning_rate": 0.000376542511660655,
      "loss": 4.5886,
      "mean_token_accuracy": 0.25200741589069364,
      "num_tokens": 10411836.0,
      "step": 13060
    },
    {
      "entropy": 4.679752588272095,
      "epoch": 3.693808312128923,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037645071106669095,
      "loss": 4.5869,
      "mean_token_accuracy": 0.24999840408563614,
      "num_tokens": 10416230.0,
      "step": 13065
    },
    {
      "entropy": 4.786351728439331,
      "epoch": 3.695221939496749,
      "grad_norm": 1.75,
      "learning_rate": 0.00037635888927015565,
      "loss": 4.5971,
      "mean_token_accuracy": 0.24785867482423782,
      "num_tokens": 10420396.0,
      "step": 13070
    },
    {
      "entropy": 4.766790437698364,
      "epoch": 3.6966355668645745,
      "grad_norm": 1.546875,
      "learning_rate": 0.00037626704629023923,
      "loss": 4.5394,
      "mean_token_accuracy": 0.25345624834299085,
      "num_tokens": 10424371.0,
      "step": 13075
    },
    {
      "entropy": 4.645627212524414,
      "epoch": 3.6980491942324,
      "grad_norm": 1.4375,
      "learning_rate": 0.00037617518214613647,
      "loss": 4.3849,
      "mean_token_accuracy": 0.2700305849313736,
      "num_tokens": 10428144.0,
      "step": 13080
    },
    {
      "entropy": 4.661588382720947,
      "epoch": 3.699462821600226,
      "grad_norm": 1.625,
      "learning_rate": 0.0003760832968570462,
      "loss": 4.5025,
      "mean_token_accuracy": 0.24727798253297806,
      "num_tokens": 10432251.0,
      "step": 13085
    },
    {
      "entropy": 4.683081007003784,
      "epoch": 3.700876448968052,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00037599139044217203,
      "loss": 4.4972,
      "mean_token_accuracy": 0.2564370259642601,
      "num_tokens": 10436244.0,
      "step": 13090
    },
    {
      "entropy": 4.776733875274658,
      "epoch": 3.7022900763358777,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003758994629207217,
      "loss": 4.6171,
      "mean_token_accuracy": 0.24849742501974106,
      "num_tokens": 10440465.0,
      "step": 13095
    },
    {
      "entropy": 4.874036121368408,
      "epoch": 3.7037037037037037,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00037580751431190765,
      "loss": 4.6328,
      "mean_token_accuracy": 0.25063999742269516,
      "num_tokens": 10444374.0,
      "step": 13100
    },
    {
      "entropy": 4.75330080986023,
      "epoch": 3.7051173310715297,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003757155446349464,
      "loss": 4.5749,
      "mean_token_accuracy": 0.2545487105846405,
      "num_tokens": 10448498.0,
      "step": 13105
    },
    {
      "entropy": 4.673137092590332,
      "epoch": 3.7065309584393553,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003756235539090591,
      "loss": 4.5493,
      "mean_token_accuracy": 0.2593318268656731,
      "num_tokens": 10452328.0,
      "step": 13110
    },
    {
      "entropy": 4.644604587554932,
      "epoch": 3.7079445858071813,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00037553154215347133,
      "loss": 4.5024,
      "mean_token_accuracy": 0.25765192210674287,
      "num_tokens": 10456037.0,
      "step": 13115
    },
    {
      "entropy": 4.66724534034729,
      "epoch": 3.709358213175007,
      "grad_norm": 1.53125,
      "learning_rate": 0.00037543950938741293,
      "loss": 4.4664,
      "mean_token_accuracy": 0.2603970214724541,
      "num_tokens": 10459951.0,
      "step": 13120
    },
    {
      "entropy": 4.804188394546509,
      "epoch": 3.710771840542833,
      "grad_norm": 1.578125,
      "learning_rate": 0.00037534745563011816,
      "loss": 4.663,
      "mean_token_accuracy": 0.23460719138383865,
      "num_tokens": 10464023.0,
      "step": 13125
    },
    {
      "entropy": 4.73428430557251,
      "epoch": 3.7121854679106585,
      "grad_norm": 1.546875,
      "learning_rate": 0.00037525538090082565,
      "loss": 4.4754,
      "mean_token_accuracy": 0.2686974138021469,
      "num_tokens": 10468161.0,
      "step": 13130
    },
    {
      "entropy": 4.724827909469605,
      "epoch": 3.7135990952784845,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0003751632852187785,
      "loss": 4.5312,
      "mean_token_accuracy": 0.2619817674160004,
      "num_tokens": 10472141.0,
      "step": 13135
    },
    {
      "entropy": 4.6808576583862305,
      "epoch": 3.7150127226463106,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003750711686032242,
      "loss": 4.5656,
      "mean_token_accuracy": 0.2607214316725731,
      "num_tokens": 10476166.0,
      "step": 13140
    },
    {
      "entropy": 4.762052154541015,
      "epoch": 3.716426350014136,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00037497903107341446,
      "loss": 4.6297,
      "mean_token_accuracy": 0.2458733931183815,
      "num_tokens": 10480320.0,
      "step": 13145
    },
    {
      "entropy": 4.70688829421997,
      "epoch": 3.717839977381962,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00037488687264860545,
      "loss": 4.5241,
      "mean_token_accuracy": 0.25273083746433256,
      "num_tokens": 10484398.0,
      "step": 13150
    },
    {
      "entropy": 4.689479494094849,
      "epoch": 3.719253604749788,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00037479469334805775,
      "loss": 4.5381,
      "mean_token_accuracy": 0.26631945222616193,
      "num_tokens": 10488240.0,
      "step": 13155
    },
    {
      "entropy": 4.7215570449829105,
      "epoch": 3.7206672321176137,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003747024931910362,
      "loss": 4.6054,
      "mean_token_accuracy": 0.25294113755226133,
      "num_tokens": 10492456.0,
      "step": 13160
    },
    {
      "entropy": 4.701136684417724,
      "epoch": 3.7220808594854398,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003746102721968102,
      "loss": 4.4833,
      "mean_token_accuracy": 0.26558475345373156,
      "num_tokens": 10496442.0,
      "step": 13165
    },
    {
      "entropy": 4.702780294418335,
      "epoch": 3.7234944868532653,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003745180303846532,
      "loss": 4.4734,
      "mean_token_accuracy": 0.2506163537502289,
      "num_tokens": 10500399.0,
      "step": 13170
    },
    {
      "entropy": 4.651794958114624,
      "epoch": 3.7249081142210914,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003744257677738433,
      "loss": 4.5215,
      "mean_token_accuracy": 0.25083540230989454,
      "num_tokens": 10504267.0,
      "step": 13175
    },
    {
      "entropy": 4.743178176879883,
      "epoch": 3.726321741588917,
      "grad_norm": 1.59375,
      "learning_rate": 0.00037433348438366276,
      "loss": 4.4971,
      "mean_token_accuracy": 0.2500301286578178,
      "num_tokens": 10508100.0,
      "step": 13180
    },
    {
      "entropy": 4.685713768005371,
      "epoch": 3.727735368956743,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003742411802333983,
      "loss": 4.474,
      "mean_token_accuracy": 0.2520857661962509,
      "num_tokens": 10512337.0,
      "step": 13185
    },
    {
      "entropy": 4.634169101715088,
      "epoch": 3.729148996324569,
      "grad_norm": 1.625,
      "learning_rate": 0.0003741488553423408,
      "loss": 4.5165,
      "mean_token_accuracy": 0.2605761781334877,
      "num_tokens": 10516478.0,
      "step": 13190
    },
    {
      "entropy": 4.719336080551147,
      "epoch": 3.7305626236923946,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00037405650972978567,
      "loss": 4.6447,
      "mean_token_accuracy": 0.24417183101177214,
      "num_tokens": 10520569.0,
      "step": 13195
    },
    {
      "entropy": 4.686707878112793,
      "epoch": 3.7319762510602206,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003739641434150324,
      "loss": 4.4848,
      "mean_token_accuracy": 0.2563915506005287,
      "num_tokens": 10524778.0,
      "step": 13200
    },
    {
      "entropy": 4.665077352523804,
      "epoch": 3.7333898784280466,
      "grad_norm": 1.65625,
      "learning_rate": 0.00037387175641738534,
      "loss": 4.5478,
      "mean_token_accuracy": 0.25484164953231814,
      "num_tokens": 10528949.0,
      "step": 13205
    },
    {
      "entropy": 4.699547576904297,
      "epoch": 3.734803505795872,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00037377934875615257,
      "loss": 4.4722,
      "mean_token_accuracy": 0.2696780875325203,
      "num_tokens": 10532950.0,
      "step": 13210
    },
    {
      "entropy": 4.746389579772949,
      "epoch": 3.736217133163698,
      "grad_norm": 1.546875,
      "learning_rate": 0.00037368692045064666,
      "loss": 4.5257,
      "mean_token_accuracy": 0.25479610711336137,
      "num_tokens": 10537140.0,
      "step": 13215
    },
    {
      "entropy": 4.713578605651856,
      "epoch": 3.7376307605315238,
      "grad_norm": 1.5625,
      "learning_rate": 0.00037359447152018465,
      "loss": 4.4874,
      "mean_token_accuracy": 0.25432832837104796,
      "num_tokens": 10541195.0,
      "step": 13220
    },
    {
      "entropy": 4.688205814361572,
      "epoch": 3.73904438789935,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00037350200198408786,
      "loss": 4.4939,
      "mean_token_accuracy": 0.2645654171705246,
      "num_tokens": 10545289.0,
      "step": 13225
    },
    {
      "entropy": 4.668663740158081,
      "epoch": 3.7404580152671754,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003734095118616816,
      "loss": 4.5704,
      "mean_token_accuracy": 0.2509492725133896,
      "num_tokens": 10549324.0,
      "step": 13230
    },
    {
      "entropy": 4.6547698974609375,
      "epoch": 3.7418716426350014,
      "grad_norm": 1.625,
      "learning_rate": 0.0003733170011722961,
      "loss": 4.4502,
      "mean_token_accuracy": 0.26323665827512743,
      "num_tokens": 10553166.0,
      "step": 13235
    },
    {
      "entropy": 4.836304950714111,
      "epoch": 3.7432852700028274,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003732244699352653,
      "loss": 4.6501,
      "mean_token_accuracy": 0.24058225899934768,
      "num_tokens": 10556834.0,
      "step": 13240
    },
    {
      "entropy": 4.7401031970977785,
      "epoch": 3.744698897370653,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037313191816992764,
      "loss": 4.5295,
      "mean_token_accuracy": 0.2501175433397293,
      "num_tokens": 10561079.0,
      "step": 13245
    },
    {
      "entropy": 4.621883773803711,
      "epoch": 3.746112524738479,
      "grad_norm": 1.578125,
      "learning_rate": 0.00037303934589562587,
      "loss": 4.4014,
      "mean_token_accuracy": 0.267057266831398,
      "num_tokens": 10564936.0,
      "step": 13250
    },
    {
      "entropy": 4.599834537506103,
      "epoch": 3.747526152106305,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003729467531317071,
      "loss": 4.5299,
      "mean_token_accuracy": 0.25159996896982195,
      "num_tokens": 10568990.0,
      "step": 13255
    },
    {
      "entropy": 4.6471700191497805,
      "epoch": 3.7489397794741306,
      "grad_norm": 1.515625,
      "learning_rate": 0.00037285413989752265,
      "loss": 4.5129,
      "mean_token_accuracy": 0.25016224980354307,
      "num_tokens": 10573046.0,
      "step": 13260
    },
    {
      "entropy": 4.895354509353638,
      "epoch": 3.7503534068419566,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00037276150621242803,
      "loss": 4.6411,
      "mean_token_accuracy": 0.23161313831806182,
      "num_tokens": 10577131.0,
      "step": 13265
    },
    {
      "entropy": 4.771190404891968,
      "epoch": 3.751767034209782,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003726688520957832,
      "loss": 4.5408,
      "mean_token_accuracy": 0.2477626159787178,
      "num_tokens": 10581153.0,
      "step": 13270
    },
    {
      "entropy": 4.708290338516235,
      "epoch": 3.753180661577608,
      "grad_norm": 1.5625,
      "learning_rate": 0.00037257617756695225,
      "loss": 4.5292,
      "mean_token_accuracy": 0.24075633883476258,
      "num_tokens": 10585080.0,
      "step": 13275
    },
    {
      "entropy": 4.665339756011963,
      "epoch": 3.754594288945434,
      "grad_norm": 1.578125,
      "learning_rate": 0.00037248348264530345,
      "loss": 4.4839,
      "mean_token_accuracy": 0.25866914391517637,
      "num_tokens": 10589087.0,
      "step": 13280
    },
    {
      "entropy": 4.811703824996949,
      "epoch": 3.75600791631326,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003723907673502097,
      "loss": 4.6878,
      "mean_token_accuracy": 0.23482495099306105,
      "num_tokens": 10593077.0,
      "step": 13285
    },
    {
      "entropy": 4.722591066360474,
      "epoch": 3.757421543681086,
      "grad_norm": 1.671875,
      "learning_rate": 0.00037229803170104784,
      "loss": 4.5219,
      "mean_token_accuracy": 0.2532244622707367,
      "num_tokens": 10596942.0,
      "step": 13290
    },
    {
      "entropy": 4.678056001663208,
      "epoch": 3.7588351710489114,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000372205275717199,
      "loss": 4.536,
      "mean_token_accuracy": 0.2516951158642769,
      "num_tokens": 10601020.0,
      "step": 13295
    },
    {
      "entropy": 4.711244010925293,
      "epoch": 3.7602487984167374,
      "grad_norm": 1.609375,
      "learning_rate": 0.00037211249941804865,
      "loss": 4.5984,
      "mean_token_accuracy": 0.2484810933470726,
      "num_tokens": 10604959.0,
      "step": 13300
    },
    {
      "entropy": 4.73989782333374,
      "epoch": 3.7616624257845634,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003720197028229864,
      "loss": 4.4835,
      "mean_token_accuracy": 0.2539603218436241,
      "num_tokens": 10609109.0,
      "step": 13305
    },
    {
      "entropy": 4.737037992477417,
      "epoch": 3.763076053152389,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003719268859514062,
      "loss": 4.586,
      "mean_token_accuracy": 0.2585724011063576,
      "num_tokens": 10612979.0,
      "step": 13310
    },
    {
      "entropy": 4.738237190246582,
      "epoch": 3.7644896805202146,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00037183404882270627,
      "loss": 4.6284,
      "mean_token_accuracy": 0.2529520243406296,
      "num_tokens": 10617277.0,
      "step": 13315
    },
    {
      "entropy": 4.675708198547364,
      "epoch": 3.7659033078880406,
      "grad_norm": 1.640625,
      "learning_rate": 0.00037174119145628895,
      "loss": 4.5292,
      "mean_token_accuracy": 0.2576525956392288,
      "num_tokens": 10621446.0,
      "step": 13320
    },
    {
      "entropy": 4.71737060546875,
      "epoch": 3.7673169352558666,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003716483138715608,
      "loss": 4.5292,
      "mean_token_accuracy": 0.2455099940299988,
      "num_tokens": 10625273.0,
      "step": 13325
    },
    {
      "entropy": 4.647572565078735,
      "epoch": 3.768730562623692,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003715554160879327,
      "loss": 4.4963,
      "mean_token_accuracy": 0.2531684398651123,
      "num_tokens": 10629092.0,
      "step": 13330
    },
    {
      "entropy": 4.745577096939087,
      "epoch": 3.7701441899915182,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003714624981248198,
      "loss": 4.5258,
      "mean_token_accuracy": 0.2613775163888931,
      "num_tokens": 10633044.0,
      "step": 13335
    },
    {
      "entropy": 4.715321826934814,
      "epoch": 3.7715578173593443,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00037136956000164117,
      "loss": 4.5663,
      "mean_token_accuracy": 0.251896370947361,
      "num_tokens": 10636993.0,
      "step": 13340
    },
    {
      "entropy": 4.668729543685913,
      "epoch": 3.77297144472717,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00037127660173782057,
      "loss": 4.4274,
      "mean_token_accuracy": 0.2719466745853424,
      "num_tokens": 10640927.0,
      "step": 13345
    },
    {
      "entropy": 4.7117105484008786,
      "epoch": 3.774385072094996,
      "grad_norm": 1.671875,
      "learning_rate": 0.00037118362335278543,
      "loss": 4.6077,
      "mean_token_accuracy": 0.25339147001504897,
      "num_tokens": 10644651.0,
      "step": 13350
    },
    {
      "entropy": 4.642689275741577,
      "epoch": 3.775798699462822,
      "grad_norm": 1.4453125,
      "learning_rate": 0.0003710906248659679,
      "loss": 4.4179,
      "mean_token_accuracy": 0.26095883548259735,
      "num_tokens": 10648731.0,
      "step": 13355
    },
    {
      "entropy": 4.7406679630279545,
      "epoch": 3.7772123268306474,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003709976062968039,
      "loss": 4.6408,
      "mean_token_accuracy": 0.2553518444299698,
      "num_tokens": 10652585.0,
      "step": 13360
    },
    {
      "entropy": 4.662831258773804,
      "epoch": 3.778625954198473,
      "grad_norm": 1.71875,
      "learning_rate": 0.00037090456766473393,
      "loss": 4.5196,
      "mean_token_accuracy": 0.251325735449791,
      "num_tokens": 10656533.0,
      "step": 13365
    },
    {
      "entropy": 4.7314221382141115,
      "epoch": 3.780039581566299,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00037081150898920226,
      "loss": 4.584,
      "mean_token_accuracy": 0.2476162239909172,
      "num_tokens": 10660551.0,
      "step": 13370
    },
    {
      "entropy": 4.744934511184693,
      "epoch": 3.781453208934125,
      "grad_norm": 1.5625,
      "learning_rate": 0.00037071843028965766,
      "loss": 4.5183,
      "mean_token_accuracy": 0.25901537835597993,
      "num_tokens": 10664489.0,
      "step": 13375
    },
    {
      "entropy": 4.720381402969361,
      "epoch": 3.7828668363019506,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00037062533158555315,
      "loss": 4.427,
      "mean_token_accuracy": 0.260734336078167,
      "num_tokens": 10668391.0,
      "step": 13380
    },
    {
      "entropy": 4.707159996032715,
      "epoch": 3.7842804636697767,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003705322128963456,
      "loss": 4.6071,
      "mean_token_accuracy": 0.2596296682953835,
      "num_tokens": 10672440.0,
      "step": 13385
    },
    {
      "entropy": 4.679089212417603,
      "epoch": 3.7856940910376027,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003704390742414964,
      "loss": 4.6119,
      "mean_token_accuracy": 0.2525580570101738,
      "num_tokens": 10676366.0,
      "step": 13390
    },
    {
      "entropy": 4.6014354705810545,
      "epoch": 3.7871077184054283,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003703459156404708,
      "loss": 4.5467,
      "mean_token_accuracy": 0.2572270080447197,
      "num_tokens": 10680204.0,
      "step": 13395
    },
    {
      "entropy": 4.6332518577575685,
      "epoch": 3.7885213457732543,
      "grad_norm": 1.484375,
      "learning_rate": 0.00037025273711273837,
      "loss": 4.4636,
      "mean_token_accuracy": 0.2717544063925743,
      "num_tokens": 10684359.0,
      "step": 13400
    },
    {
      "entropy": 4.8223954200744625,
      "epoch": 3.78993497314108,
      "grad_norm": 1.53125,
      "learning_rate": 0.00037015953867777297,
      "loss": 4.706,
      "mean_token_accuracy": 0.2449349895119667,
      "num_tokens": 10688551.0,
      "step": 13405
    },
    {
      "entropy": 4.714366865158081,
      "epoch": 3.791348600508906,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003700663203550524,
      "loss": 4.4678,
      "mean_token_accuracy": 0.2661253288388252,
      "num_tokens": 10692231.0,
      "step": 13410
    },
    {
      "entropy": 4.702810478210449,
      "epoch": 3.7927622278767315,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003699730821640587,
      "loss": 4.4689,
      "mean_token_accuracy": 0.2553928479552269,
      "num_tokens": 10696156.0,
      "step": 13415
    },
    {
      "entropy": 4.857287454605102,
      "epoch": 3.7941758552445575,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003698798241242783,
      "loss": 4.802,
      "mean_token_accuracy": 0.235882006585598,
      "num_tokens": 10700016.0,
      "step": 13420
    },
    {
      "entropy": 4.7657801628112795,
      "epoch": 3.7955894826123835,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00036978654625520127,
      "loss": 4.645,
      "mean_token_accuracy": 0.24572035372257234,
      "num_tokens": 10704089.0,
      "step": 13425
    },
    {
      "entropy": 4.781498861312866,
      "epoch": 3.797003109980209,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00036969324857632227,
      "loss": 4.5269,
      "mean_token_accuracy": 0.24792766124010085,
      "num_tokens": 10708008.0,
      "step": 13430
    },
    {
      "entropy": 4.70299277305603,
      "epoch": 3.798416737348035,
      "grad_norm": 1.546875,
      "learning_rate": 0.00036959993110713976,
      "loss": 4.4887,
      "mean_token_accuracy": 0.25832368731498717,
      "num_tokens": 10711874.0,
      "step": 13435
    },
    {
      "entropy": 4.6556250095367435,
      "epoch": 3.799830364715861,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00036950659386715675,
      "loss": 4.5034,
      "mean_token_accuracy": 0.25679989606142045,
      "num_tokens": 10716297.0,
      "step": 13440
    },
    {
      "entropy": 4.701553726196289,
      "epoch": 3.8012439920836867,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00036941323687587994,
      "loss": 4.5465,
      "mean_token_accuracy": 0.25136248022317886,
      "num_tokens": 10720605.0,
      "step": 13445
    },
    {
      "entropy": 4.710104751586914,
      "epoch": 3.8026576194515127,
      "grad_norm": 1.546875,
      "learning_rate": 0.00036931986015282056,
      "loss": 4.5692,
      "mean_token_accuracy": 0.26373191177845,
      "num_tokens": 10724498.0,
      "step": 13450
    },
    {
      "entropy": 4.734765338897705,
      "epoch": 3.8040712468193383,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00036922646371749355,
      "loss": 4.5535,
      "mean_token_accuracy": 0.2564811438322067,
      "num_tokens": 10728396.0,
      "step": 13455
    },
    {
      "entropy": 4.602238368988037,
      "epoch": 3.8054848741871643,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003691330475894184,
      "loss": 4.5048,
      "mean_token_accuracy": 0.2574146628379822,
      "num_tokens": 10732257.0,
      "step": 13460
    },
    {
      "entropy": 4.680519485473633,
      "epoch": 3.80689850155499,
      "grad_norm": 1.671875,
      "learning_rate": 0.00036903961178811836,
      "loss": 4.5012,
      "mean_token_accuracy": 0.2627511784434319,
      "num_tokens": 10736090.0,
      "step": 13465
    },
    {
      "entropy": 4.648702573776245,
      "epoch": 3.808312128922816,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000368946156333121,
      "loss": 4.4379,
      "mean_token_accuracy": 0.27304104864597323,
      "num_tokens": 10739878.0,
      "step": 13470
    },
    {
      "entropy": 4.707232093811035,
      "epoch": 3.809725756290642,
      "grad_norm": 1.5625,
      "learning_rate": 0.00036885268124395786,
      "loss": 4.5784,
      "mean_token_accuracy": 0.25505097061395643,
      "num_tokens": 10744053.0,
      "step": 13475
    },
    {
      "entropy": 4.689425611495972,
      "epoch": 3.8111393836584675,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0003687591865401648,
      "loss": 4.4723,
      "mean_token_accuracy": 0.256505286693573,
      "num_tokens": 10747987.0,
      "step": 13480
    },
    {
      "entropy": 4.695379829406738,
      "epoch": 3.8125530110262935,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003686656722412814,
      "loss": 4.5378,
      "mean_token_accuracy": 0.2586742386221886,
      "num_tokens": 10752019.0,
      "step": 13485
    },
    {
      "entropy": 4.6616432666778564,
      "epoch": 3.8139666383941195,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003685721383668519,
      "loss": 4.4949,
      "mean_token_accuracy": 0.2647451862692833,
      "num_tokens": 10756156.0,
      "step": 13490
    },
    {
      "entropy": 4.707197189331055,
      "epoch": 3.815380265761945,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000368478584936424,
      "loss": 4.4563,
      "mean_token_accuracy": 0.2596962884068489,
      "num_tokens": 10759875.0,
      "step": 13495
    },
    {
      "entropy": 4.708128118515015,
      "epoch": 3.816793893129771,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003683850119695499,
      "loss": 4.5909,
      "mean_token_accuracy": 0.2546014726161957,
      "num_tokens": 10763786.0,
      "step": 13500
    },
    {
      "entropy": 4.819493818283081,
      "epoch": 3.8182075204975967,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00036829141948578576,
      "loss": 4.6961,
      "mean_token_accuracy": 0.23661780953407288,
      "num_tokens": 10767856.0,
      "step": 13505
    },
    {
      "entropy": 4.731859970092773,
      "epoch": 3.8196211478654227,
      "grad_norm": 1.515625,
      "learning_rate": 0.00036819780750469187,
      "loss": 4.5137,
      "mean_token_accuracy": 0.2538356021046638,
      "num_tokens": 10772041.0,
      "step": 13510
    },
    {
      "entropy": 4.645097541809082,
      "epoch": 3.8210347752332483,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003681041760458326,
      "loss": 4.4226,
      "mean_token_accuracy": 0.2690298229455948,
      "num_tokens": 10775850.0,
      "step": 13515
    },
    {
      "entropy": 4.722962856292725,
      "epoch": 3.8224484026010743,
      "grad_norm": 1.59375,
      "learning_rate": 0.00036801052512877623,
      "loss": 4.5856,
      "mean_token_accuracy": 0.24789937138557433,
      "num_tokens": 10779710.0,
      "step": 13520
    },
    {
      "entropy": 4.751594829559326,
      "epoch": 3.8238620299689003,
      "grad_norm": 1.984375,
      "learning_rate": 0.0003679168547730952,
      "loss": 4.6281,
      "mean_token_accuracy": 0.2423988625407219,
      "num_tokens": 10783627.0,
      "step": 13525
    },
    {
      "entropy": 4.830720329284668,
      "epoch": 3.825275657336726,
      "grad_norm": 1.71875,
      "learning_rate": 0.00036782316499836613,
      "loss": 4.6508,
      "mean_token_accuracy": 0.240307055413723,
      "num_tokens": 10787675.0,
      "step": 13530
    },
    {
      "entropy": 4.780664587020874,
      "epoch": 3.826689284704552,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003677294558241696,
      "loss": 4.5054,
      "mean_token_accuracy": 0.2568957552313805,
      "num_tokens": 10791709.0,
      "step": 13535
    },
    {
      "entropy": 4.783442401885987,
      "epoch": 3.828102912072378,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00036763572727009023,
      "loss": 4.5715,
      "mean_token_accuracy": 0.25845580250024797,
      "num_tokens": 10795848.0,
      "step": 13540
    },
    {
      "entropy": 4.688379764556885,
      "epoch": 3.8295165394402035,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003675419793557168,
      "loss": 4.5092,
      "mean_token_accuracy": 0.25617930591106414,
      "num_tokens": 10799662.0,
      "step": 13545
    },
    {
      "entropy": 4.658485984802246,
      "epoch": 3.8309301668080296,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003674482121006418,
      "loss": 4.5406,
      "mean_token_accuracy": 0.25622835457324983,
      "num_tokens": 10803968.0,
      "step": 13550
    },
    {
      "entropy": 4.657243013381958,
      "epoch": 3.832343794175855,
      "grad_norm": 1.625,
      "learning_rate": 0.0003673544255244622,
      "loss": 4.5293,
      "mean_token_accuracy": 0.25068102180957796,
      "num_tokens": 10807810.0,
      "step": 13555
    },
    {
      "entropy": 4.823262739181518,
      "epoch": 3.833757421543681,
      "grad_norm": 1.625,
      "learning_rate": 0.00036726061964677877,
      "loss": 4.6663,
      "mean_token_accuracy": 0.24167069494724275,
      "num_tokens": 10811624.0,
      "step": 13560
    },
    {
      "entropy": 4.826300764083863,
      "epoch": 3.8351710489115067,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003671667944871963,
      "loss": 4.6269,
      "mean_token_accuracy": 0.2423076167702675,
      "num_tokens": 10815806.0,
      "step": 13565
    },
    {
      "entropy": 4.684281969070435,
      "epoch": 3.8365846762793328,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00036707295006532375,
      "loss": 4.4162,
      "mean_token_accuracy": 0.27242272794246675,
      "num_tokens": 10819942.0,
      "step": 13570
    },
    {
      "entropy": 4.623341846466064,
      "epoch": 3.8379983036471588,
      "grad_norm": 1.46875,
      "learning_rate": 0.0003669790864007741,
      "loss": 4.4054,
      "mean_token_accuracy": 0.25811880081892014,
      "num_tokens": 10823975.0,
      "step": 13575
    },
    {
      "entropy": 4.589093589782715,
      "epoch": 3.8394119310149843,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003668852035131641,
      "loss": 4.4289,
      "mean_token_accuracy": 0.26519284695386885,
      "num_tokens": 10827729.0,
      "step": 13580
    },
    {
      "entropy": 4.698354816436767,
      "epoch": 3.8408255583828104,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00036679130142211474,
      "loss": 4.6191,
      "mean_token_accuracy": 0.24654134213924409,
      "num_tokens": 10831710.0,
      "step": 13585
    },
    {
      "entropy": 4.604273176193237,
      "epoch": 3.8422391857506364,
      "grad_norm": 1.734375,
      "learning_rate": 0.000366697380147251,
      "loss": 4.447,
      "mean_token_accuracy": 0.26096131950616835,
      "num_tokens": 10835626.0,
      "step": 13590
    },
    {
      "entropy": 4.717008113861084,
      "epoch": 3.843652813118462,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003666034397082019,
      "loss": 4.5091,
      "mean_token_accuracy": 0.24994847923517227,
      "num_tokens": 10839444.0,
      "step": 13595
    },
    {
      "entropy": 4.647537612915039,
      "epoch": 3.8450664404862875,
      "grad_norm": 1.90625,
      "learning_rate": 0.00036650948012460034,
      "loss": 4.526,
      "mean_token_accuracy": 0.2675424799323082,
      "num_tokens": 10843356.0,
      "step": 13600
    },
    {
      "entropy": 4.800908470153809,
      "epoch": 3.8464800678541136,
      "grad_norm": 1.796875,
      "learning_rate": 0.00036641550141608333,
      "loss": 4.6141,
      "mean_token_accuracy": 0.23876935690641404,
      "num_tokens": 10847441.0,
      "step": 13605
    },
    {
      "entropy": 4.696867799758911,
      "epoch": 3.8478936952219396,
      "grad_norm": 1.515625,
      "learning_rate": 0.00036632150360229183,
      "loss": 4.5085,
      "mean_token_accuracy": 0.25811854004859924,
      "num_tokens": 10851335.0,
      "step": 13610
    },
    {
      "entropy": 4.65193362236023,
      "epoch": 3.849307322589765,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00036622748670287077,
      "loss": 4.4839,
      "mean_token_accuracy": 0.263127276301384,
      "num_tokens": 10855358.0,
      "step": 13615
    },
    {
      "entropy": 4.734093761444091,
      "epoch": 3.850720949957591,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003661334507374691,
      "loss": 4.6573,
      "mean_token_accuracy": 0.25243519693613053,
      "num_tokens": 10859396.0,
      "step": 13620
    },
    {
      "entropy": 4.779131126403809,
      "epoch": 3.852134577325417,
      "grad_norm": 1.53125,
      "learning_rate": 0.00036603939572573974,
      "loss": 4.581,
      "mean_token_accuracy": 0.2476092219352722,
      "num_tokens": 10863471.0,
      "step": 13625
    },
    {
      "entropy": 4.720531702041626,
      "epoch": 3.8535482046932428,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0003659453216873397,
      "loss": 4.4426,
      "mean_token_accuracy": 0.2573708087205887,
      "num_tokens": 10867147.0,
      "step": 13630
    },
    {
      "entropy": 4.761018133163452,
      "epoch": 3.854961832061069,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00036585122864192987,
      "loss": 4.6269,
      "mean_token_accuracy": 0.25386121273040774,
      "num_tokens": 10871423.0,
      "step": 13635
    },
    {
      "entropy": 4.627726316452026,
      "epoch": 3.856375459428895,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000365757116609175,
      "loss": 4.436,
      "mean_token_accuracy": 0.26882972866296767,
      "num_tokens": 10875526.0,
      "step": 13640
    },
    {
      "entropy": 4.727829360961914,
      "epoch": 3.8577890867967204,
      "grad_norm": 1.4375,
      "learning_rate": 0.0003656629856087439,
      "loss": 4.5571,
      "mean_token_accuracy": 0.2405529499053955,
      "num_tokens": 10879598.0,
      "step": 13645
    },
    {
      "entropy": 4.832670450210571,
      "epoch": 3.859202714164546,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003655688356603096,
      "loss": 4.5489,
      "mean_token_accuracy": 0.24314497113227845,
      "num_tokens": 10883536.0,
      "step": 13650
    },
    {
      "entropy": 4.79330153465271,
      "epoch": 3.860616341532372,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00036547466678354864,
      "loss": 4.5238,
      "mean_token_accuracy": 0.24550267308950424,
      "num_tokens": 10887367.0,
      "step": 13655
    },
    {
      "entropy": 4.750546312332153,
      "epoch": 3.862029968900198,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00036538047899814177,
      "loss": 4.5702,
      "mean_token_accuracy": 0.25028165727853774,
      "num_tokens": 10891398.0,
      "step": 13660
    },
    {
      "entropy": 4.689729118347168,
      "epoch": 3.8634435962680236,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00036528627232377376,
      "loss": 4.6253,
      "mean_token_accuracy": 0.2443688228726387,
      "num_tokens": 10895602.0,
      "step": 13665
    },
    {
      "entropy": 4.712592554092407,
      "epoch": 3.8648572236358496,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003651920467801331,
      "loss": 4.4786,
      "mean_token_accuracy": 0.26014590859413145,
      "num_tokens": 10899647.0,
      "step": 13670
    },
    {
      "entropy": 4.732739877700806,
      "epoch": 3.8662708510036756,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00036509780238691244,
      "loss": 4.6037,
      "mean_token_accuracy": 0.24546402096748351,
      "num_tokens": 10903697.0,
      "step": 13675
    },
    {
      "entropy": 4.745795869827271,
      "epoch": 3.867684478371501,
      "grad_norm": 1.671875,
      "learning_rate": 0.00036500353916380814,
      "loss": 4.5779,
      "mean_token_accuracy": 0.25354787409305574,
      "num_tokens": 10907797.0,
      "step": 13680
    },
    {
      "entropy": 4.706536102294922,
      "epoch": 3.869098105739327,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003649092571305208,
      "loss": 4.4039,
      "mean_token_accuracy": 0.2708014205098152,
      "num_tokens": 10911461.0,
      "step": 13685
    },
    {
      "entropy": 4.590805006027222,
      "epoch": 3.870511733107153,
      "grad_norm": 1.703125,
      "learning_rate": 0.00036481495630675463,
      "loss": 4.4574,
      "mean_token_accuracy": 0.2664432987570763,
      "num_tokens": 10915160.0,
      "step": 13690
    },
    {
      "entropy": 4.701899671554566,
      "epoch": 3.871925360474979,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003647206367122181,
      "loss": 4.4927,
      "mean_token_accuracy": 0.262466062605381,
      "num_tokens": 10919317.0,
      "step": 13695
    },
    {
      "entropy": 4.702951908111572,
      "epoch": 3.8733389878428044,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00036462629836662324,
      "loss": 4.4426,
      "mean_token_accuracy": 0.2592417478561401,
      "num_tokens": 10923042.0,
      "step": 13700
    },
    {
      "entropy": 4.759500551223755,
      "epoch": 3.8747526152106304,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003645319412896863,
      "loss": 4.6193,
      "mean_token_accuracy": 0.24694173783063889,
      "num_tokens": 10927169.0,
      "step": 13705
    },
    {
      "entropy": 4.628828620910644,
      "epoch": 3.8761662425784564,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003644375655011272,
      "loss": 4.465,
      "mean_token_accuracy": 0.26444774866104126,
      "num_tokens": 10931332.0,
      "step": 13710
    },
    {
      "entropy": 4.71375527381897,
      "epoch": 3.877579869946282,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00036434317102067,
      "loss": 4.6369,
      "mean_token_accuracy": 0.24876486957073213,
      "num_tokens": 10935179.0,
      "step": 13715
    },
    {
      "entropy": 4.768712902069092,
      "epoch": 3.878993497314108,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003642487578680426,
      "loss": 4.7069,
      "mean_token_accuracy": 0.23950154334306717,
      "num_tokens": 10939205.0,
      "step": 13720
    },
    {
      "entropy": 4.72045783996582,
      "epoch": 3.880407124681934,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003641543260629766,
      "loss": 4.5485,
      "mean_token_accuracy": 0.26176293939352036,
      "num_tokens": 10943355.0,
      "step": 13725
    },
    {
      "entropy": 4.70264253616333,
      "epoch": 3.8818207520497596,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003640598756252079,
      "loss": 4.4972,
      "mean_token_accuracy": 0.25145269334316256,
      "num_tokens": 10947522.0,
      "step": 13730
    },
    {
      "entropy": 4.701862049102783,
      "epoch": 3.8832343794175856,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00036396540657447586,
      "loss": 4.5151,
      "mean_token_accuracy": 0.25462574511766434,
      "num_tokens": 10951639.0,
      "step": 13735
    },
    {
      "entropy": 4.671893548965454,
      "epoch": 3.8846480067854112,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003638709189305239,
      "loss": 4.5706,
      "mean_token_accuracy": 0.2611603081226349,
      "num_tokens": 10955874.0,
      "step": 13740
    },
    {
      "entropy": 4.710419416427612,
      "epoch": 3.8860616341532372,
      "grad_norm": 1.625,
      "learning_rate": 0.0003637764127130996,
      "loss": 4.4994,
      "mean_token_accuracy": 0.2509218782186508,
      "num_tokens": 10959397.0,
      "step": 13745
    },
    {
      "entropy": 4.794018316268921,
      "epoch": 3.887475261521063,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003636818879419539,
      "loss": 4.5376,
      "mean_token_accuracy": 0.25946378856897356,
      "num_tokens": 10963260.0,
      "step": 13750
    },
    {
      "entropy": 4.747557735443115,
      "epoch": 3.888888888888889,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00036358734463684203,
      "loss": 4.5591,
      "mean_token_accuracy": 0.2589696556329727,
      "num_tokens": 10967343.0,
      "step": 13755
    },
    {
      "entropy": 4.658880519866943,
      "epoch": 3.890302516256715,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000363492782817523,
      "loss": 4.444,
      "mean_token_accuracy": 0.26548199355602264,
      "num_tokens": 10971560.0,
      "step": 13760
    },
    {
      "entropy": 4.743999719619751,
      "epoch": 3.8917161436245404,
      "grad_norm": 1.5625,
      "learning_rate": 0.00036339820250375955,
      "loss": 4.6402,
      "mean_token_accuracy": 0.251124282181263,
      "num_tokens": 10975785.0,
      "step": 13765
    },
    {
      "entropy": 4.735117244720459,
      "epoch": 3.8931297709923665,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003633036037153184,
      "loss": 4.5142,
      "mean_token_accuracy": 0.25194386690855025,
      "num_tokens": 10980009.0,
      "step": 13770
    },
    {
      "entropy": 4.770359659194947,
      "epoch": 3.8945433983601925,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00036320898647197015,
      "loss": 4.5513,
      "mean_token_accuracy": 0.2461131438612938,
      "num_tokens": 10983661.0,
      "step": 13775
    },
    {
      "entropy": 4.739265632629395,
      "epoch": 3.895957025728018,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003631143507934892,
      "loss": 4.6487,
      "mean_token_accuracy": 0.24465200006961824,
      "num_tokens": 10987580.0,
      "step": 13780
    },
    {
      "entropy": 4.685715007781982,
      "epoch": 3.897370653095844,
      "grad_norm": 1.5625,
      "learning_rate": 0.00036301969669965377,
      "loss": 4.4781,
      "mean_token_accuracy": 0.2601613000035286,
      "num_tokens": 10991294.0,
      "step": 13785
    },
    {
      "entropy": 4.655831432342529,
      "epoch": 3.8987842804636696,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003629250242102461,
      "loss": 4.4474,
      "mean_token_accuracy": 0.25936087816953657,
      "num_tokens": 10995084.0,
      "step": 13790
    },
    {
      "entropy": 4.716451025009155,
      "epoch": 3.9001979078314957,
      "grad_norm": 1.640625,
      "learning_rate": 0.000362830333345052,
      "loss": 4.5739,
      "mean_token_accuracy": 0.25664752870798113,
      "num_tokens": 10999137.0,
      "step": 13795
    },
    {
      "entropy": 4.76203932762146,
      "epoch": 3.9016115351993212,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003627356241238614,
      "loss": 4.7535,
      "mean_token_accuracy": 0.2349080488085747,
      "num_tokens": 11003111.0,
      "step": 13800
    },
    {
      "entropy": 4.871933031082153,
      "epoch": 3.9030251625671473,
      "grad_norm": 1.53125,
      "learning_rate": 0.00036264089656646794,
      "loss": 4.6274,
      "mean_token_accuracy": 0.24091254323720931,
      "num_tokens": 11007150.0,
      "step": 13805
    },
    {
      "entropy": 4.824051189422607,
      "epoch": 3.9044387899349733,
      "grad_norm": 1.7578125,
      "learning_rate": 0.000362546150692669,
      "loss": 4.5771,
      "mean_token_accuracy": 0.25935512632131574,
      "num_tokens": 11011001.0,
      "step": 13810
    },
    {
      "entropy": 4.684522104263306,
      "epoch": 3.905852417302799,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00036245138652226587,
      "loss": 4.4042,
      "mean_token_accuracy": 0.26542300134897234,
      "num_tokens": 11015087.0,
      "step": 13815
    },
    {
      "entropy": 4.641892671585083,
      "epoch": 3.907266044670625,
      "grad_norm": 1.625,
      "learning_rate": 0.00036235660407506393,
      "loss": 4.4756,
      "mean_token_accuracy": 0.2651048481464386,
      "num_tokens": 11018983.0,
      "step": 13820
    },
    {
      "entropy": 4.612433958053589,
      "epoch": 3.908679672038451,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003622618033708718,
      "loss": 4.4417,
      "mean_token_accuracy": 0.2618775025010109,
      "num_tokens": 11022680.0,
      "step": 13825
    },
    {
      "entropy": 4.6027185916900635,
      "epoch": 3.9100932994062765,
      "grad_norm": 1.640625,
      "learning_rate": 0.00036216698442950235,
      "loss": 4.4549,
      "mean_token_accuracy": 0.25995580703020094,
      "num_tokens": 11026485.0,
      "step": 13830
    },
    {
      "entropy": 4.731858777999878,
      "epoch": 3.9115069267741025,
      "grad_norm": 1.5625,
      "learning_rate": 0.00036207214727077225,
      "loss": 4.5938,
      "mean_token_accuracy": 0.25330419689416883,
      "num_tokens": 11030481.0,
      "step": 13835
    },
    {
      "entropy": 4.8124227046966555,
      "epoch": 3.912920554141928,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003619772919145017,
      "loss": 4.6869,
      "mean_token_accuracy": 0.2474244937300682,
      "num_tokens": 11034362.0,
      "step": 13840
    },
    {
      "entropy": 4.699274921417237,
      "epoch": 3.914334181509754,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003618824183805152,
      "loss": 4.523,
      "mean_token_accuracy": 0.25991071164608004,
      "num_tokens": 11038481.0,
      "step": 13845
    },
    {
      "entropy": 4.7684625625610355,
      "epoch": 3.9157478088775797,
      "grad_norm": 1.703125,
      "learning_rate": 0.00036178752668864037,
      "loss": 4.5653,
      "mean_token_accuracy": 0.24899394810199738,
      "num_tokens": 11042380.0,
      "step": 13850
    },
    {
      "entropy": 4.7072892665863035,
      "epoch": 3.9171614362454057,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003616926168587092,
      "loss": 4.495,
      "mean_token_accuracy": 0.254258306324482,
      "num_tokens": 11046427.0,
      "step": 13855
    },
    {
      "entropy": 4.755464029312134,
      "epoch": 3.9185750636132317,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003615976889105571,
      "loss": 4.5506,
      "mean_token_accuracy": 0.26559149026870726,
      "num_tokens": 11050420.0,
      "step": 13860
    },
    {
      "entropy": 4.748749256134033,
      "epoch": 3.9199886909810573,
      "grad_norm": 1.609375,
      "learning_rate": 0.00036150274286402364,
      "loss": 4.6067,
      "mean_token_accuracy": 0.254745677113533,
      "num_tokens": 11054155.0,
      "step": 13865
    },
    {
      "entropy": 4.643721866607666,
      "epoch": 3.9214023183488833,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00036140777873895175,
      "loss": 4.4944,
      "mean_token_accuracy": 0.26806291192770004,
      "num_tokens": 11057994.0,
      "step": 13870
    },
    {
      "entropy": 4.6595570087432865,
      "epoch": 3.9228159457167093,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003613127965551884,
      "loss": 4.478,
      "mean_token_accuracy": 0.26679911762475966,
      "num_tokens": 11062093.0,
      "step": 13875
    },
    {
      "entropy": 4.674387121200562,
      "epoch": 3.924229573084535,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00036121779633258445,
      "loss": 4.6032,
      "mean_token_accuracy": 0.25576215982437134,
      "num_tokens": 11066000.0,
      "step": 13880
    },
    {
      "entropy": 4.651311016082763,
      "epoch": 3.9256432004523605,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003611227780909942,
      "loss": 4.4779,
      "mean_token_accuracy": 0.256543955206871,
      "num_tokens": 11069585.0,
      "step": 13885
    },
    {
      "entropy": 4.672791481018066,
      "epoch": 3.9270568278201865,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003610277418502758,
      "loss": 4.4732,
      "mean_token_accuracy": 0.2595260918140411,
      "num_tokens": 11073645.0,
      "step": 13890
    },
    {
      "entropy": 4.643919181823731,
      "epoch": 3.9284704551880125,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003609326876302914,
      "loss": 4.5075,
      "mean_token_accuracy": 0.2665888205170631,
      "num_tokens": 11077831.0,
      "step": 13895
    },
    {
      "entropy": 4.695799493789673,
      "epoch": 3.929884082555838,
      "grad_norm": 1.90625,
      "learning_rate": 0.00036083761545090663,
      "loss": 4.6069,
      "mean_token_accuracy": 0.2395849868655205,
      "num_tokens": 11081745.0,
      "step": 13900
    },
    {
      "entropy": 4.803744840621948,
      "epoch": 3.931297709923664,
      "grad_norm": 1.71875,
      "learning_rate": 0.00036074252533199107,
      "loss": 4.6077,
      "mean_token_accuracy": 0.25490506291389464,
      "num_tokens": 11085719.0,
      "step": 13905
    },
    {
      "entropy": 4.827231597900391,
      "epoch": 3.93271133729149,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00036064741729341794,
      "loss": 4.5358,
      "mean_token_accuracy": 0.2578920409083366,
      "num_tokens": 11089657.0,
      "step": 13910
    },
    {
      "entropy": 4.674139356613159,
      "epoch": 3.9341249646593157,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00036055229135506425,
      "loss": 4.449,
      "mean_token_accuracy": 0.264768648147583,
      "num_tokens": 11093535.0,
      "step": 13915
    },
    {
      "entropy": 4.684307479858399,
      "epoch": 3.9355385920271417,
      "grad_norm": 1.546875,
      "learning_rate": 0.00036045714753681063,
      "loss": 4.5793,
      "mean_token_accuracy": 0.2524320766329765,
      "num_tokens": 11097720.0,
      "step": 13920
    },
    {
      "entropy": 4.610172748565674,
      "epoch": 3.9369522193949678,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003603619858585417,
      "loss": 4.6039,
      "mean_token_accuracy": 0.24972153306007386,
      "num_tokens": 11101852.0,
      "step": 13925
    },
    {
      "entropy": 4.64522500038147,
      "epoch": 3.9383658467627933,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003602668063401455,
      "loss": 4.4243,
      "mean_token_accuracy": 0.27786844223737717,
      "num_tokens": 11105769.0,
      "step": 13930
    },
    {
      "entropy": 4.776746845245361,
      "epoch": 3.939779474130619,
      "grad_norm": 1.625,
      "learning_rate": 0.0003601716090015141,
      "loss": 4.5834,
      "mean_token_accuracy": 0.2530807003378868,
      "num_tokens": 11109650.0,
      "step": 13935
    },
    {
      "entropy": 4.7766119003295895,
      "epoch": 3.941193101498445,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003600763938625431,
      "loss": 4.56,
      "mean_token_accuracy": 0.25370156615972517,
      "num_tokens": 11113672.0,
      "step": 13940
    },
    {
      "entropy": 4.680834674835205,
      "epoch": 3.942606728866271,
      "grad_norm": 1.515625,
      "learning_rate": 0.00035998116094313186,
      "loss": 4.4887,
      "mean_token_accuracy": 0.2614923223853111,
      "num_tokens": 11117589.0,
      "step": 13945
    },
    {
      "entropy": 4.73240327835083,
      "epoch": 3.9440203562340965,
      "grad_norm": 1.640625,
      "learning_rate": 0.00035988591026318357,
      "loss": 4.5147,
      "mean_token_accuracy": 0.2624636247754097,
      "num_tokens": 11121311.0,
      "step": 13950
    },
    {
      "entropy": 4.739521217346192,
      "epoch": 3.9454339836019225,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003597906418426049,
      "loss": 4.5335,
      "mean_token_accuracy": 0.2511450365185738,
      "num_tokens": 11125087.0,
      "step": 13955
    },
    {
      "entropy": 4.693289279937744,
      "epoch": 3.9468476109697486,
      "grad_norm": 1.625,
      "learning_rate": 0.0003596953557013064,
      "loss": 4.4361,
      "mean_token_accuracy": 0.2641616716980934,
      "num_tokens": 11129020.0,
      "step": 13960
    },
    {
      "entropy": 4.685501956939698,
      "epoch": 3.948261238337574,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00035960005185920233,
      "loss": 4.4772,
      "mean_token_accuracy": 0.2564189746975899,
      "num_tokens": 11133082.0,
      "step": 13965
    },
    {
      "entropy": 4.642407846450806,
      "epoch": 3.9496748657054,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0003595047303362105,
      "loss": 4.5061,
      "mean_token_accuracy": 0.25369016975164416,
      "num_tokens": 11136863.0,
      "step": 13970
    },
    {
      "entropy": 4.796057224273682,
      "epoch": 3.9510884930732257,
      "grad_norm": 1.703125,
      "learning_rate": 0.00035940939115225266,
      "loss": 4.5496,
      "mean_token_accuracy": 0.25820010900497437,
      "num_tokens": 11140824.0,
      "step": 13975
    },
    {
      "entropy": 4.6784255027771,
      "epoch": 3.9525021204410518,
      "grad_norm": 1.578125,
      "learning_rate": 0.00035931403432725405,
      "loss": 4.4753,
      "mean_token_accuracy": 0.25732877254486086,
      "num_tokens": 11144749.0,
      "step": 13980
    },
    {
      "entropy": 4.65656270980835,
      "epoch": 3.9539157478088773,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003592186598811436,
      "loss": 4.5657,
      "mean_token_accuracy": 0.2549586981534958,
      "num_tokens": 11148755.0,
      "step": 13985
    },
    {
      "entropy": 4.6274574279785154,
      "epoch": 3.9553293751767034,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003591232678338541,
      "loss": 4.4023,
      "mean_token_accuracy": 0.2560295104980469,
      "num_tokens": 11152540.0,
      "step": 13990
    },
    {
      "entropy": 4.758601570129395,
      "epoch": 3.9567430025445294,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00035902785820532174,
      "loss": 4.5323,
      "mean_token_accuracy": 0.2622156009078026,
      "num_tokens": 11156358.0,
      "step": 13995
    },
    {
      "entropy": 4.731723785400391,
      "epoch": 3.958156629912355,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00035893243101548666,
      "loss": 4.5618,
      "mean_token_accuracy": 0.2525500953197479,
      "num_tokens": 11160395.0,
      "step": 14000
    },
    {
      "epoch": 3.958156629912355,
      "eval_entropy": 4.525367460210301,
      "eval_loss": 5.033152103424072,
      "eval_mean_token_accuracy": 0.24448419501179924,
      "eval_num_tokens": 11160395.0,
      "eval_runtime": 6.701,
      "eval_samples_per_second": 1683.929,
      "eval_steps_per_second": 210.566,
      "step": 14000
    },
    {
      "entropy": 4.564944553375244,
      "epoch": 3.959570257280181,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00035883698628429254,
      "loss": 4.444,
      "mean_token_accuracy": 0.26314639747142793,
      "num_tokens": 11164215.0,
      "step": 14005
    },
    {
      "entropy": 4.692602300643921,
      "epoch": 3.960983884648007,
      "grad_norm": 1.625,
      "learning_rate": 0.0003587415240316867,
      "loss": 4.5461,
      "mean_token_accuracy": 0.24492590576410295,
      "num_tokens": 11168144.0,
      "step": 14010
    },
    {
      "entropy": 4.795106983184814,
      "epoch": 3.9623975120158326,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035864604427762015,
      "loss": 4.5564,
      "mean_token_accuracy": 0.2524279609322548,
      "num_tokens": 11172097.0,
      "step": 14015
    },
    {
      "entropy": 4.7735954284667965,
      "epoch": 3.9638111393836586,
      "grad_norm": 1.5625,
      "learning_rate": 0.00035855054704204763,
      "loss": 4.5956,
      "mean_token_accuracy": 0.2519525095820427,
      "num_tokens": 11175921.0,
      "step": 14020
    },
    {
      "entropy": 4.829806184768676,
      "epoch": 3.965224766751484,
      "grad_norm": 1.484375,
      "learning_rate": 0.00035845503234492745,
      "loss": 4.6494,
      "mean_token_accuracy": 0.2417868122458458,
      "num_tokens": 11180389.0,
      "step": 14025
    },
    {
      "entropy": 4.74340410232544,
      "epoch": 3.96663839411931,
      "grad_norm": 1.8125,
      "learning_rate": 0.00035835950020622154,
      "loss": 4.5962,
      "mean_token_accuracy": 0.2405349940061569,
      "num_tokens": 11184089.0,
      "step": 14030
    },
    {
      "entropy": 4.785094356536865,
      "epoch": 3.9680520214871358,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00035826395064589555,
      "loss": 4.6222,
      "mean_token_accuracy": 0.2507664978504181,
      "num_tokens": 11188151.0,
      "step": 14035
    },
    {
      "entropy": 4.772581624984741,
      "epoch": 3.969465648854962,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003581683836839189,
      "loss": 4.5316,
      "mean_token_accuracy": 0.24916147589683532,
      "num_tokens": 11192239.0,
      "step": 14040
    },
    {
      "entropy": 4.685791015625,
      "epoch": 3.970879276222788,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00035807279934026425,
      "loss": 4.4602,
      "mean_token_accuracy": 0.25974666476249697,
      "num_tokens": 11196259.0,
      "step": 14045
    },
    {
      "entropy": 4.714363098144531,
      "epoch": 3.9722929035906134,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00035797719763490827,
      "loss": 4.4888,
      "mean_token_accuracy": 0.25227752327919006,
      "num_tokens": 11200112.0,
      "step": 14050
    },
    {
      "entropy": 4.73885555267334,
      "epoch": 3.9737065309584394,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035788157858783107,
      "loss": 4.5853,
      "mean_token_accuracy": 0.2523780897259712,
      "num_tokens": 11204051.0,
      "step": 14055
    },
    {
      "entropy": 4.655237102508545,
      "epoch": 3.9751201583262654,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003577859422190165,
      "loss": 4.5092,
      "mean_token_accuracy": 0.253102970123291,
      "num_tokens": 11207815.0,
      "step": 14060
    },
    {
      "entropy": 4.710619401931763,
      "epoch": 3.976533785694091,
      "grad_norm": 1.6875,
      "learning_rate": 0.00035769028854845187,
      "loss": 4.5405,
      "mean_token_accuracy": 0.25317621529102324,
      "num_tokens": 11211785.0,
      "step": 14065
    },
    {
      "entropy": 4.710752725601196,
      "epoch": 3.977947413061917,
      "grad_norm": 1.671875,
      "learning_rate": 0.00035759461759612846,
      "loss": 4.5411,
      "mean_token_accuracy": 0.24933832734823227,
      "num_tokens": 11215778.0,
      "step": 14070
    },
    {
      "entropy": 4.794663190841675,
      "epoch": 3.9793610404297426,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00035749892938204055,
      "loss": 4.4888,
      "mean_token_accuracy": 0.2633426323533058,
      "num_tokens": 11220010.0,
      "step": 14075
    },
    {
      "entropy": 4.655690860748291,
      "epoch": 3.9807746677975686,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003574032239261866,
      "loss": 4.4977,
      "mean_token_accuracy": 0.2546457052230835,
      "num_tokens": 11224174.0,
      "step": 14080
    },
    {
      "entropy": 4.757253742218017,
      "epoch": 3.982188295165394,
      "grad_norm": 1.578125,
      "learning_rate": 0.00035730750124856844,
      "loss": 4.5998,
      "mean_token_accuracy": 0.25451924800872805,
      "num_tokens": 11228294.0,
      "step": 14085
    },
    {
      "entropy": 4.637308359146118,
      "epoch": 3.98360192253322,
      "grad_norm": 1.625,
      "learning_rate": 0.0003572117613691914,
      "loss": 4.4652,
      "mean_token_accuracy": 0.2567900106310844,
      "num_tokens": 11232038.0,
      "step": 14090
    },
    {
      "entropy": 4.718468809127808,
      "epoch": 3.9850155499010462,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003571160043080647,
      "loss": 4.5959,
      "mean_token_accuracy": 0.2553757309913635,
      "num_tokens": 11236059.0,
      "step": 14095
    },
    {
      "entropy": 4.684141016006469,
      "epoch": 3.986429177268872,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003570202300852009,
      "loss": 4.5068,
      "mean_token_accuracy": 0.2597695991396904,
      "num_tokens": 11240093.0,
      "step": 14100
    },
    {
      "entropy": 4.759722328186035,
      "epoch": 3.987842804636698,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003569244387206162,
      "loss": 4.6077,
      "mean_token_accuracy": 0.2516933292150497,
      "num_tokens": 11244128.0,
      "step": 14105
    },
    {
      "entropy": 4.758224725723267,
      "epoch": 3.989256432004524,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003568286302343304,
      "loss": 4.4174,
      "mean_token_accuracy": 0.26733039915561674,
      "num_tokens": 11248090.0,
      "step": 14110
    },
    {
      "entropy": 4.650763940811157,
      "epoch": 3.9906700593723494,
      "grad_norm": 1.625,
      "learning_rate": 0.00035673280464636686,
      "loss": 4.4028,
      "mean_token_accuracy": 0.2709064304828644,
      "num_tokens": 11252163.0,
      "step": 14115
    },
    {
      "entropy": 4.619721078872681,
      "epoch": 3.9920836867401754,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003566369619767526,
      "loss": 4.43,
      "mean_token_accuracy": 0.2696743875741959,
      "num_tokens": 11255952.0,
      "step": 14120
    },
    {
      "entropy": 4.630394411087036,
      "epoch": 3.993497314108001,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003565411022455181,
      "loss": 4.445,
      "mean_token_accuracy": 0.25772920399904253,
      "num_tokens": 11260078.0,
      "step": 14125
    },
    {
      "entropy": 4.768359756469726,
      "epoch": 3.994910941475827,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003564452254726975,
      "loss": 4.6743,
      "mean_token_accuracy": 0.23542530089616776,
      "num_tokens": 11264243.0,
      "step": 14130
    },
    {
      "entropy": 4.837106323242187,
      "epoch": 3.9963245688436526,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003563493316783284,
      "loss": 4.539,
      "mean_token_accuracy": 0.26083938777446747,
      "num_tokens": 11268399.0,
      "step": 14135
    },
    {
      "entropy": 4.632800674438476,
      "epoch": 3.9977381962114786,
      "grad_norm": 1.625,
      "learning_rate": 0.00035625342088245205,
      "loss": 4.4666,
      "mean_token_accuracy": 0.2759882763028145,
      "num_tokens": 11272315.0,
      "step": 14140
    },
    {
      "entropy": 4.699775886535645,
      "epoch": 3.9991518235793047,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003561574931051131,
      "loss": 4.5503,
      "mean_token_accuracy": 0.24732542634010315,
      "num_tokens": 11276418.0,
      "step": 14145
    },
    {
      "entropy": 4.806932210922241,
      "epoch": 4.00056545094713,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003560615483663601,
      "loss": 4.564,
      "mean_token_accuracy": 0.24293984919786454,
      "num_tokens": 11280528.0,
      "step": 14150
    },
    {
      "entropy": 4.76903977394104,
      "epoch": 4.001979078314956,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00035596558668624464,
      "loss": 4.3961,
      "mean_token_accuracy": 0.2620241865515709,
      "num_tokens": 11284744.0,
      "step": 14155
    },
    {
      "entropy": 4.661413478851318,
      "epoch": 4.003392705682782,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003558696080848224,
      "loss": 4.271,
      "mean_token_accuracy": 0.26673697829246523,
      "num_tokens": 11288974.0,
      "step": 14160
    },
    {
      "entropy": 4.657903575897217,
      "epoch": 4.004806333050608,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00035577361258215203,
      "loss": 4.3596,
      "mean_token_accuracy": 0.25839888155460355,
      "num_tokens": 11293220.0,
      "step": 14165
    },
    {
      "entropy": 4.660568809509277,
      "epoch": 4.006219960418433,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003556776001982961,
      "loss": 4.2887,
      "mean_token_accuracy": 0.2629737541079521,
      "num_tokens": 11297097.0,
      "step": 14170
    },
    {
      "entropy": 4.73074860572815,
      "epoch": 4.00763358778626,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003555815709533206,
      "loss": 4.3451,
      "mean_token_accuracy": 0.2616847544908524,
      "num_tokens": 11301051.0,
      "step": 14175
    },
    {
      "entropy": 4.693859720230103,
      "epoch": 4.0090472151540855,
      "grad_norm": 1.5625,
      "learning_rate": 0.00035548552486729514,
      "loss": 4.36,
      "mean_token_accuracy": 0.25604725778102877,
      "num_tokens": 11304908.0,
      "step": 14180
    },
    {
      "entropy": 4.663771057128907,
      "epoch": 4.010460842521911,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003553894619602927,
      "loss": 4.2906,
      "mean_token_accuracy": 0.2763788655400276,
      "num_tokens": 11309080.0,
      "step": 14185
    },
    {
      "entropy": 4.657538175582886,
      "epoch": 4.0118744698897375,
      "grad_norm": 1.703125,
      "learning_rate": 0.00035529338225238975,
      "loss": 4.33,
      "mean_token_accuracy": 0.28188861459493636,
      "num_tokens": 11313028.0,
      "step": 14190
    },
    {
      "entropy": 4.57024507522583,
      "epoch": 4.013288097257563,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003551972857636665,
      "loss": 4.1935,
      "mean_token_accuracy": 0.281586717069149,
      "num_tokens": 11316921.0,
      "step": 14195
    },
    {
      "entropy": 4.652089643478393,
      "epoch": 4.014701724625389,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00035510117251420645,
      "loss": 4.3529,
      "mean_token_accuracy": 0.2657230168581009,
      "num_tokens": 11321278.0,
      "step": 14200
    },
    {
      "entropy": 4.737310791015625,
      "epoch": 4.016115351993214,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003550050425240966,
      "loss": 4.4427,
      "mean_token_accuracy": 0.2586666375398636,
      "num_tokens": 11325204.0,
      "step": 14205
    },
    {
      "entropy": 4.781671714782715,
      "epoch": 4.017528979361041,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003549088958134277,
      "loss": 4.3765,
      "mean_token_accuracy": 0.26622362434864044,
      "num_tokens": 11329015.0,
      "step": 14210
    },
    {
      "entropy": 4.598283815383911,
      "epoch": 4.018942606728866,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00035481273240229363,
      "loss": 4.2037,
      "mean_token_accuracy": 0.2911985725164413,
      "num_tokens": 11333169.0,
      "step": 14215
    },
    {
      "entropy": 4.681335020065307,
      "epoch": 4.020356234096692,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003547165523107921,
      "loss": 4.3728,
      "mean_token_accuracy": 0.260310035943985,
      "num_tokens": 11336944.0,
      "step": 14220
    },
    {
      "entropy": 4.643592977523804,
      "epoch": 4.021769861464518,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00035462035555902405,
      "loss": 4.4144,
      "mean_token_accuracy": 0.2542284563183784,
      "num_tokens": 11340934.0,
      "step": 14225
    },
    {
      "entropy": 4.704354906082154,
      "epoch": 4.023183488832344,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003545241421670941,
      "loss": 4.4603,
      "mean_token_accuracy": 0.2527497932314873,
      "num_tokens": 11345214.0,
      "step": 14230
    },
    {
      "entropy": 4.835513067245484,
      "epoch": 4.0245971162001695,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035442791215511017,
      "loss": 4.4258,
      "mean_token_accuracy": 0.24980147331953048,
      "num_tokens": 11349294.0,
      "step": 14235
    },
    {
      "entropy": 4.7001971244812015,
      "epoch": 4.026010743567995,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003543316655431837,
      "loss": 4.2908,
      "mean_token_accuracy": 0.2767662093043327,
      "num_tokens": 11353152.0,
      "step": 14240
    },
    {
      "entropy": 4.625879192352295,
      "epoch": 4.0274243709358215,
      "grad_norm": 1.6875,
      "learning_rate": 0.00035423540235142973,
      "loss": 4.3327,
      "mean_token_accuracy": 0.2632051408290863,
      "num_tokens": 11357143.0,
      "step": 14245
    },
    {
      "entropy": 4.465175819396973,
      "epoch": 4.028837998303647,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003541391225999667,
      "loss": 4.2284,
      "mean_token_accuracy": 0.28252696692943574,
      "num_tokens": 11361256.0,
      "step": 14250
    },
    {
      "entropy": 4.633312702178955,
      "epoch": 4.030251625671473,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003540428263089164,
      "loss": 4.291,
      "mean_token_accuracy": 0.27144998759031297,
      "num_tokens": 11365104.0,
      "step": 14255
    },
    {
      "entropy": 4.640343856811524,
      "epoch": 4.031665253039299,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00035394651349840423,
      "loss": 4.2111,
      "mean_token_accuracy": 0.27496292889118196,
      "num_tokens": 11369401.0,
      "step": 14260
    },
    {
      "entropy": 4.7146963596344,
      "epoch": 4.033078880407125,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003538501841885589,
      "loss": 4.4116,
      "mean_token_accuracy": 0.25408882200717925,
      "num_tokens": 11373453.0,
      "step": 14265
    },
    {
      "entropy": 4.691504859924317,
      "epoch": 4.03449250777495,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003537538383995126,
      "loss": 4.305,
      "mean_token_accuracy": 0.26722117215394975,
      "num_tokens": 11377198.0,
      "step": 14270
    },
    {
      "entropy": 4.701403474807739,
      "epoch": 4.035906135142777,
      "grad_norm": 1.5625,
      "learning_rate": 0.00035365747615140114,
      "loss": 4.3798,
      "mean_token_accuracy": 0.2629324674606323,
      "num_tokens": 11381302.0,
      "step": 14275
    },
    {
      "entropy": 4.616743087768555,
      "epoch": 4.037319762510602,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003535610974643635,
      "loss": 4.2863,
      "mean_token_accuracy": 0.27507011890411376,
      "num_tokens": 11385786.0,
      "step": 14280
    },
    {
      "entropy": 4.650902509689331,
      "epoch": 4.038733389878428,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003534647023585425,
      "loss": 4.2588,
      "mean_token_accuracy": 0.2746400773525238,
      "num_tokens": 11389642.0,
      "step": 14285
    },
    {
      "entropy": 4.659925317764282,
      "epoch": 4.0401470172462535,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003533682908540837,
      "loss": 4.3469,
      "mean_token_accuracy": 0.2624311909079552,
      "num_tokens": 11393415.0,
      "step": 14290
    },
    {
      "entropy": 4.647394418716431,
      "epoch": 4.04156064461408,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003532718629711368,
      "loss": 4.3215,
      "mean_token_accuracy": 0.26710559576749804,
      "num_tokens": 11397198.0,
      "step": 14295
    },
    {
      "entropy": 4.6507750988006595,
      "epoch": 4.0429742719819055,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00035317541872985456,
      "loss": 4.2926,
      "mean_token_accuracy": 0.2649114653468132,
      "num_tokens": 11401318.0,
      "step": 14300
    },
    {
      "entropy": 4.673664093017578,
      "epoch": 4.044387899349731,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003530789581503933,
      "loss": 4.378,
      "mean_token_accuracy": 0.2713590428233147,
      "num_tokens": 11405193.0,
      "step": 14305
    },
    {
      "entropy": 4.637875175476074,
      "epoch": 4.0458015267175576,
      "grad_norm": 1.8125,
      "learning_rate": 0.0003529824812529124,
      "loss": 4.3887,
      "mean_token_accuracy": 0.24953527450561525,
      "num_tokens": 11409358.0,
      "step": 14310
    },
    {
      "entropy": 4.601602172851562,
      "epoch": 4.047215154085383,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003528859880575754,
      "loss": 4.2768,
      "mean_token_accuracy": 0.27399891018867495,
      "num_tokens": 11413033.0,
      "step": 14315
    },
    {
      "entropy": 4.625516557693482,
      "epoch": 4.048628781453209,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003527894785845484,
      "loss": 4.3276,
      "mean_token_accuracy": 0.26950441747903825,
      "num_tokens": 11416845.0,
      "step": 14320
    },
    {
      "entropy": 4.707536506652832,
      "epoch": 4.050042408821035,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003526929528540014,
      "loss": 4.332,
      "mean_token_accuracy": 0.26587475538253785,
      "num_tokens": 11421051.0,
      "step": 14325
    },
    {
      "entropy": 4.631922340393066,
      "epoch": 4.051456036188861,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00035259641088610774,
      "loss": 4.3049,
      "mean_token_accuracy": 0.26402136087417605,
      "num_tokens": 11425156.0,
      "step": 14330
    },
    {
      "entropy": 4.6630606174469,
      "epoch": 4.052869663556686,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000352499852701044,
      "loss": 4.3429,
      "mean_token_accuracy": 0.26523649245500563,
      "num_tokens": 11429198.0,
      "step": 14335
    },
    {
      "entropy": 4.659229898452759,
      "epoch": 4.054283290924512,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003524032783189903,
      "loss": 4.4554,
      "mean_token_accuracy": 0.24732918292284012,
      "num_tokens": 11433395.0,
      "step": 14340
    },
    {
      "entropy": 4.756145477294922,
      "epoch": 4.055696918292338,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003523066877601301,
      "loss": 4.3996,
      "mean_token_accuracy": 0.253556963801384,
      "num_tokens": 11437342.0,
      "step": 14345
    },
    {
      "entropy": 4.823345518112182,
      "epoch": 4.057110545660164,
      "grad_norm": 1.90625,
      "learning_rate": 0.00035221008104465026,
      "loss": 4.4743,
      "mean_token_accuracy": 0.25172090232372285,
      "num_tokens": 11441287.0,
      "step": 14350
    },
    {
      "entropy": 4.670554304122925,
      "epoch": 4.0585241730279895,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00035211345819274084,
      "loss": 4.3187,
      "mean_token_accuracy": 0.2712868541479111,
      "num_tokens": 11445305.0,
      "step": 14355
    },
    {
      "entropy": 4.609780311584473,
      "epoch": 4.059937800395816,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003520168192245955,
      "loss": 4.2947,
      "mean_token_accuracy": 0.2746932163834572,
      "num_tokens": 11449261.0,
      "step": 14360
    },
    {
      "entropy": 4.665049266815186,
      "epoch": 4.0613514277636416,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00035192016416041126,
      "loss": 4.2414,
      "mean_token_accuracy": 0.2830650627613068,
      "num_tokens": 11453183.0,
      "step": 14365
    },
    {
      "entropy": 4.635500240325928,
      "epoch": 4.062765055131467,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035182349302038837,
      "loss": 4.259,
      "mean_token_accuracy": 0.2734477683901787,
      "num_tokens": 11457340.0,
      "step": 14370
    },
    {
      "entropy": 4.596258687973022,
      "epoch": 4.064178682499294,
      "grad_norm": 1.671875,
      "learning_rate": 0.00035172680582473047,
      "loss": 4.4615,
      "mean_token_accuracy": 0.26026866734027865,
      "num_tokens": 11461031.0,
      "step": 14375
    },
    {
      "entropy": 4.674797058105469,
      "epoch": 4.065592309867119,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003516301025936447,
      "loss": 4.4239,
      "mean_token_accuracy": 0.2653328388929367,
      "num_tokens": 11465013.0,
      "step": 14380
    },
    {
      "entropy": 4.773935604095459,
      "epoch": 4.067005937234945,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003515333833473413,
      "loss": 4.3812,
      "mean_token_accuracy": 0.25749184787273405,
      "num_tokens": 11469023.0,
      "step": 14385
    },
    {
      "entropy": 4.736489915847779,
      "epoch": 4.06841956460277,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000351436648106034,
      "loss": 4.3983,
      "mean_token_accuracy": 0.25546216070652006,
      "num_tokens": 11473223.0,
      "step": 14390
    },
    {
      "entropy": 4.70972843170166,
      "epoch": 4.069833191970597,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00035133989688994,
      "loss": 4.4019,
      "mean_token_accuracy": 0.2667390078306198,
      "num_tokens": 11477183.0,
      "step": 14395
    },
    {
      "entropy": 4.562456560134888,
      "epoch": 4.071246819338422,
      "grad_norm": 1.65625,
      "learning_rate": 0.00035124312971927953,
      "loss": 4.2349,
      "mean_token_accuracy": 0.27837352454662323,
      "num_tokens": 11481089.0,
      "step": 14400
    },
    {
      "entropy": 4.654204988479615,
      "epoch": 4.072660446706248,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003511463466142765,
      "loss": 4.407,
      "mean_token_accuracy": 0.26190420985221863,
      "num_tokens": 11484861.0,
      "step": 14405
    },
    {
      "entropy": 4.658562755584716,
      "epoch": 4.074074074074074,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003510495475951579,
      "loss": 4.289,
      "mean_token_accuracy": 0.26496831327676773,
      "num_tokens": 11488894.0,
      "step": 14410
    },
    {
      "entropy": 4.696594333648681,
      "epoch": 4.0754877014419,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00035095273268215407,
      "loss": 4.3523,
      "mean_token_accuracy": 0.26219944953918456,
      "num_tokens": 11493032.0,
      "step": 14415
    },
    {
      "entropy": 4.791533279418945,
      "epoch": 4.0769013288097256,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00035085590189549884,
      "loss": 4.4781,
      "mean_token_accuracy": 0.2567607626318932,
      "num_tokens": 11496907.0,
      "step": 14420
    },
    {
      "entropy": 4.7210608959198,
      "epoch": 4.078314956177552,
      "grad_norm": 1.8125,
      "learning_rate": 0.00035075905525542916,
      "loss": 4.4668,
      "mean_token_accuracy": 0.2600831568241119,
      "num_tokens": 11501111.0,
      "step": 14425
    },
    {
      "entropy": 4.600313949584961,
      "epoch": 4.079728583545378,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003506621927821854,
      "loss": 4.2618,
      "mean_token_accuracy": 0.2712284952402115,
      "num_tokens": 11505238.0,
      "step": 14430
    },
    {
      "entropy": 4.6603367805480955,
      "epoch": 4.081142210913203,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035056531449601127,
      "loss": 4.2393,
      "mean_token_accuracy": 0.27284787893295287,
      "num_tokens": 11509199.0,
      "step": 14435
    },
    {
      "entropy": 4.632684278488159,
      "epoch": 4.082555838281029,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003504684204171537,
      "loss": 4.3561,
      "mean_token_accuracy": 0.28497077971696855,
      "num_tokens": 11513166.0,
      "step": 14440
    },
    {
      "entropy": 4.74118070602417,
      "epoch": 4.083969465648855,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003503715105658629,
      "loss": 4.3647,
      "mean_token_accuracy": 0.259723374247551,
      "num_tokens": 11516979.0,
      "step": 14445
    },
    {
      "entropy": 4.650669384002685,
      "epoch": 4.085383093016681,
      "grad_norm": 1.59375,
      "learning_rate": 0.00035027458496239246,
      "loss": 4.4018,
      "mean_token_accuracy": 0.2717857763171196,
      "num_tokens": 11521198.0,
      "step": 14450
    },
    {
      "entropy": 4.581076955795288,
      "epoch": 4.086796720384506,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00035017764362699926,
      "loss": 4.2357,
      "mean_token_accuracy": 0.27302692979574206,
      "num_tokens": 11525148.0,
      "step": 14455
    },
    {
      "entropy": 4.6465085506439205,
      "epoch": 4.088210347752333,
      "grad_norm": 1.671875,
      "learning_rate": 0.00035008068657994346,
      "loss": 4.3345,
      "mean_token_accuracy": 0.2742710098624229,
      "num_tokens": 11528945.0,
      "step": 14460
    },
    {
      "entropy": 4.713344240188599,
      "epoch": 4.089623975120158,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00034998371384148847,
      "loss": 4.3391,
      "mean_token_accuracy": 0.2589407920837402,
      "num_tokens": 11533046.0,
      "step": 14465
    },
    {
      "entropy": 4.598863840103149,
      "epoch": 4.091037602487984,
      "grad_norm": 1.625,
      "learning_rate": 0.0003498867254319009,
      "loss": 4.3485,
      "mean_token_accuracy": 0.27119449228048326,
      "num_tokens": 11537230.0,
      "step": 14470
    },
    {
      "entropy": 4.689518213272095,
      "epoch": 4.09245122985581,
      "grad_norm": 1.609375,
      "learning_rate": 0.00034978972137145094,
      "loss": 4.4528,
      "mean_token_accuracy": 0.2525009259581566,
      "num_tokens": 11541455.0,
      "step": 14475
    },
    {
      "entropy": 4.6969749450683596,
      "epoch": 4.093864857223636,
      "grad_norm": 1.546875,
      "learning_rate": 0.00034969270168041155,
      "loss": 4.4903,
      "mean_token_accuracy": 0.26305725425481796,
      "num_tokens": 11545334.0,
      "step": 14480
    },
    {
      "entropy": 4.637935066223145,
      "epoch": 4.095278484591462,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003495956663790594,
      "loss": 4.3002,
      "mean_token_accuracy": 0.2691005438566208,
      "num_tokens": 11549454.0,
      "step": 14485
    },
    {
      "entropy": 4.721942615509033,
      "epoch": 4.096692111959287,
      "grad_norm": 1.890625,
      "learning_rate": 0.0003494986154876743,
      "loss": 4.431,
      "mean_token_accuracy": 0.24858421981334686,
      "num_tokens": 11553150.0,
      "step": 14490
    },
    {
      "entropy": 4.554998397827148,
      "epoch": 4.098105739327114,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003494015490265393,
      "loss": 4.3416,
      "mean_token_accuracy": 0.2708120822906494,
      "num_tokens": 11557153.0,
      "step": 14495
    },
    {
      "entropy": 4.584628534317017,
      "epoch": 4.099519366694939,
      "grad_norm": 1.625,
      "learning_rate": 0.0003493044670159406,
      "loss": 4.3012,
      "mean_token_accuracy": 0.26992763578891754,
      "num_tokens": 11561048.0,
      "step": 14500
    },
    {
      "entropy": 4.654827404022217,
      "epoch": 4.100932994062765,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00034920736947616785,
      "loss": 4.394,
      "mean_token_accuracy": 0.27304490804672243,
      "num_tokens": 11564962.0,
      "step": 14505
    },
    {
      "entropy": 4.684949684143066,
      "epoch": 4.102346621430591,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003491102564275136,
      "loss": 4.3465,
      "mean_token_accuracy": 0.26240100115537646,
      "num_tokens": 11569183.0,
      "step": 14510
    },
    {
      "entropy": 4.694346523284912,
      "epoch": 4.103760248798417,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0003490131278902741,
      "loss": 4.4724,
      "mean_token_accuracy": 0.25097903460264204,
      "num_tokens": 11573229.0,
      "step": 14515
    },
    {
      "entropy": 4.760846090316773,
      "epoch": 4.105173876166242,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003489159838847487,
      "loss": 4.3886,
      "mean_token_accuracy": 0.2637793645262718,
      "num_tokens": 11577303.0,
      "step": 14520
    },
    {
      "entropy": 4.571243619918823,
      "epoch": 4.106587503534068,
      "grad_norm": 1.46875,
      "learning_rate": 0.00034881882443123956,
      "loss": 4.2902,
      "mean_token_accuracy": 0.27558847665786745,
      "num_tokens": 11581461.0,
      "step": 14525
    },
    {
      "entropy": 4.632758951187133,
      "epoch": 4.1080011309018944,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00034872164955005266,
      "loss": 4.2729,
      "mean_token_accuracy": 0.28061726689338684,
      "num_tokens": 11585335.0,
      "step": 14530
    },
    {
      "entropy": 4.725997686386108,
      "epoch": 4.10941475826972,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003486244592614969,
      "loss": 4.4636,
      "mean_token_accuracy": 0.25143745392560957,
      "num_tokens": 11589142.0,
      "step": 14535
    },
    {
      "entropy": 4.634647607803345,
      "epoch": 4.110828385637546,
      "grad_norm": 1.53125,
      "learning_rate": 0.0003485272535858845,
      "loss": 4.3163,
      "mean_token_accuracy": 0.2693181797862053,
      "num_tokens": 11592972.0,
      "step": 14540
    },
    {
      "entropy": 4.707914161682129,
      "epoch": 4.112242013005372,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00034843003254353064,
      "loss": 4.4899,
      "mean_token_accuracy": 0.25437737703323365,
      "num_tokens": 11596952.0,
      "step": 14545
    },
    {
      "entropy": 4.628875017166138,
      "epoch": 4.113655640373198,
      "grad_norm": 1.53125,
      "learning_rate": 0.00034833279615475413,
      "loss": 4.2839,
      "mean_token_accuracy": 0.2718482717871666,
      "num_tokens": 11601133.0,
      "step": 14550
    },
    {
      "entropy": 4.705890655517578,
      "epoch": 4.115069267741023,
      "grad_norm": 1.9375,
      "learning_rate": 0.00034823554443987666,
      "loss": 4.4312,
      "mean_token_accuracy": 0.2629788801074028,
      "num_tokens": 11605093.0,
      "step": 14555
    },
    {
      "entropy": 4.5935807704925535,
      "epoch": 4.11648289510885,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00034813827741922334,
      "loss": 4.283,
      "mean_token_accuracy": 0.27783487290143966,
      "num_tokens": 11608877.0,
      "step": 14560
    },
    {
      "entropy": 4.5683684825897215,
      "epoch": 4.117896522476675,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003480409951131223,
      "loss": 4.251,
      "mean_token_accuracy": 0.2716818481683731,
      "num_tokens": 11613124.0,
      "step": 14565
    },
    {
      "entropy": 4.570466423034668,
      "epoch": 4.119310149844501,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003479436975419049,
      "loss": 4.2526,
      "mean_token_accuracy": 0.27556669116020205,
      "num_tokens": 11616801.0,
      "step": 14570
    },
    {
      "entropy": 4.629071044921875,
      "epoch": 4.120723777212326,
      "grad_norm": 1.8515625,
      "learning_rate": 0.0003478463847259058,
      "loss": 4.3772,
      "mean_token_accuracy": 0.25197464823722837,
      "num_tokens": 11620677.0,
      "step": 14575
    },
    {
      "entropy": 4.770156049728394,
      "epoch": 4.122137404580153,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003477490566854628,
      "loss": 4.4931,
      "mean_token_accuracy": 0.25051552802324295,
      "num_tokens": 11624659.0,
      "step": 14580
    },
    {
      "entropy": 4.616741132736206,
      "epoch": 4.1235510319479785,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003476517134409168,
      "loss": 4.2735,
      "mean_token_accuracy": 0.2785224437713623,
      "num_tokens": 11628573.0,
      "step": 14585
    },
    {
      "entropy": 4.702906942367553,
      "epoch": 4.124964659315804,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00034755435501261195,
      "loss": 4.3354,
      "mean_token_accuracy": 0.2635082557797432,
      "num_tokens": 11632464.0,
      "step": 14590
    },
    {
      "entropy": 4.676360940933227,
      "epoch": 4.1263782866836305,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003474569814208956,
      "loss": 4.4147,
      "mean_token_accuracy": 0.2599305689334869,
      "num_tokens": 11636616.0,
      "step": 14595
    },
    {
      "entropy": 4.716134071350098,
      "epoch": 4.127791914051456,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003473595926861182,
      "loss": 4.4652,
      "mean_token_accuracy": 0.2616117179393768,
      "num_tokens": 11640966.0,
      "step": 14600
    },
    {
      "entropy": 4.626281499862671,
      "epoch": 4.129205541419282,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003472621888286333,
      "loss": 4.3329,
      "mean_token_accuracy": 0.2637219771742821,
      "num_tokens": 11644847.0,
      "step": 14605
    },
    {
      "entropy": 4.580921030044555,
      "epoch": 4.130619168787108,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00034716476986879785,
      "loss": 4.3233,
      "mean_token_accuracy": 0.2703400805592537,
      "num_tokens": 11648861.0,
      "step": 14610
    },
    {
      "entropy": 4.620072984695435,
      "epoch": 4.132032796154934,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00034706733582697176,
      "loss": 4.2767,
      "mean_token_accuracy": 0.27548509240150454,
      "num_tokens": 11652802.0,
      "step": 14615
    },
    {
      "entropy": 4.809795999526978,
      "epoch": 4.133446423522759,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003469698867235181,
      "loss": 4.4191,
      "mean_token_accuracy": 0.2549046382308006,
      "num_tokens": 11656633.0,
      "step": 14620
    },
    {
      "entropy": 4.758797597885132,
      "epoch": 4.134860050890585,
      "grad_norm": 1.9765625,
      "learning_rate": 0.0003468724225788032,
      "loss": 4.433,
      "mean_token_accuracy": 0.2507268100976944,
      "num_tokens": 11660737.0,
      "step": 14625
    },
    {
      "entropy": 4.6670228958129885,
      "epoch": 4.136273678258411,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00034677494341319636,
      "loss": 4.3312,
      "mean_token_accuracy": 0.27587525695562365,
      "num_tokens": 11664685.0,
      "step": 14630
    },
    {
      "entropy": 4.701639032363891,
      "epoch": 4.137687305626237,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00034667744924707025,
      "loss": 4.4887,
      "mean_token_accuracy": 0.25783083140850066,
      "num_tokens": 11668501.0,
      "step": 14635
    },
    {
      "entropy": 4.61709680557251,
      "epoch": 4.1391009329940625,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003465799401008005,
      "loss": 4.2368,
      "mean_token_accuracy": 0.26954754889011384,
      "num_tokens": 11672343.0,
      "step": 14640
    },
    {
      "entropy": 4.71200532913208,
      "epoch": 4.140514560361889,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003464824159947659,
      "loss": 4.4797,
      "mean_token_accuracy": 0.25032242983579633,
      "num_tokens": 11676436.0,
      "step": 14645
    },
    {
      "entropy": 4.557952880859375,
      "epoch": 4.1419281877297145,
      "grad_norm": 1.703125,
      "learning_rate": 0.00034638487694934835,
      "loss": 4.287,
      "mean_token_accuracy": 0.2795265346765518,
      "num_tokens": 11680223.0,
      "step": 14650
    },
    {
      "entropy": 4.566676187515259,
      "epoch": 4.14334181509754,
      "grad_norm": 1.625,
      "learning_rate": 0.000346287322984933,
      "loss": 4.3183,
      "mean_token_accuracy": 0.2656062483787537,
      "num_tokens": 11684375.0,
      "step": 14655
    },
    {
      "entropy": 4.577923536300659,
      "epoch": 4.1447554424653665,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003461897541219079,
      "loss": 4.2523,
      "mean_token_accuracy": 0.28197304904460907,
      "num_tokens": 11688251.0,
      "step": 14660
    },
    {
      "entropy": 4.649516677856445,
      "epoch": 4.146169069833192,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00034609217038066443,
      "loss": 4.4424,
      "mean_token_accuracy": 0.26903561800718306,
      "num_tokens": 11692435.0,
      "step": 14665
    },
    {
      "entropy": 4.622374391555786,
      "epoch": 4.147582697201018,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000345994571781597,
      "loss": 4.3613,
      "mean_token_accuracy": 0.2595497086644173,
      "num_tokens": 11696408.0,
      "step": 14670
    },
    {
      "entropy": 4.591560983657837,
      "epoch": 4.148996324568843,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003458969583451031,
      "loss": 4.2504,
      "mean_token_accuracy": 0.2853626161813736,
      "num_tokens": 11700457.0,
      "step": 14675
    },
    {
      "entropy": 4.625127077102661,
      "epoch": 4.15040995193667,
      "grad_norm": 1.5625,
      "learning_rate": 0.00034579933009158334,
      "loss": 4.3705,
      "mean_token_accuracy": 0.26398824602365495,
      "num_tokens": 11704836.0,
      "step": 14680
    },
    {
      "entropy": 4.64619550704956,
      "epoch": 4.151823579304495,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00034570168704144133,
      "loss": 4.3836,
      "mean_token_accuracy": 0.25943306982517245,
      "num_tokens": 11708648.0,
      "step": 14685
    },
    {
      "entropy": 4.673339939117431,
      "epoch": 4.153237206672321,
      "grad_norm": 1.984375,
      "learning_rate": 0.00034560402921508397,
      "loss": 4.3826,
      "mean_token_accuracy": 0.2669199839234352,
      "num_tokens": 11712728.0,
      "step": 14690
    },
    {
      "entropy": 4.666187572479248,
      "epoch": 4.154650834040147,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003455063566329211,
      "loss": 4.2826,
      "mean_token_accuracy": 0.26894493103027345,
      "num_tokens": 11716706.0,
      "step": 14695
    },
    {
      "entropy": 4.765179109573364,
      "epoch": 4.156064461407973,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003454086693153657,
      "loss": 4.4455,
      "mean_token_accuracy": 0.24919744431972504,
      "num_tokens": 11720829.0,
      "step": 14700
    },
    {
      "entropy": 4.679324531555176,
      "epoch": 4.1574780887757985,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003453109672828338,
      "loss": 4.4265,
      "mean_token_accuracy": 0.26550883501768113,
      "num_tokens": 11724879.0,
      "step": 14705
    },
    {
      "entropy": 4.682565021514892,
      "epoch": 4.158891716143625,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003452132505557445,
      "loss": 4.3565,
      "mean_token_accuracy": 0.2586074352264404,
      "num_tokens": 11728805.0,
      "step": 14710
    },
    {
      "entropy": 4.747020816802978,
      "epoch": 4.1603053435114505,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003451155191545201,
      "loss": 4.451,
      "mean_token_accuracy": 0.263311031460762,
      "num_tokens": 11732909.0,
      "step": 14715
    },
    {
      "entropy": 4.650549602508545,
      "epoch": 4.161718970879276,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003450177730995857,
      "loss": 4.2902,
      "mean_token_accuracy": 0.2845249205827713,
      "num_tokens": 11736637.0,
      "step": 14720
    },
    {
      "entropy": 4.6716724872589115,
      "epoch": 4.163132598247102,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003449200124113697,
      "loss": 4.321,
      "mean_token_accuracy": 0.27282078713178637,
      "num_tokens": 11740417.0,
      "step": 14725
    },
    {
      "entropy": 4.702000904083252,
      "epoch": 4.164546225614928,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003448222371103035,
      "loss": 4.4155,
      "mean_token_accuracy": 0.2494064286351204,
      "num_tokens": 11744351.0,
      "step": 14730
    },
    {
      "entropy": 4.684924173355102,
      "epoch": 4.165959852982754,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003447244472168215,
      "loss": 4.4159,
      "mean_token_accuracy": 0.25982269644737244,
      "num_tokens": 11748596.0,
      "step": 14735
    },
    {
      "entropy": 4.739539241790771,
      "epoch": 4.167373480350579,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00034462664275136125,
      "loss": 4.4272,
      "mean_token_accuracy": 0.2625486865639687,
      "num_tokens": 11752670.0,
      "step": 14740
    },
    {
      "entropy": 4.6173066139221195,
      "epoch": 4.168787107718406,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003445288237343632,
      "loss": 4.2843,
      "mean_token_accuracy": 0.27080139368772504,
      "num_tokens": 11756721.0,
      "step": 14745
    },
    {
      "entropy": 4.7119852066040036,
      "epoch": 4.170200735086231,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00034443099018627093,
      "loss": 4.3266,
      "mean_token_accuracy": 0.2608434945344925,
      "num_tokens": 11760566.0,
      "step": 14750
    },
    {
      "entropy": 4.585563659667969,
      "epoch": 4.171614362454057,
      "grad_norm": 1.8125,
      "learning_rate": 0.00034433314212753104,
      "loss": 4.2154,
      "mean_token_accuracy": 0.2772764086723328,
      "num_tokens": 11764421.0,
      "step": 14755
    },
    {
      "entropy": 4.667434167861939,
      "epoch": 4.1730279898218825,
      "grad_norm": 1.65625,
      "learning_rate": 0.00034423527957859327,
      "loss": 4.3584,
      "mean_token_accuracy": 0.27055888772010805,
      "num_tokens": 11768286.0,
      "step": 14760
    },
    {
      "entropy": 4.602902269363403,
      "epoch": 4.174441617189709,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00034413740255991005,
      "loss": 4.3392,
      "mean_token_accuracy": 0.2667020305991173,
      "num_tokens": 11772288.0,
      "step": 14765
    },
    {
      "entropy": 4.616268539428711,
      "epoch": 4.1758552445575345,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003440395110919373,
      "loss": 4.4154,
      "mean_token_accuracy": 0.2544607728719711,
      "num_tokens": 11775910.0,
      "step": 14770
    },
    {
      "entropy": 4.688276481628418,
      "epoch": 4.17726887192536,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00034394160519513376,
      "loss": 4.4579,
      "mean_token_accuracy": 0.2560638442635536,
      "num_tokens": 11780053.0,
      "step": 14775
    },
    {
      "entropy": 4.6634800910949705,
      "epoch": 4.178682499293187,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00034384368488996097,
      "loss": 4.3483,
      "mean_token_accuracy": 0.27410229444503786,
      "num_tokens": 11784061.0,
      "step": 14780
    },
    {
      "entropy": 4.633969926834107,
      "epoch": 4.180096126661012,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003437457501968837,
      "loss": 4.3048,
      "mean_token_accuracy": 0.2624826982617378,
      "num_tokens": 11787947.0,
      "step": 14785
    },
    {
      "entropy": 4.671413707733154,
      "epoch": 4.181509754028838,
      "grad_norm": 1.65625,
      "learning_rate": 0.00034364780113636985,
      "loss": 4.3959,
      "mean_token_accuracy": 0.26080692559480667,
      "num_tokens": 11791867.0,
      "step": 14790
    },
    {
      "entropy": 4.638953399658203,
      "epoch": 4.182923381396664,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00034354983772888996,
      "loss": 4.336,
      "mean_token_accuracy": 0.2660954728722572,
      "num_tokens": 11795602.0,
      "step": 14795
    },
    {
      "entropy": 4.586797094345092,
      "epoch": 4.18433700876449,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00034345185999491793,
      "loss": 4.3106,
      "mean_token_accuracy": 0.27309995591640474,
      "num_tokens": 11799207.0,
      "step": 14800
    },
    {
      "entropy": 4.669696235656739,
      "epoch": 4.185750636132315,
      "grad_norm": 1.625,
      "learning_rate": 0.0003433538679549305,
      "loss": 4.4465,
      "mean_token_accuracy": 0.2664122015237808,
      "num_tokens": 11803106.0,
      "step": 14805
    },
    {
      "entropy": 4.7154065608978275,
      "epoch": 4.187164263500142,
      "grad_norm": 1.6875,
      "learning_rate": 0.00034325586162940736,
      "loss": 4.431,
      "mean_token_accuracy": 0.2594901278614998,
      "num_tokens": 11807141.0,
      "step": 14810
    },
    {
      "entropy": 4.664580392837524,
      "epoch": 4.188577890867967,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00034315784103883126,
      "loss": 4.352,
      "mean_token_accuracy": 0.27187355011701586,
      "num_tokens": 11811232.0,
      "step": 14815
    },
    {
      "entropy": 4.6498565673828125,
      "epoch": 4.189991518235793,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003430598062036878,
      "loss": 4.443,
      "mean_token_accuracy": 0.25996207147836686,
      "num_tokens": 11815134.0,
      "step": 14820
    },
    {
      "entropy": 4.551423358917236,
      "epoch": 4.1914051456036185,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00034296175714446564,
      "loss": 4.2412,
      "mean_token_accuracy": 0.2827671840786934,
      "num_tokens": 11819271.0,
      "step": 14825
    },
    {
      "entropy": 4.67864727973938,
      "epoch": 4.192818772971445,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003428636938816565,
      "loss": 4.3953,
      "mean_token_accuracy": 0.2600191831588745,
      "num_tokens": 11823094.0,
      "step": 14830
    },
    {
      "entropy": 4.571554708480835,
      "epoch": 4.194232400339271,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003427656164357551,
      "loss": 4.3008,
      "mean_token_accuracy": 0.269410240650177,
      "num_tokens": 11827021.0,
      "step": 14835
    },
    {
      "entropy": 4.669420433044434,
      "epoch": 4.195646027707096,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003426675248272589,
      "loss": 4.3391,
      "mean_token_accuracy": 0.2609356105327606,
      "num_tokens": 11831286.0,
      "step": 14840
    },
    {
      "entropy": 4.610635566711426,
      "epoch": 4.197059655074923,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003425694190766684,
      "loss": 4.2442,
      "mean_token_accuracy": 0.27137421518564225,
      "num_tokens": 11835036.0,
      "step": 14845
    },
    {
      "entropy": 4.649283075332642,
      "epoch": 4.198473282442748,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00034247129920448715,
      "loss": 4.4142,
      "mean_token_accuracy": 0.2549251139163971,
      "num_tokens": 11838881.0,
      "step": 14850
    },
    {
      "entropy": 4.678107166290284,
      "epoch": 4.199886909810574,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00034237316523122165,
      "loss": 4.3729,
      "mean_token_accuracy": 0.26425751298666,
      "num_tokens": 11843144.0,
      "step": 14855
    },
    {
      "entropy": 4.577150297164917,
      "epoch": 4.201300537178399,
      "grad_norm": 1.78125,
      "learning_rate": 0.00034227501717738117,
      "loss": 4.2821,
      "mean_token_accuracy": 0.2774759724736214,
      "num_tokens": 11847024.0,
      "step": 14860
    },
    {
      "entropy": 4.72305474281311,
      "epoch": 4.202714164546226,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003421768550634782,
      "loss": 4.4985,
      "mean_token_accuracy": 0.25000310838222506,
      "num_tokens": 11850874.0,
      "step": 14865
    },
    {
      "entropy": 4.659708786010742,
      "epoch": 4.204127791914051,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003420786789100279,
      "loss": 4.3449,
      "mean_token_accuracy": 0.2692570298910141,
      "num_tokens": 11854824.0,
      "step": 14870
    },
    {
      "entropy": 4.589246034622192,
      "epoch": 4.205541419281877,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003419804887375485,
      "loss": 4.2742,
      "mean_token_accuracy": 0.2830927848815918,
      "num_tokens": 11858780.0,
      "step": 14875
    },
    {
      "entropy": 4.6029212951660154,
      "epoch": 4.206955046649703,
      "grad_norm": 1.671875,
      "learning_rate": 0.00034188228456656113,
      "loss": 4.3754,
      "mean_token_accuracy": 0.2723086550831795,
      "num_tokens": 11862694.0,
      "step": 14880
    },
    {
      "entropy": 4.539843368530273,
      "epoch": 4.208368674017529,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003417840664175899,
      "loss": 4.3044,
      "mean_token_accuracy": 0.2756637200713158,
      "num_tokens": 11866375.0,
      "step": 14885
    },
    {
      "entropy": 4.62773962020874,
      "epoch": 4.209782301385355,
      "grad_norm": 1.65625,
      "learning_rate": 0.00034168583431116176,
      "loss": 4.4367,
      "mean_token_accuracy": 0.2622852593660355,
      "num_tokens": 11870035.0,
      "step": 14890
    },
    {
      "entropy": 4.634240198135376,
      "epoch": 4.211195928753181,
      "grad_norm": 1.625,
      "learning_rate": 0.0003415875882678066,
      "loss": 4.2977,
      "mean_token_accuracy": 0.2672229766845703,
      "num_tokens": 11874198.0,
      "step": 14895
    },
    {
      "entropy": 4.637016439437867,
      "epoch": 4.212609556121007,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003414893283080573,
      "loss": 4.3542,
      "mean_token_accuracy": 0.2691663518548012,
      "num_tokens": 11878398.0,
      "step": 14900
    },
    {
      "entropy": 4.596614980697632,
      "epoch": 4.214023183488832,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003413910544524496,
      "loss": 4.2897,
      "mean_token_accuracy": 0.2846513748168945,
      "num_tokens": 11882286.0,
      "step": 14905
    },
    {
      "entropy": 4.606705093383789,
      "epoch": 4.215436810856658,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003412927667215219,
      "loss": 4.4036,
      "mean_token_accuracy": 0.25982747972011566,
      "num_tokens": 11886227.0,
      "step": 14910
    },
    {
      "entropy": 4.6260082721710205,
      "epoch": 4.216850438224484,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00034119446513581605,
      "loss": 4.3684,
      "mean_token_accuracy": 0.2615576237440109,
      "num_tokens": 11890524.0,
      "step": 14915
    },
    {
      "entropy": 4.690567684173584,
      "epoch": 4.21826406559231,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00034109614971587613,
      "loss": 4.4487,
      "mean_token_accuracy": 0.25919603556394577,
      "num_tokens": 11894220.0,
      "step": 14920
    },
    {
      "entropy": 4.675605726242066,
      "epoch": 4.219677692960135,
      "grad_norm": 1.625,
      "learning_rate": 0.00034099782048224985,
      "loss": 4.308,
      "mean_token_accuracy": 0.2704439863562584,
      "num_tokens": 11898142.0,
      "step": 14925
    },
    {
      "entropy": 4.722121095657348,
      "epoch": 4.221091320327962,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003408994774554871,
      "loss": 4.4061,
      "mean_token_accuracy": 0.27277589589357376,
      "num_tokens": 11902303.0,
      "step": 14930
    },
    {
      "entropy": 4.567144823074341,
      "epoch": 4.222504947695787,
      "grad_norm": 1.765625,
      "learning_rate": 0.00034080112065614117,
      "loss": 4.242,
      "mean_token_accuracy": 0.2795108571648598,
      "num_tokens": 11906452.0,
      "step": 14935
    },
    {
      "entropy": 4.518485689163208,
      "epoch": 4.223918575063613,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00034070275010476776,
      "loss": 4.2237,
      "mean_token_accuracy": 0.27962863594293597,
      "num_tokens": 11910257.0,
      "step": 14940
    },
    {
      "entropy": 4.66899266242981,
      "epoch": 4.2253322024314395,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003406043658219261,
      "loss": 4.3956,
      "mean_token_accuracy": 0.26157755553722384,
      "num_tokens": 11914339.0,
      "step": 14945
    },
    {
      "entropy": 4.690749168395996,
      "epoch": 4.226745829799265,
      "grad_norm": 1.90625,
      "learning_rate": 0.0003405059678281774,
      "loss": 4.3451,
      "mean_token_accuracy": 0.2618440255522728,
      "num_tokens": 11918253.0,
      "step": 14950
    },
    {
      "entropy": 4.749414300918579,
      "epoch": 4.228159457167091,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003404075561440867,
      "loss": 4.3777,
      "mean_token_accuracy": 0.2642712816596031,
      "num_tokens": 11922017.0,
      "step": 14955
    },
    {
      "entropy": 4.604523658752441,
      "epoch": 4.229573084534916,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003403091307902211,
      "loss": 4.3253,
      "mean_token_accuracy": 0.2720665320754051,
      "num_tokens": 11926047.0,
      "step": 14960
    },
    {
      "entropy": 4.650488805770874,
      "epoch": 4.230986711902743,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00034021069178715113,
      "loss": 4.3411,
      "mean_token_accuracy": 0.2690587744116783,
      "num_tokens": 11930059.0,
      "step": 14965
    },
    {
      "entropy": 4.714971590042114,
      "epoch": 4.232400339270568,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003401122391554498,
      "loss": 4.4138,
      "mean_token_accuracy": 0.2587359055876732,
      "num_tokens": 11934087.0,
      "step": 14970
    },
    {
      "entropy": 4.669944763183594,
      "epoch": 4.233813966638394,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00034001377291569316,
      "loss": 4.4004,
      "mean_token_accuracy": 0.2655493527650833,
      "num_tokens": 11937763.0,
      "step": 14975
    },
    {
      "entropy": 4.684559392929077,
      "epoch": 4.23522759400622,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003399152930884599,
      "loss": 4.4022,
      "mean_token_accuracy": 0.2686028331518173,
      "num_tokens": 11941644.0,
      "step": 14980
    },
    {
      "entropy": 4.647962951660157,
      "epoch": 4.236641221374046,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003398167996943319,
      "loss": 4.3357,
      "mean_token_accuracy": 0.27624250650405885,
      "num_tokens": 11945946.0,
      "step": 14985
    },
    {
      "entropy": 4.655934381484985,
      "epoch": 4.238054848741871,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0003397182927538936,
      "loss": 4.3405,
      "mean_token_accuracy": 0.2609539195895195,
      "num_tokens": 11949847.0,
      "step": 14990
    },
    {
      "entropy": 4.61395959854126,
      "epoch": 4.239468476109698,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00033961977228773234,
      "loss": 4.2604,
      "mean_token_accuracy": 0.2757533386349678,
      "num_tokens": 11954092.0,
      "step": 14995
    },
    {
      "entropy": 4.614962244033814,
      "epoch": 4.2408821034775235,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00033952123831643816,
      "loss": 4.3321,
      "mean_token_accuracy": 0.2641711339354515,
      "num_tokens": 11957764.0,
      "step": 15000
    },
    {
      "epoch": 4.2408821034775235,
      "eval_entropy": 4.4800035130631235,
      "eval_loss": 5.023704528808594,
      "eval_mean_token_accuracy": 0.24572687034780974,
      "eval_num_tokens": 11957764.0,
      "eval_runtime": 6.8728,
      "eval_samples_per_second": 1641.839,
      "eval_steps_per_second": 205.303,
      "step": 15000
    },
    {
      "entropy": 4.638383960723877,
      "epoch": 4.242295730845349,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003394226908606042,
      "loss": 4.4117,
      "mean_token_accuracy": 0.2665575236082077,
      "num_tokens": 11961899.0,
      "step": 15005
    },
    {
      "entropy": 4.687260341644287,
      "epoch": 4.243709358213175,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003393241299408262,
      "loss": 4.3123,
      "mean_token_accuracy": 0.27236763685941695,
      "num_tokens": 11965860.0,
      "step": 15010
    },
    {
      "entropy": 4.653101825714112,
      "epoch": 4.245122985581001,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003392255555777029,
      "loss": 4.3633,
      "mean_token_accuracy": 0.27132226526737213,
      "num_tokens": 11969910.0,
      "step": 15015
    },
    {
      "entropy": 4.607816076278686,
      "epoch": 4.246536612948827,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003391269677918356,
      "loss": 4.3018,
      "mean_token_accuracy": 0.2591134801506996,
      "num_tokens": 11973827.0,
      "step": 15020
    },
    {
      "entropy": 4.6577582359313965,
      "epoch": 4.247950240316652,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003390283666038286,
      "loss": 4.3733,
      "mean_token_accuracy": 0.266715045273304,
      "num_tokens": 11978015.0,
      "step": 15025
    },
    {
      "entropy": 4.780586910247803,
      "epoch": 4.249363867684479,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003389297520342888,
      "loss": 4.4628,
      "mean_token_accuracy": 0.25646388828754424,
      "num_tokens": 11981929.0,
      "step": 15030
    },
    {
      "entropy": 4.692286539077759,
      "epoch": 4.250777495052304,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00033883112410382633,
      "loss": 4.3394,
      "mean_token_accuracy": 0.26872755885124205,
      "num_tokens": 11985637.0,
      "step": 15035
    },
    {
      "entropy": 4.679859447479248,
      "epoch": 4.25219112242013,
      "grad_norm": 1.671875,
      "learning_rate": 0.00033873248283305355,
      "loss": 4.5039,
      "mean_token_accuracy": 0.25370749086141586,
      "num_tokens": 11989782.0,
      "step": 15040
    },
    {
      "entropy": 4.659099531173706,
      "epoch": 4.253604749787955,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0003386338282425859,
      "loss": 4.435,
      "mean_token_accuracy": 0.2592045322060585,
      "num_tokens": 11993823.0,
      "step": 15045
    },
    {
      "entropy": 4.639576005935669,
      "epoch": 4.255018377155782,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003385351603530418,
      "loss": 4.281,
      "mean_token_accuracy": 0.2760147303342819,
      "num_tokens": 11997158.0,
      "step": 15050
    },
    {
      "entropy": 4.742485761642456,
      "epoch": 4.2564320045236075,
      "grad_norm": 1.8046875,
      "learning_rate": 0.000338436479185042,
      "loss": 4.3398,
      "mean_token_accuracy": 0.26690368801355363,
      "num_tokens": 12001059.0,
      "step": 15055
    },
    {
      "entropy": 4.729960536956787,
      "epoch": 4.257845631891433,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003383377847592103,
      "loss": 4.5423,
      "mean_token_accuracy": 0.2518656015396118,
      "num_tokens": 12004831.0,
      "step": 15060
    },
    {
      "entropy": 4.68672251701355,
      "epoch": 4.2592592592592595,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003382390770961733,
      "loss": 4.3527,
      "mean_token_accuracy": 0.2753472402691841,
      "num_tokens": 12008797.0,
      "step": 15065
    },
    {
      "entropy": 4.7781823635101315,
      "epoch": 4.260672886627085,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003381403562165603,
      "loss": 4.4799,
      "mean_token_accuracy": 0.252541309595108,
      "num_tokens": 12012663.0,
      "step": 15070
    },
    {
      "entropy": 4.673221731185913,
      "epoch": 4.262086513994911,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00033804162214100313,
      "loss": 4.3881,
      "mean_token_accuracy": 0.26497416347265246,
      "num_tokens": 12016308.0,
      "step": 15075
    },
    {
      "entropy": 4.625990867614746,
      "epoch": 4.263500141362737,
      "grad_norm": 1.75,
      "learning_rate": 0.00033794287489013694,
      "loss": 4.4203,
      "mean_token_accuracy": 0.2637854918837547,
      "num_tokens": 12020308.0,
      "step": 15080
    },
    {
      "entropy": 4.587663316726685,
      "epoch": 4.264913768730563,
      "grad_norm": 1.71875,
      "learning_rate": 0.00033784411448459906,
      "loss": 4.2864,
      "mean_token_accuracy": 0.2750094816088676,
      "num_tokens": 12024208.0,
      "step": 15085
    },
    {
      "entropy": 4.669900941848755,
      "epoch": 4.266327396098388,
      "grad_norm": 1.671875,
      "learning_rate": 0.00033774534094502983,
      "loss": 4.3736,
      "mean_token_accuracy": 0.25336671322584153,
      "num_tokens": 12028161.0,
      "step": 15090
    },
    {
      "entropy": 4.574107503890991,
      "epoch": 4.267741023466215,
      "grad_norm": 1.703125,
      "learning_rate": 0.00033764655429207235,
      "loss": 4.4021,
      "mean_token_accuracy": 0.2573527187108994,
      "num_tokens": 12032181.0,
      "step": 15095
    },
    {
      "entropy": 4.732128286361695,
      "epoch": 4.26915465083404,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003375477545463725,
      "loss": 4.4157,
      "mean_token_accuracy": 0.2524444207549095,
      "num_tokens": 12035966.0,
      "step": 15100
    },
    {
      "entropy": 4.663439416885376,
      "epoch": 4.270568278201866,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00033744894172857865,
      "loss": 4.3072,
      "mean_token_accuracy": 0.2756292596459389,
      "num_tokens": 12039980.0,
      "step": 15105
    },
    {
      "entropy": 4.521182060241699,
      "epoch": 4.2719819055696915,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003373501158593422,
      "loss": 4.2227,
      "mean_token_accuracy": 0.27444323152303696,
      "num_tokens": 12043686.0,
      "step": 15110
    },
    {
      "entropy": 4.5887675285339355,
      "epoch": 4.273395532937518,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003372512769593171,
      "loss": 4.3039,
      "mean_token_accuracy": 0.2775538727641106,
      "num_tokens": 12047532.0,
      "step": 15115
    },
    {
      "entropy": 4.578398847579956,
      "epoch": 4.2748091603053435,
      "grad_norm": 1.6875,
      "learning_rate": 0.00033715242504916005,
      "loss": 4.3374,
      "mean_token_accuracy": 0.26252163797616956,
      "num_tokens": 12051449.0,
      "step": 15120
    },
    {
      "entropy": 4.662602043151855,
      "epoch": 4.276222787673169,
      "grad_norm": 1.75,
      "learning_rate": 0.00033705356014953047,
      "loss": 4.4024,
      "mean_token_accuracy": 0.25512285679578783,
      "num_tokens": 12055272.0,
      "step": 15125
    },
    {
      "entropy": 4.5256514072418215,
      "epoch": 4.277636415040996,
      "grad_norm": 1.578125,
      "learning_rate": 0.00033695468228109055,
      "loss": 4.307,
      "mean_token_accuracy": 0.2738457679748535,
      "num_tokens": 12059304.0,
      "step": 15130
    },
    {
      "entropy": 4.663923883438111,
      "epoch": 4.279050042408821,
      "grad_norm": 1.59375,
      "learning_rate": 0.00033685579146450515,
      "loss": 4.4022,
      "mean_token_accuracy": 0.2599486142396927,
      "num_tokens": 12063488.0,
      "step": 15135
    },
    {
      "entropy": 4.664436340332031,
      "epoch": 4.280463669776647,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003367568877204419,
      "loss": 4.3571,
      "mean_token_accuracy": 0.26240302473306654,
      "num_tokens": 12067347.0,
      "step": 15140
    },
    {
      "entropy": 4.718522596359253,
      "epoch": 4.281877297144472,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000336657971069571,
      "loss": 4.4776,
      "mean_token_accuracy": 0.2611654192209244,
      "num_tokens": 12071325.0,
      "step": 15145
    },
    {
      "entropy": 4.678318071365356,
      "epoch": 4.283290924512299,
      "grad_norm": 1.65625,
      "learning_rate": 0.00033655904153256537,
      "loss": 4.3298,
      "mean_token_accuracy": 0.272107195854187,
      "num_tokens": 12075194.0,
      "step": 15150
    },
    {
      "entropy": 4.709887456893921,
      "epoch": 4.284704551880124,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003364600991301008,
      "loss": 4.4883,
      "mean_token_accuracy": 0.26358645856380464,
      "num_tokens": 12079134.0,
      "step": 15155
    },
    {
      "entropy": 4.6766520023345945,
      "epoch": 4.28611817924795,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00033636114388285553,
      "loss": 4.4094,
      "mean_token_accuracy": 0.26485465914011,
      "num_tokens": 12083130.0,
      "step": 15160
    },
    {
      "entropy": 4.6963366031646725,
      "epoch": 4.287531806615776,
      "grad_norm": 1.75,
      "learning_rate": 0.00033626217581151063,
      "loss": 4.462,
      "mean_token_accuracy": 0.2643813297152519,
      "num_tokens": 12087475.0,
      "step": 15165
    },
    {
      "entropy": 4.676368379592896,
      "epoch": 4.288945433983602,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003361631949367498,
      "loss": 4.3346,
      "mean_token_accuracy": 0.27066474556922915,
      "num_tokens": 12091345.0,
      "step": 15170
    },
    {
      "entropy": 4.593088436126709,
      "epoch": 4.2903590613514275,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00033606420127925944,
      "loss": 4.3462,
      "mean_token_accuracy": 0.27168858200311663,
      "num_tokens": 12095424.0,
      "step": 15175
    },
    {
      "entropy": 4.721170139312744,
      "epoch": 4.291772688719254,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003359651948597287,
      "loss": 4.4368,
      "mean_token_accuracy": 0.25656689554452894,
      "num_tokens": 12099504.0,
      "step": 15180
    },
    {
      "entropy": 4.6603196144104,
      "epoch": 4.29318631608708,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00033586617569884917,
      "loss": 4.3382,
      "mean_token_accuracy": 0.27614943832159045,
      "num_tokens": 12103245.0,
      "step": 15185
    },
    {
      "entropy": 4.70788688659668,
      "epoch": 4.294599943454905,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00033576714381731537,
      "loss": 4.3972,
      "mean_token_accuracy": 0.25574812293052673,
      "num_tokens": 12107300.0,
      "step": 15190
    },
    {
      "entropy": 4.6791582107543945,
      "epoch": 4.296013570822731,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003356680992358242,
      "loss": 4.4046,
      "mean_token_accuracy": 0.25267298966646196,
      "num_tokens": 12111424.0,
      "step": 15195
    },
    {
      "entropy": 4.57160701751709,
      "epoch": 4.297427198190557,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00033556904197507546,
      "loss": 4.3613,
      "mean_token_accuracy": 0.251092529296875,
      "num_tokens": 12115074.0,
      "step": 15200
    },
    {
      "entropy": 4.687270641326904,
      "epoch": 4.298840825558383,
      "grad_norm": 1.75,
      "learning_rate": 0.00033546997205577156,
      "loss": 4.4247,
      "mean_token_accuracy": 0.2554011046886444,
      "num_tokens": 12119185.0,
      "step": 15205
    },
    {
      "entropy": 4.661234951019287,
      "epoch": 4.300254452926208,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00033537088949861743,
      "loss": 4.2685,
      "mean_token_accuracy": 0.28098344057798386,
      "num_tokens": 12123396.0,
      "step": 15210
    },
    {
      "entropy": 4.5543111801147464,
      "epoch": 4.301668080294035,
      "grad_norm": 1.640625,
      "learning_rate": 0.00033527179432432067,
      "loss": 4.2916,
      "mean_token_accuracy": 0.27609900534152987,
      "num_tokens": 12127441.0,
      "step": 15215
    },
    {
      "entropy": 4.586578798294068,
      "epoch": 4.30308170766186,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003351726865535916,
      "loss": 4.2567,
      "mean_token_accuracy": 0.28159120082855227,
      "num_tokens": 12131469.0,
      "step": 15220
    },
    {
      "entropy": 4.679605484008789,
      "epoch": 4.304495335029686,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00033507356620714315,
      "loss": 4.3762,
      "mean_token_accuracy": 0.2576820060610771,
      "num_tokens": 12135417.0,
      "step": 15225
    },
    {
      "entropy": 4.746995687484741,
      "epoch": 4.305908962397512,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00033497443330569084,
      "loss": 4.4225,
      "mean_token_accuracy": 0.25963999778032304,
      "num_tokens": 12139314.0,
      "step": 15230
    },
    {
      "entropy": 4.736866426467896,
      "epoch": 4.307322589765338,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003348752878699529,
      "loss": 4.3483,
      "mean_token_accuracy": 0.24795842915773392,
      "num_tokens": 12143322.0,
      "step": 15235
    },
    {
      "entropy": 4.6541821479797365,
      "epoch": 4.308736217133164,
      "grad_norm": 1.609375,
      "learning_rate": 0.00033477612992065,
      "loss": 4.373,
      "mean_token_accuracy": 0.25923986434936525,
      "num_tokens": 12147429.0,
      "step": 15240
    },
    {
      "entropy": 4.647263479232788,
      "epoch": 4.310149844500989,
      "grad_norm": 1.578125,
      "learning_rate": 0.00033467695947850555,
      "loss": 4.3802,
      "mean_token_accuracy": 0.25989028364419936,
      "num_tokens": 12151732.0,
      "step": 15245
    },
    {
      "entropy": 4.707294940948486,
      "epoch": 4.311563471868816,
      "grad_norm": 1.65625,
      "learning_rate": 0.00033457777656424565,
      "loss": 4.4496,
      "mean_token_accuracy": 0.2529548674821854,
      "num_tokens": 12155850.0,
      "step": 15250
    },
    {
      "entropy": 4.67233452796936,
      "epoch": 4.312977099236641,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003344785811985989,
      "loss": 4.3813,
      "mean_token_accuracy": 0.2594792008399963,
      "num_tokens": 12160170.0,
      "step": 15255
    },
    {
      "entropy": 4.578860950469971,
      "epoch": 4.314390726604467,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0003343793734022964,
      "loss": 4.2585,
      "mean_token_accuracy": 0.27998505234718324,
      "num_tokens": 12164232.0,
      "step": 15260
    },
    {
      "entropy": 4.636598443984985,
      "epoch": 4.315804353972293,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00033428015319607215,
      "loss": 4.4121,
      "mean_token_accuracy": 0.2626410096883774,
      "num_tokens": 12168198.0,
      "step": 15265
    },
    {
      "entropy": 4.824108552932739,
      "epoch": 4.317217981340119,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003341809206006624,
      "loss": 4.5282,
      "mean_token_accuracy": 0.25110555440187454,
      "num_tokens": 12172040.0,
      "step": 15270
    },
    {
      "entropy": 4.586532068252564,
      "epoch": 4.318631608707944,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003340816756368062,
      "loss": 4.1984,
      "mean_token_accuracy": 0.27760314047336576,
      "num_tokens": 12176013.0,
      "step": 15275
    },
    {
      "entropy": 4.642512798309326,
      "epoch": 4.320045236075771,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003339824183252452,
      "loss": 4.41,
      "mean_token_accuracy": 0.25600753277540206,
      "num_tokens": 12180499.0,
      "step": 15280
    },
    {
      "entropy": 4.636323499679565,
      "epoch": 4.321458863443596,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003338831486867234,
      "loss": 4.374,
      "mean_token_accuracy": 0.2777717962861061,
      "num_tokens": 12184340.0,
      "step": 15285
    },
    {
      "entropy": 4.665165281295776,
      "epoch": 4.322872490811422,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00033378386674198774,
      "loss": 4.3289,
      "mean_token_accuracy": 0.26004510670900344,
      "num_tokens": 12188293.0,
      "step": 15290
    },
    {
      "entropy": 4.659646034240723,
      "epoch": 4.324286118179248,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003336845725117875,
      "loss": 4.3271,
      "mean_token_accuracy": 0.2812427833676338,
      "num_tokens": 12192666.0,
      "step": 15295
    },
    {
      "entropy": 4.695091009140015,
      "epoch": 4.325699745547074,
      "grad_norm": 1.78125,
      "learning_rate": 0.00033358526601687434,
      "loss": 4.4334,
      "mean_token_accuracy": 0.25873361229896547,
      "num_tokens": 12196728.0,
      "step": 15300
    },
    {
      "entropy": 4.658290863037109,
      "epoch": 4.3271133729149,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003334859472780029,
      "loss": 4.3538,
      "mean_token_accuracy": 0.2678295478224754,
      "num_tokens": 12200659.0,
      "step": 15305
    },
    {
      "entropy": 4.643149709701538,
      "epoch": 4.328527000282725,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0003333866163159302,
      "loss": 4.3753,
      "mean_token_accuracy": 0.2579259514808655,
      "num_tokens": 12205023.0,
      "step": 15310
    },
    {
      "entropy": 4.653281736373901,
      "epoch": 4.329940627650552,
      "grad_norm": 1.625,
      "learning_rate": 0.0003332872731514157,
      "loss": 4.4358,
      "mean_token_accuracy": 0.256537427008152,
      "num_tokens": 12208689.0,
      "step": 15315
    },
    {
      "entropy": 4.594387483596802,
      "epoch": 4.331354255018377,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00033318791780522156,
      "loss": 4.3471,
      "mean_token_accuracy": 0.26940288841724397,
      "num_tokens": 12212639.0,
      "step": 15320
    },
    {
      "entropy": 4.625309801101684,
      "epoch": 4.332767882386203,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003330885502981124,
      "loss": 4.3222,
      "mean_token_accuracy": 0.2791112020611763,
      "num_tokens": 12216456.0,
      "step": 15325
    },
    {
      "entropy": 4.693771123886108,
      "epoch": 4.334181509754028,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00033298917065085554,
      "loss": 4.4826,
      "mean_token_accuracy": 0.25337713956832886,
      "num_tokens": 12220560.0,
      "step": 15330
    },
    {
      "entropy": 4.78141222000122,
      "epoch": 4.335595137121855,
      "grad_norm": 1.59375,
      "learning_rate": 0.00033288977888422047,
      "loss": 4.4934,
      "mean_token_accuracy": 0.24958913326263427,
      "num_tokens": 12224735.0,
      "step": 15335
    },
    {
      "entropy": 4.682506704330445,
      "epoch": 4.33700876448968,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003327903750189797,
      "loss": 4.2983,
      "mean_token_accuracy": 0.27259191274642947,
      "num_tokens": 12228719.0,
      "step": 15340
    },
    {
      "entropy": 4.663098621368408,
      "epoch": 4.338422391857506,
      "grad_norm": 1.625,
      "learning_rate": 0.00033269095907590784,
      "loss": 4.4345,
      "mean_token_accuracy": 0.2484143480658531,
      "num_tokens": 12232807.0,
      "step": 15345
    },
    {
      "entropy": 4.686139488220215,
      "epoch": 4.3398360192253325,
      "grad_norm": 1.53125,
      "learning_rate": 0.00033259153107578234,
      "loss": 4.3093,
      "mean_token_accuracy": 0.2753065228462219,
      "num_tokens": 12236700.0,
      "step": 15350
    },
    {
      "entropy": 4.563955497741699,
      "epoch": 4.341249646593158,
      "grad_norm": 1.765625,
      "learning_rate": 0.000332492091039383,
      "loss": 4.3117,
      "mean_token_accuracy": 0.27547753006219866,
      "num_tokens": 12240634.0,
      "step": 15355
    },
    {
      "entropy": 4.694627809524536,
      "epoch": 4.342663273960984,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00033239263898749214,
      "loss": 4.3877,
      "mean_token_accuracy": 0.2700698062777519,
      "num_tokens": 12244604.0,
      "step": 15360
    },
    {
      "entropy": 4.614536380767822,
      "epoch": 4.34407690132881,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003322931749408945,
      "loss": 4.4232,
      "mean_token_accuracy": 0.26537179946899414,
      "num_tokens": 12248259.0,
      "step": 15365
    },
    {
      "entropy": 4.695988321304322,
      "epoch": 4.345490528696636,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003321936989203778,
      "loss": 4.4327,
      "mean_token_accuracy": 0.2585691958665848,
      "num_tokens": 12252591.0,
      "step": 15370
    },
    {
      "entropy": 4.6022051811218265,
      "epoch": 4.346904156064461,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003320942109467316,
      "loss": 4.3969,
      "mean_token_accuracy": 0.2657610222697258,
      "num_tokens": 12256479.0,
      "step": 15375
    },
    {
      "entropy": 4.6828735828399655,
      "epoch": 4.348317783432288,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00033199471104074826,
      "loss": 4.3795,
      "mean_token_accuracy": 0.2677401348948479,
      "num_tokens": 12260302.0,
      "step": 15380
    },
    {
      "entropy": 4.681718254089356,
      "epoch": 4.349731410800113,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003318951992232229,
      "loss": 4.4256,
      "mean_token_accuracy": 0.25841780453920365,
      "num_tokens": 12263992.0,
      "step": 15385
    },
    {
      "entropy": 4.686974430084229,
      "epoch": 4.351145038167939,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00033179567551495275,
      "loss": 4.4315,
      "mean_token_accuracy": 0.2506022363901138,
      "num_tokens": 12268096.0,
      "step": 15390
    },
    {
      "entropy": 4.646786737442016,
      "epoch": 4.352558665535764,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0003316961399367376,
      "loss": 4.4854,
      "mean_token_accuracy": 0.2704179257154465,
      "num_tokens": 12272098.0,
      "step": 15395
    },
    {
      "entropy": 4.709814357757568,
      "epoch": 4.353972292903591,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00033159659250937975,
      "loss": 4.3758,
      "mean_token_accuracy": 0.25384555011987686,
      "num_tokens": 12275815.0,
      "step": 15400
    },
    {
      "entropy": 4.658036899566651,
      "epoch": 4.3553859202714165,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003314970332536841,
      "loss": 4.3177,
      "mean_token_accuracy": 0.27027151435613633,
      "num_tokens": 12279976.0,
      "step": 15405
    },
    {
      "entropy": 4.588439655303955,
      "epoch": 4.356799547639242,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003313974621904578,
      "loss": 4.26,
      "mean_token_accuracy": 0.2770970478653908,
      "num_tokens": 12283714.0,
      "step": 15410
    },
    {
      "entropy": 4.608199644088745,
      "epoch": 4.3582131750070685,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0003312978793405107,
      "loss": 4.3356,
      "mean_token_accuracy": 0.27484062016010286,
      "num_tokens": 12287448.0,
      "step": 15415
    },
    {
      "entropy": 4.599783039093017,
      "epoch": 4.359626802374894,
      "grad_norm": 1.8203125,
      "learning_rate": 0.000331198284724655,
      "loss": 4.3656,
      "mean_token_accuracy": 0.2630252495408058,
      "num_tokens": 12291087.0,
      "step": 15420
    },
    {
      "entropy": 4.618776321411133,
      "epoch": 4.36104042974272,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003310986783637052,
      "loss": 4.3443,
      "mean_token_accuracy": 0.26947192549705506,
      "num_tokens": 12294889.0,
      "step": 15425
    },
    {
      "entropy": 4.586993646621704,
      "epoch": 4.362454057110545,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003309990602784786,
      "loss": 4.3112,
      "mean_token_accuracy": 0.2669772684574127,
      "num_tokens": 12298973.0,
      "step": 15430
    },
    {
      "entropy": 4.534275388717651,
      "epoch": 4.363867684478372,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003308994304897947,
      "loss": 4.2487,
      "mean_token_accuracy": 0.2756001964211464,
      "num_tokens": 12302842.0,
      "step": 15435
    },
    {
      "entropy": 4.661095142364502,
      "epoch": 4.365281311846197,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00033079978901847545,
      "loss": 4.491,
      "mean_token_accuracy": 0.25247663259506226,
      "num_tokens": 12306782.0,
      "step": 15440
    },
    {
      "entropy": 4.643551015853882,
      "epoch": 4.366694939214023,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00033070013588534536,
      "loss": 4.396,
      "mean_token_accuracy": 0.25948548763990403,
      "num_tokens": 12310823.0,
      "step": 15445
    },
    {
      "entropy": 4.600639772415161,
      "epoch": 4.368108566581849,
      "grad_norm": 1.578125,
      "learning_rate": 0.00033060047111123136,
      "loss": 4.3809,
      "mean_token_accuracy": 0.27218521386384964,
      "num_tokens": 12314643.0,
      "step": 15450
    },
    {
      "entropy": 4.556357574462891,
      "epoch": 4.369522193949675,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003305007947169627,
      "loss": 4.2769,
      "mean_token_accuracy": 0.27224640250205995,
      "num_tokens": 12318320.0,
      "step": 15455
    },
    {
      "entropy": 4.642270898818969,
      "epoch": 4.3709358213175005,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00033040110672337116,
      "loss": 4.4615,
      "mean_token_accuracy": 0.25431008040905,
      "num_tokens": 12322107.0,
      "step": 15460
    },
    {
      "entropy": 4.673546266555786,
      "epoch": 4.372349448685327,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003303014071512909,
      "loss": 4.3096,
      "mean_token_accuracy": 0.2744332477450371,
      "num_tokens": 12326216.0,
      "step": 15465
    },
    {
      "entropy": 4.661748456954956,
      "epoch": 4.3737630760531525,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0003302016960215585,
      "loss": 4.4043,
      "mean_token_accuracy": 0.26477097272872924,
      "num_tokens": 12330317.0,
      "step": 15470
    },
    {
      "entropy": 4.604436254501342,
      "epoch": 4.375176703420978,
      "grad_norm": 1.6875,
      "learning_rate": 0.000330101973355013,
      "loss": 4.3141,
      "mean_token_accuracy": 0.2699108347296715,
      "num_tokens": 12334316.0,
      "step": 15475
    },
    {
      "entropy": 4.657952880859375,
      "epoch": 4.376590330788804,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003300022391724958,
      "loss": 4.3254,
      "mean_token_accuracy": 0.26807580292224886,
      "num_tokens": 12338165.0,
      "step": 15480
    },
    {
      "entropy": 4.6943638801574705,
      "epoch": 4.37800395815663,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00032990249349485075,
      "loss": 4.4916,
      "mean_token_accuracy": 0.2594763979315758,
      "num_tokens": 12342271.0,
      "step": 15485
    },
    {
      "entropy": 4.650580024719238,
      "epoch": 4.379417585524456,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000329802736342924,
      "loss": 4.338,
      "mean_token_accuracy": 0.26981528997421267,
      "num_tokens": 12345945.0,
      "step": 15490
    },
    {
      "entropy": 4.560748100280762,
      "epoch": 4.380831212892281,
      "grad_norm": 1.859375,
      "learning_rate": 0.00032970296773756437,
      "loss": 4.3901,
      "mean_token_accuracy": 0.2693938508629799,
      "num_tokens": 12349903.0,
      "step": 15495
    },
    {
      "entropy": 4.613334035873413,
      "epoch": 4.382244840260108,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003296031876996226,
      "loss": 4.3466,
      "mean_token_accuracy": 0.2637952744960785,
      "num_tokens": 12354110.0,
      "step": 15500
    },
    {
      "entropy": 4.573693561553955,
      "epoch": 4.383658467627933,
      "grad_norm": 1.859375,
      "learning_rate": 0.00032950339624995237,
      "loss": 4.3226,
      "mean_token_accuracy": 0.2738985687494278,
      "num_tokens": 12357868.0,
      "step": 15505
    },
    {
      "entropy": 4.734817218780518,
      "epoch": 4.385072094995759,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003294035934094093,
      "loss": 4.5273,
      "mean_token_accuracy": 0.254166541993618,
      "num_tokens": 12362096.0,
      "step": 15510
    },
    {
      "entropy": 4.708562755584717,
      "epoch": 4.386485722363585,
      "grad_norm": 1.59375,
      "learning_rate": 0.00032930377919885174,
      "loss": 4.2988,
      "mean_token_accuracy": 0.27131799757480624,
      "num_tokens": 12365993.0,
      "step": 15515
    },
    {
      "entropy": 4.69531478881836,
      "epoch": 4.387899349731411,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00032920395363913996,
      "loss": 4.312,
      "mean_token_accuracy": 0.2748084858059883,
      "num_tokens": 12369544.0,
      "step": 15520
    },
    {
      "entropy": 4.591352033615112,
      "epoch": 4.3893129770992365,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003291041167511372,
      "loss": 4.2779,
      "mean_token_accuracy": 0.2820843756198883,
      "num_tokens": 12373569.0,
      "step": 15525
    },
    {
      "entropy": 4.648910236358643,
      "epoch": 4.390726604467062,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003290042685557085,
      "loss": 4.3078,
      "mean_token_accuracy": 0.2752492278814316,
      "num_tokens": 12377683.0,
      "step": 15530
    },
    {
      "entropy": 4.722354507446289,
      "epoch": 4.3921402318348886,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003289044090737217,
      "loss": 4.4874,
      "mean_token_accuracy": 0.24800448268651962,
      "num_tokens": 12381837.0,
      "step": 15535
    },
    {
      "entropy": 4.625592184066773,
      "epoch": 4.393553859202714,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003288045383260468,
      "loss": 4.3012,
      "mean_token_accuracy": 0.2844582721590996,
      "num_tokens": 12385872.0,
      "step": 15540
    },
    {
      "entropy": 4.648695278167724,
      "epoch": 4.39496748657054,
      "grad_norm": 1.65625,
      "learning_rate": 0.000328704656333556,
      "loss": 4.3764,
      "mean_token_accuracy": 0.2624357044696808,
      "num_tokens": 12389842.0,
      "step": 15545
    },
    {
      "entropy": 4.701553726196289,
      "epoch": 4.396381113938366,
      "grad_norm": 1.59375,
      "learning_rate": 0.00032860476311712415,
      "loss": 4.49,
      "mean_token_accuracy": 0.263100628554821,
      "num_tokens": 12394271.0,
      "step": 15550
    },
    {
      "entropy": 4.754794645309448,
      "epoch": 4.397794741306192,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00032850485869762833,
      "loss": 4.38,
      "mean_token_accuracy": 0.26041917502880096,
      "num_tokens": 12398603.0,
      "step": 15555
    },
    {
      "entropy": 4.693798160552978,
      "epoch": 4.399208368674017,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003284049430959478,
      "loss": 4.3484,
      "mean_token_accuracy": 0.26674404591321943,
      "num_tokens": 12402778.0,
      "step": 15560
    },
    {
      "entropy": 4.671718597412109,
      "epoch": 4.400621996041844,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003283050163329645,
      "loss": 4.3472,
      "mean_token_accuracy": 0.2782501056790352,
      "num_tokens": 12406604.0,
      "step": 15565
    },
    {
      "entropy": 4.560561084747315,
      "epoch": 4.402035623409669,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003282050784295624,
      "loss": 4.3476,
      "mean_token_accuracy": 0.27606046497821807,
      "num_tokens": 12410368.0,
      "step": 15570
    },
    {
      "entropy": 4.6116859912872314,
      "epoch": 4.403449250777495,
      "grad_norm": 1.609375,
      "learning_rate": 0.000328105129406628,
      "loss": 4.3861,
      "mean_token_accuracy": 0.2611613541841507,
      "num_tokens": 12414226.0,
      "step": 15575
    },
    {
      "entropy": 4.578901481628418,
      "epoch": 4.4048628781453205,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003280051692850498,
      "loss": 4.2752,
      "mean_token_accuracy": 0.2761152878403664,
      "num_tokens": 12418351.0,
      "step": 15580
    },
    {
      "entropy": 4.654712867736817,
      "epoch": 4.406276505513147,
      "grad_norm": 1.75,
      "learning_rate": 0.0003279051980857191,
      "loss": 4.3259,
      "mean_token_accuracy": 0.27395582646131517,
      "num_tokens": 12422315.0,
      "step": 15585
    },
    {
      "entropy": 4.599606466293335,
      "epoch": 4.4076901328809726,
      "grad_norm": 1.59375,
      "learning_rate": 0.000327805215829529,
      "loss": 4.4247,
      "mean_token_accuracy": 0.2685958921909332,
      "num_tokens": 12426190.0,
      "step": 15590
    },
    {
      "entropy": 4.655700969696045,
      "epoch": 4.409103760248798,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003277052225373755,
      "loss": 4.3848,
      "mean_token_accuracy": 0.25424831807613374,
      "num_tokens": 12430249.0,
      "step": 15595
    },
    {
      "entropy": 4.6339325428009035,
      "epoch": 4.410517387616625,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003276052182301564,
      "loss": 4.4358,
      "mean_token_accuracy": 0.26135321259498595,
      "num_tokens": 12434478.0,
      "step": 15600
    },
    {
      "entropy": 4.652605104446411,
      "epoch": 4.41193101498445,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00032750520292877195,
      "loss": 4.263,
      "mean_token_accuracy": 0.2868884280323982,
      "num_tokens": 12438407.0,
      "step": 15605
    },
    {
      "entropy": 4.562043857574463,
      "epoch": 4.413344642352276,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003274051766541247,
      "loss": 4.3215,
      "mean_token_accuracy": 0.2814920753240585,
      "num_tokens": 12442372.0,
      "step": 15610
    },
    {
      "entropy": 4.584943532943726,
      "epoch": 4.414758269720101,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0003273051394271197,
      "loss": 4.4354,
      "mean_token_accuracy": 0.26405309438705443,
      "num_tokens": 12446200.0,
      "step": 15615
    },
    {
      "entropy": 4.724782180786133,
      "epoch": 4.416171897087928,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003272050912686639,
      "loss": 4.4374,
      "mean_token_accuracy": 0.251626181602478,
      "num_tokens": 12450188.0,
      "step": 15620
    },
    {
      "entropy": 4.689814233779908,
      "epoch": 4.417585524455753,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00032710503219966693,
      "loss": 4.4028,
      "mean_token_accuracy": 0.25586794018745423,
      "num_tokens": 12454263.0,
      "step": 15625
    },
    {
      "entropy": 4.731481742858887,
      "epoch": 4.418999151823579,
      "grad_norm": 1.8125,
      "learning_rate": 0.0003270049622410404,
      "loss": 4.4821,
      "mean_token_accuracy": 0.2514324218034744,
      "num_tokens": 12458243.0,
      "step": 15630
    },
    {
      "entropy": 4.760683345794678,
      "epoch": 4.420412779191405,
      "grad_norm": 1.609375,
      "learning_rate": 0.00032690488141369847,
      "loss": 4.4637,
      "mean_token_accuracy": 0.2543594479560852,
      "num_tokens": 12462397.0,
      "step": 15635
    },
    {
      "entropy": 4.713118267059326,
      "epoch": 4.421826406559231,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003268047897385571,
      "loss": 4.4023,
      "mean_token_accuracy": 0.26945195347070694,
      "num_tokens": 12466389.0,
      "step": 15640
    },
    {
      "entropy": 4.655123329162597,
      "epoch": 4.423240033927057,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00032670468723653513,
      "loss": 4.333,
      "mean_token_accuracy": 0.2688908502459526,
      "num_tokens": 12470610.0,
      "step": 15645
    },
    {
      "entropy": 4.791089153289795,
      "epoch": 4.424653661294883,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00032660457392855326,
      "loss": 4.4617,
      "mean_token_accuracy": 0.2590916737914085,
      "num_tokens": 12474662.0,
      "step": 15650
    },
    {
      "entropy": 4.6278955936431885,
      "epoch": 4.426067288662709,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003265044498355345,
      "loss": 4.3642,
      "mean_token_accuracy": 0.26205014139413835,
      "num_tokens": 12478642.0,
      "step": 15655
    },
    {
      "entropy": 4.681504726409912,
      "epoch": 4.427480916030534,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003264043149784043,
      "loss": 4.3753,
      "mean_token_accuracy": 0.2666431128978729,
      "num_tokens": 12482337.0,
      "step": 15660
    },
    {
      "entropy": 4.684455108642578,
      "epoch": 4.428894543398361,
      "grad_norm": 1.609375,
      "learning_rate": 0.00032630416937809014,
      "loss": 4.4368,
      "mean_token_accuracy": 0.2504228115081787,
      "num_tokens": 12486411.0,
      "step": 15665
    },
    {
      "entropy": 4.758188009262085,
      "epoch": 4.430308170766186,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003262040130555218,
      "loss": 4.4674,
      "mean_token_accuracy": 0.247445710003376,
      "num_tokens": 12490556.0,
      "step": 15670
    },
    {
      "entropy": 4.625837564468384,
      "epoch": 4.431721798134012,
      "grad_norm": 1.546875,
      "learning_rate": 0.00032610384603163137,
      "loss": 4.3492,
      "mean_token_accuracy": 0.2725079506635666,
      "num_tokens": 12494709.0,
      "step": 15675
    },
    {
      "entropy": 4.6481101512908936,
      "epoch": 4.433135425501837,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003260036683273531,
      "loss": 4.4225,
      "mean_token_accuracy": 0.26225157231092455,
      "num_tokens": 12498739.0,
      "step": 15680
    },
    {
      "entropy": 4.705290460586548,
      "epoch": 4.434549052869664,
      "grad_norm": 1.75,
      "learning_rate": 0.0003259034799636236,
      "loss": 4.4735,
      "mean_token_accuracy": 0.2473021298646927,
      "num_tokens": 12502788.0,
      "step": 15685
    },
    {
      "entropy": 4.626170825958252,
      "epoch": 4.435962680237489,
      "grad_norm": 1.625,
      "learning_rate": 0.00032580328096138154,
      "loss": 4.346,
      "mean_token_accuracy": 0.2648080036044121,
      "num_tokens": 12506768.0,
      "step": 15690
    },
    {
      "entropy": 4.687180471420288,
      "epoch": 4.437376307605315,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00032570307134156796,
      "loss": 4.4823,
      "mean_token_accuracy": 0.25860633999109267,
      "num_tokens": 12510767.0,
      "step": 15695
    },
    {
      "entropy": 4.661324977874756,
      "epoch": 4.4387899349731414,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0003256028511251259,
      "loss": 4.352,
      "mean_token_accuracy": 0.2712166145443916,
      "num_tokens": 12514647.0,
      "step": 15700
    },
    {
      "entropy": 4.585727787017822,
      "epoch": 4.440203562340967,
      "grad_norm": 1.640625,
      "learning_rate": 0.000325502620333001,
      "loss": 4.3861,
      "mean_token_accuracy": 0.2662515267729759,
      "num_tokens": 12518960.0,
      "step": 15705
    },
    {
      "entropy": 4.634370756149292,
      "epoch": 4.441617189708793,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003254023789861407,
      "loss": 4.3503,
      "mean_token_accuracy": 0.26343195587396623,
      "num_tokens": 12522958.0,
      "step": 15710
    },
    {
      "entropy": 4.66596474647522,
      "epoch": 4.443030817076618,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003253021271054948,
      "loss": 4.4479,
      "mean_token_accuracy": 0.2534408688545227,
      "num_tokens": 12526941.0,
      "step": 15715
    },
    {
      "entropy": 4.602621173858642,
      "epoch": 4.444444444444445,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003252018647120155,
      "loss": 4.4308,
      "mean_token_accuracy": 0.26129013001918794,
      "num_tokens": 12531043.0,
      "step": 15720
    },
    {
      "entropy": 4.67469596862793,
      "epoch": 4.44585807181227,
      "grad_norm": 1.8125,
      "learning_rate": 0.00032510159182665686,
      "loss": 4.4285,
      "mean_token_accuracy": 0.25666120648384094,
      "num_tokens": 12534712.0,
      "step": 15725
    },
    {
      "entropy": 4.726824331283569,
      "epoch": 4.447271699180096,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003250013084703754,
      "loss": 4.4186,
      "mean_token_accuracy": 0.25678087770938873,
      "num_tokens": 12538897.0,
      "step": 15730
    },
    {
      "entropy": 4.694701528549194,
      "epoch": 4.448685326547922,
      "grad_norm": 1.65625,
      "learning_rate": 0.00032490101466412956,
      "loss": 4.3347,
      "mean_token_accuracy": 0.2668992027640343,
      "num_tokens": 12542798.0,
      "step": 15735
    },
    {
      "entropy": 4.629713773727417,
      "epoch": 4.450098953915748,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003248007104288803,
      "loss": 4.3041,
      "mean_token_accuracy": 0.27692056745290755,
      "num_tokens": 12547156.0,
      "step": 15740
    },
    {
      "entropy": 4.631553745269775,
      "epoch": 4.451512581283573,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00032470039578559053,
      "loss": 4.4862,
      "mean_token_accuracy": 0.25902546793222425,
      "num_tokens": 12551094.0,
      "step": 15745
    },
    {
      "entropy": 4.6671209812164305,
      "epoch": 4.4529262086514,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003246000707552254,
      "loss": 4.462,
      "mean_token_accuracy": 0.258967787027359,
      "num_tokens": 12555241.0,
      "step": 15750
    },
    {
      "entropy": 4.7625938892364506,
      "epoch": 4.4543398360192255,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00032449973535875215,
      "loss": 4.4721,
      "mean_token_accuracy": 0.2533877268433571,
      "num_tokens": 12559200.0,
      "step": 15755
    },
    {
      "entropy": 4.694896793365478,
      "epoch": 4.455753463387051,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003243993896171403,
      "loss": 4.3175,
      "mean_token_accuracy": 0.2695566311478615,
      "num_tokens": 12563125.0,
      "step": 15760
    },
    {
      "entropy": 4.7120684623718265,
      "epoch": 4.457167090754877,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003242990335513615,
      "loss": 4.4263,
      "mean_token_accuracy": 0.2546339839696884,
      "num_tokens": 12566927.0,
      "step": 15765
    },
    {
      "entropy": 4.691117143630981,
      "epoch": 4.458580718122703,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003241986671823896,
      "loss": 4.5248,
      "mean_token_accuracy": 0.2487319990992546,
      "num_tokens": 12570875.0,
      "step": 15770
    },
    {
      "entropy": 4.689526557922363,
      "epoch": 4.459994345490529,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00032409829053120043,
      "loss": 4.3843,
      "mean_token_accuracy": 0.2671830415725708,
      "num_tokens": 12574863.0,
      "step": 15775
    },
    {
      "entropy": 4.715511417388916,
      "epoch": 4.461407972858354,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00032399790361877217,
      "loss": 4.4439,
      "mean_token_accuracy": 0.26151473373174666,
      "num_tokens": 12579052.0,
      "step": 15780
    },
    {
      "entropy": 4.719805574417114,
      "epoch": 4.462821600226181,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00032389750646608513,
      "loss": 4.4229,
      "mean_token_accuracy": 0.2587859630584717,
      "num_tokens": 12583134.0,
      "step": 15785
    },
    {
      "entropy": 4.640354824066162,
      "epoch": 4.464235227594006,
      "grad_norm": 1.640625,
      "learning_rate": 0.00032379709909412154,
      "loss": 4.3607,
      "mean_token_accuracy": 0.2699592739343643,
      "num_tokens": 12587435.0,
      "step": 15790
    },
    {
      "entropy": 4.625859975814819,
      "epoch": 4.465648854961832,
      "grad_norm": 1.703125,
      "learning_rate": 0.00032369668152386593,
      "loss": 4.4465,
      "mean_token_accuracy": 0.2669498577713966,
      "num_tokens": 12591783.0,
      "step": 15795
    },
    {
      "entropy": 4.717779111862183,
      "epoch": 4.467062482329658,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00032359625377630514,
      "loss": 4.422,
      "mean_token_accuracy": 0.25818881690502166,
      "num_tokens": 12595650.0,
      "step": 15800
    },
    {
      "entropy": 4.735414361953735,
      "epoch": 4.468476109697484,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003234958158724277,
      "loss": 4.4606,
      "mean_token_accuracy": 0.2517810195684433,
      "num_tokens": 12599839.0,
      "step": 15805
    },
    {
      "entropy": 4.639988899230957,
      "epoch": 4.4698897370653095,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003233953678332247,
      "loss": 4.2596,
      "mean_token_accuracy": 0.28316949903964994,
      "num_tokens": 12603742.0,
      "step": 15810
    },
    {
      "entropy": 4.614012670516968,
      "epoch": 4.471303364433135,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003232949096796891,
      "loss": 4.2822,
      "mean_token_accuracy": 0.2716996505856514,
      "num_tokens": 12607556.0,
      "step": 15815
    },
    {
      "entropy": 4.637189245223999,
      "epoch": 4.4727169918009615,
      "grad_norm": 1.53125,
      "learning_rate": 0.00032319444143281604,
      "loss": 4.3232,
      "mean_token_accuracy": 0.2627250403165817,
      "num_tokens": 12611584.0,
      "step": 15820
    },
    {
      "entropy": 4.660959911346436,
      "epoch": 4.474130619168787,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003230939631136027,
      "loss": 4.3489,
      "mean_token_accuracy": 0.26662867069244384,
      "num_tokens": 12615618.0,
      "step": 15825
    },
    {
      "entropy": 4.762927389144897,
      "epoch": 4.475544246536613,
      "grad_norm": 1.734375,
      "learning_rate": 0.00032299347474304846,
      "loss": 4.4809,
      "mean_token_accuracy": 0.25928491950035093,
      "num_tokens": 12619874.0,
      "step": 15830
    },
    {
      "entropy": 4.684660243988037,
      "epoch": 4.476957873904439,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003228929763421548,
      "loss": 4.2991,
      "mean_token_accuracy": 0.2686818614602089,
      "num_tokens": 12623845.0,
      "step": 15835
    },
    {
      "entropy": 4.693379306793213,
      "epoch": 4.478371501272265,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003227924679319252,
      "loss": 4.4562,
      "mean_token_accuracy": 0.2601957738399506,
      "num_tokens": 12627930.0,
      "step": 15840
    },
    {
      "entropy": 4.657646703720093,
      "epoch": 4.47978512864009,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00032269194953336546,
      "loss": 4.3986,
      "mean_token_accuracy": 0.2704332277178764,
      "num_tokens": 12632047.0,
      "step": 15845
    },
    {
      "entropy": 4.6801595211029055,
      "epoch": 4.481198756007917,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003225914211674832,
      "loss": 4.3515,
      "mean_token_accuracy": 0.2604921832680702,
      "num_tokens": 12635804.0,
      "step": 15850
    },
    {
      "entropy": 4.6604835987091064,
      "epoch": 4.482612383375742,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00032249088285528805,
      "loss": 4.3553,
      "mean_token_accuracy": 0.2761675387620926,
      "num_tokens": 12639595.0,
      "step": 15855
    },
    {
      "entropy": 4.611642837524414,
      "epoch": 4.484026010743568,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00032239033461779217,
      "loss": 4.328,
      "mean_token_accuracy": 0.2713270038366318,
      "num_tokens": 12643511.0,
      "step": 15860
    },
    {
      "entropy": 4.666283655166626,
      "epoch": 4.4854396381113935,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003222897764760093,
      "loss": 4.3969,
      "mean_token_accuracy": 0.2555485963821411,
      "num_tokens": 12647247.0,
      "step": 15865
    },
    {
      "entropy": 4.720825624465943,
      "epoch": 4.48685326547922,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003221892084509557,
      "loss": 4.522,
      "mean_token_accuracy": 0.24527915716171264,
      "num_tokens": 12651074.0,
      "step": 15870
    },
    {
      "entropy": 4.684508657455444,
      "epoch": 4.4882668928470455,
      "grad_norm": 1.65625,
      "learning_rate": 0.00032208863056364936,
      "loss": 4.3576,
      "mean_token_accuracy": 0.26309540718793867,
      "num_tokens": 12654910.0,
      "step": 15875
    },
    {
      "entropy": 4.531758785247803,
      "epoch": 4.489680520214871,
      "grad_norm": 1.640625,
      "learning_rate": 0.00032198804283511034,
      "loss": 4.2191,
      "mean_token_accuracy": 0.27934612780809404,
      "num_tokens": 12658909.0,
      "step": 15880
    },
    {
      "entropy": 4.621357679367065,
      "epoch": 4.4910941475826975,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003218874452863609,
      "loss": 4.3529,
      "mean_token_accuracy": 0.27231679409742354,
      "num_tokens": 12662765.0,
      "step": 15885
    },
    {
      "entropy": 4.651836681365967,
      "epoch": 4.492507774950523,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003217868379384254,
      "loss": 4.4555,
      "mean_token_accuracy": 0.2582566738128662,
      "num_tokens": 12667048.0,
      "step": 15890
    },
    {
      "entropy": 4.594143056869507,
      "epoch": 4.493921402318349,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003216862208123301,
      "loss": 4.3325,
      "mean_token_accuracy": 0.25862763822078705,
      "num_tokens": 12670636.0,
      "step": 15895
    },
    {
      "entropy": 4.6581981658935545,
      "epoch": 4.495335029686174,
      "grad_norm": 1.59375,
      "learning_rate": 0.00032158559392910337,
      "loss": 4.3365,
      "mean_token_accuracy": 0.25964539498090744,
      "num_tokens": 12674857.0,
      "step": 15900
    },
    {
      "entropy": 4.64360032081604,
      "epoch": 4.496748657054001,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003214849573097755,
      "loss": 4.3682,
      "mean_token_accuracy": 0.26310827285051347,
      "num_tokens": 12678921.0,
      "step": 15905
    },
    {
      "entropy": 4.693691205978394,
      "epoch": 4.498162284421826,
      "grad_norm": 1.640625,
      "learning_rate": 0.00032138431097537904,
      "loss": 4.4808,
      "mean_token_accuracy": 0.2541467934846878,
      "num_tokens": 12682782.0,
      "step": 15910
    },
    {
      "entropy": 4.674745225906372,
      "epoch": 4.499575911789652,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00032128365494694846,
      "loss": 4.368,
      "mean_token_accuracy": 0.268680739402771,
      "num_tokens": 12686728.0,
      "step": 15915
    },
    {
      "entropy": 4.592482471466065,
      "epoch": 4.500989539157478,
      "grad_norm": 1.625,
      "learning_rate": 0.0003211829892455201,
      "loss": 4.3002,
      "mean_token_accuracy": 0.2826909214258194,
      "num_tokens": 12690547.0,
      "step": 15920
    },
    {
      "entropy": 4.759224557876587,
      "epoch": 4.502403166525304,
      "grad_norm": 1.734375,
      "learning_rate": 0.00032108231389213265,
      "loss": 4.4717,
      "mean_token_accuracy": 0.2581175222992897,
      "num_tokens": 12694432.0,
      "step": 15925
    },
    {
      "entropy": 4.642607164382935,
      "epoch": 4.5038167938931295,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003209816289078266,
      "loss": 4.2928,
      "mean_token_accuracy": 0.2677662014961243,
      "num_tokens": 12698870.0,
      "step": 15930
    },
    {
      "entropy": 4.635746192932129,
      "epoch": 4.505230421260956,
      "grad_norm": 1.703125,
      "learning_rate": 0.00032088093431364436,
      "loss": 4.3891,
      "mean_token_accuracy": 0.27170401364564895,
      "num_tokens": 12702653.0,
      "step": 15935
    },
    {
      "entropy": 4.615369606018066,
      "epoch": 4.5066440486287815,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00032078023013063075,
      "loss": 4.3145,
      "mean_token_accuracy": 0.2721775844693184,
      "num_tokens": 12706955.0,
      "step": 15940
    },
    {
      "entropy": 4.682344961166382,
      "epoch": 4.508057675996607,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00032067951637983194,
      "loss": 4.4513,
      "mean_token_accuracy": 0.25059485286474226,
      "num_tokens": 12710731.0,
      "step": 15945
    },
    {
      "entropy": 4.772052860260009,
      "epoch": 4.509471303364434,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00032057879308229674,
      "loss": 4.5268,
      "mean_token_accuracy": 0.2473808690905571,
      "num_tokens": 12714572.0,
      "step": 15950
    },
    {
      "entropy": 4.618041229248047,
      "epoch": 4.510884930732259,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003204780602590757,
      "loss": 4.3546,
      "mean_token_accuracy": 0.2634569123387337,
      "num_tokens": 12718603.0,
      "step": 15955
    },
    {
      "entropy": 4.522592496871948,
      "epoch": 4.512298558100085,
      "grad_norm": 1.71875,
      "learning_rate": 0.00032037731793122117,
      "loss": 4.2922,
      "mean_token_accuracy": 0.27907453179359437,
      "num_tokens": 12722807.0,
      "step": 15960
    },
    {
      "entropy": 4.5705303192138675,
      "epoch": 4.51371218546791,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003202765661197879,
      "loss": 4.3342,
      "mean_token_accuracy": 0.26796761751174925,
      "num_tokens": 12726876.0,
      "step": 15965
    },
    {
      "entropy": 4.590322875976563,
      "epoch": 4.515125812835737,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00032017580484583225,
      "loss": 4.3702,
      "mean_token_accuracy": 0.27391931116580964,
      "num_tokens": 12730972.0,
      "step": 15970
    },
    {
      "entropy": 4.771474742889405,
      "epoch": 4.516539440203562,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00032007503413041276,
      "loss": 4.4664,
      "mean_token_accuracy": 0.25466011464595795,
      "num_tokens": 12734789.0,
      "step": 15975
    },
    {
      "entropy": 4.694821500778199,
      "epoch": 4.517953067571388,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00031997425399458984,
      "loss": 4.3668,
      "mean_token_accuracy": 0.26284207999706266,
      "num_tokens": 12738525.0,
      "step": 15980
    },
    {
      "entropy": 4.729448318481445,
      "epoch": 4.519366694939214,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031987346445942603,
      "loss": 4.4458,
      "mean_token_accuracy": 0.2658174768090248,
      "num_tokens": 12742546.0,
      "step": 15985
    },
    {
      "entropy": 4.678018474578858,
      "epoch": 4.52078032230704,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031977266554598547,
      "loss": 4.3494,
      "mean_token_accuracy": 0.2676664084196091,
      "num_tokens": 12746184.0,
      "step": 15990
    },
    {
      "entropy": 4.63763198852539,
      "epoch": 4.5221939496748655,
      "grad_norm": 1.671875,
      "learning_rate": 0.00031967185727533486,
      "loss": 4.3896,
      "mean_token_accuracy": 0.26811966150999067,
      "num_tokens": 12750161.0,
      "step": 15995
    },
    {
      "entropy": 4.617028665542603,
      "epoch": 4.523607577042691,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00031957103966854226,
      "loss": 4.4089,
      "mean_token_accuracy": 0.2661013096570969,
      "num_tokens": 12754188.0,
      "step": 16000
    },
    {
      "epoch": 4.523607577042691,
      "eval_entropy": 4.5117043325193045,
      "eval_loss": 5.005865097045898,
      "eval_mean_token_accuracy": 0.2476137896945593,
      "eval_num_tokens": 12754188.0,
      "eval_runtime": 6.6884,
      "eval_samples_per_second": 1687.107,
      "eval_steps_per_second": 210.963,
      "step": 16000
    },
    {
      "entropy": 4.650805902481079,
      "epoch": 4.525021204410518,
      "grad_norm": 1.5546875,
      "learning_rate": 0.000319470212746678,
      "loss": 4.4057,
      "mean_token_accuracy": 0.266128471493721,
      "num_tokens": 12758279.0,
      "step": 16005
    },
    {
      "entropy": 4.707661390304565,
      "epoch": 4.526434831778343,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003193693765308142,
      "loss": 4.4745,
      "mean_token_accuracy": 0.2450110599398613,
      "num_tokens": 12762221.0,
      "step": 16010
    },
    {
      "entropy": 4.722390174865723,
      "epoch": 4.527848459146169,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003192685310420251,
      "loss": 4.5111,
      "mean_token_accuracy": 0.25305157750844953,
      "num_tokens": 12765957.0,
      "step": 16015
    },
    {
      "entropy": 4.700660514831543,
      "epoch": 4.529262086513995,
      "grad_norm": 1.625,
      "learning_rate": 0.00031916767630138683,
      "loss": 4.3564,
      "mean_token_accuracy": 0.26745724827051165,
      "num_tokens": 12769763.0,
      "step": 16020
    },
    {
      "entropy": 4.633463716506958,
      "epoch": 4.530675713881821,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00031906681232997733,
      "loss": 4.3284,
      "mean_token_accuracy": 0.2605445757508278,
      "num_tokens": 12773808.0,
      "step": 16025
    },
    {
      "entropy": 4.461117792129516,
      "epoch": 4.532089341249646,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003189659391488765,
      "loss": 4.2524,
      "mean_token_accuracy": 0.2713987693190575,
      "num_tokens": 12777889.0,
      "step": 16030
    },
    {
      "entropy": 4.684966897964477,
      "epoch": 4.533502968617473,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003188650567791663,
      "loss": 4.4431,
      "mean_token_accuracy": 0.25671309381723406,
      "num_tokens": 12781925.0,
      "step": 16035
    },
    {
      "entropy": 4.60801215171814,
      "epoch": 4.534916595985298,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003187641652419304,
      "loss": 4.3619,
      "mean_token_accuracy": 0.269709450006485,
      "num_tokens": 12785716.0,
      "step": 16040
    },
    {
      "entropy": 4.617230415344238,
      "epoch": 4.536330223353124,
      "grad_norm": 1.59375,
      "learning_rate": 0.0003186632645582547,
      "loss": 4.366,
      "mean_token_accuracy": 0.2624741390347481,
      "num_tokens": 12789872.0,
      "step": 16045
    },
    {
      "entropy": 4.711508798599243,
      "epoch": 4.53774385072095,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0003185623547492267,
      "loss": 4.5039,
      "mean_token_accuracy": 0.2512928739190102,
      "num_tokens": 12793816.0,
      "step": 16050
    },
    {
      "entropy": 4.615178537368775,
      "epoch": 4.539157478088776,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003184614358359359,
      "loss": 4.3217,
      "mean_token_accuracy": 0.2818364784121513,
      "num_tokens": 12797793.0,
      "step": 16055
    },
    {
      "entropy": 4.656971025466919,
      "epoch": 4.540571105456602,
      "grad_norm": 1.640625,
      "learning_rate": 0.00031836050783947377,
      "loss": 4.3185,
      "mean_token_accuracy": 0.26608929932117464,
      "num_tokens": 12801949.0,
      "step": 16060
    },
    {
      "entropy": 4.663751268386841,
      "epoch": 4.541984732824427,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003182595707809337,
      "loss": 4.3608,
      "mean_token_accuracy": 0.25608091056346893,
      "num_tokens": 12805667.0,
      "step": 16065
    },
    {
      "entropy": 4.649747896194458,
      "epoch": 4.543398360192254,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003181586246814108,
      "loss": 4.2848,
      "mean_token_accuracy": 0.28083213716745375,
      "num_tokens": 12809713.0,
      "step": 16070
    },
    {
      "entropy": 4.684282493591309,
      "epoch": 4.544811987560079,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003180576695620022,
      "loss": 4.3842,
      "mean_token_accuracy": 0.26710040122270584,
      "num_tokens": 12813494.0,
      "step": 16075
    },
    {
      "entropy": 4.681137323379517,
      "epoch": 4.546225614927905,
      "grad_norm": 1.625,
      "learning_rate": 0.0003179567054438069,
      "loss": 4.4537,
      "mean_token_accuracy": 0.25973714590072633,
      "num_tokens": 12817517.0,
      "step": 16080
    },
    {
      "entropy": 4.674393081665039,
      "epoch": 4.54763924229573,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00031785573234792587,
      "loss": 4.3584,
      "mean_token_accuracy": 0.26133899837732316,
      "num_tokens": 12821632.0,
      "step": 16085
    },
    {
      "entropy": 4.617424869537354,
      "epoch": 4.549052869663557,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003177547502954617,
      "loss": 4.2205,
      "mean_token_accuracy": 0.28029831796884536,
      "num_tokens": 12825513.0,
      "step": 16090
    },
    {
      "entropy": 4.652894830703735,
      "epoch": 4.550466497031382,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031765375930751925,
      "loss": 4.2586,
      "mean_token_accuracy": 0.27874798625707625,
      "num_tokens": 12829295.0,
      "step": 16095
    },
    {
      "entropy": 4.704120063781739,
      "epoch": 4.551880124399208,
      "grad_norm": 1.84375,
      "learning_rate": 0.00031755275940520476,
      "loss": 4.362,
      "mean_token_accuracy": 0.26527885496616366,
      "num_tokens": 12833257.0,
      "step": 16100
    },
    {
      "entropy": 4.684330034255981,
      "epoch": 4.553293751767034,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003174517506096267,
      "loss": 4.5072,
      "mean_token_accuracy": 0.25935696363449096,
      "num_tokens": 12837209.0,
      "step": 16105
    },
    {
      "entropy": 4.592080926895141,
      "epoch": 4.55470737913486,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003173507329418953,
      "loss": 4.2782,
      "mean_token_accuracy": 0.2748946323990822,
      "num_tokens": 12841013.0,
      "step": 16110
    },
    {
      "entropy": 4.571475505828857,
      "epoch": 4.556121006502686,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003172497064231226,
      "loss": 4.3334,
      "mean_token_accuracy": 0.2652077481150627,
      "num_tokens": 12844937.0,
      "step": 16115
    },
    {
      "entropy": 4.625619745254516,
      "epoch": 4.557534633870512,
      "grad_norm": 1.703125,
      "learning_rate": 0.00031714867107442255,
      "loss": 4.3399,
      "mean_token_accuracy": 0.2786511093378067,
      "num_tokens": 12849046.0,
      "step": 16120
    },
    {
      "entropy": 4.64364104270935,
      "epoch": 4.558948261238338,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031704762691691084,
      "loss": 4.3749,
      "mean_token_accuracy": 0.27038692235946654,
      "num_tokens": 12852939.0,
      "step": 16125
    },
    {
      "entropy": 4.601371383666992,
      "epoch": 4.560361888606163,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003169465739717051,
      "loss": 4.3499,
      "mean_token_accuracy": 0.27393107712268827,
      "num_tokens": 12856888.0,
      "step": 16130
    },
    {
      "entropy": 4.706913232803345,
      "epoch": 4.56177551597399,
      "grad_norm": 1.671875,
      "learning_rate": 0.00031684551225992486,
      "loss": 4.4549,
      "mean_token_accuracy": 0.2561491698026657,
      "num_tokens": 12860860.0,
      "step": 16135
    },
    {
      "entropy": 4.599346494674682,
      "epoch": 4.563189143341815,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0003167444418026913,
      "loss": 4.3429,
      "mean_token_accuracy": 0.2740123733878136,
      "num_tokens": 12864884.0,
      "step": 16140
    },
    {
      "entropy": 4.539031648635865,
      "epoch": 4.564602770709641,
      "grad_norm": 1.734375,
      "learning_rate": 0.00031664336262112754,
      "loss": 4.3189,
      "mean_token_accuracy": 0.2754500761628151,
      "num_tokens": 12868985.0,
      "step": 16145
    },
    {
      "entropy": 4.633904933929443,
      "epoch": 4.566016398077466,
      "grad_norm": 1.5,
      "learning_rate": 0.00031654227473635845,
      "loss": 4.376,
      "mean_token_accuracy": 0.26220581531524656,
      "num_tokens": 12873192.0,
      "step": 16150
    },
    {
      "entropy": 4.665906858444214,
      "epoch": 4.567430025445293,
      "grad_norm": 1.65625,
      "learning_rate": 0.000316441178169511,
      "loss": 4.3536,
      "mean_token_accuracy": 0.2704638734459877,
      "num_tokens": 12877360.0,
      "step": 16155
    },
    {
      "entropy": 4.680084753036499,
      "epoch": 4.568843652813118,
      "grad_norm": 1.65625,
      "learning_rate": 0.00031634007294171335,
      "loss": 4.3605,
      "mean_token_accuracy": 0.26528591066598894,
      "num_tokens": 12881339.0,
      "step": 16160
    },
    {
      "entropy": 4.719594860076905,
      "epoch": 4.570257280180944,
      "grad_norm": 1.765625,
      "learning_rate": 0.0003162389590740962,
      "loss": 4.5175,
      "mean_token_accuracy": 0.24678635448217393,
      "num_tokens": 12885368.0,
      "step": 16165
    },
    {
      "entropy": 4.677007436752319,
      "epoch": 4.5716709075487705,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003161378365877916,
      "loss": 4.4222,
      "mean_token_accuracy": 0.26202949285507204,
      "num_tokens": 12889539.0,
      "step": 16170
    },
    {
      "entropy": 4.704844188690186,
      "epoch": 4.573084534916596,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003160367055039336,
      "loss": 4.4913,
      "mean_token_accuracy": 0.2514870822429657,
      "num_tokens": 12893845.0,
      "step": 16175
    },
    {
      "entropy": 4.656061458587646,
      "epoch": 4.574498162284422,
      "grad_norm": 1.765625,
      "learning_rate": 0.00031593556584365783,
      "loss": 4.4017,
      "mean_token_accuracy": 0.2664449453353882,
      "num_tokens": 12897693.0,
      "step": 16180
    },
    {
      "entropy": 4.696681261062622,
      "epoch": 4.575911789652247,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031583441762810205,
      "loss": 4.406,
      "mean_token_accuracy": 0.2572014585137367,
      "num_tokens": 12901661.0,
      "step": 16185
    },
    {
      "entropy": 4.642475986480713,
      "epoch": 4.577325417020074,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00031573326087840537,
      "loss": 4.2673,
      "mean_token_accuracy": 0.27192078828811644,
      "num_tokens": 12905463.0,
      "step": 16190
    },
    {
      "entropy": 4.547554302215576,
      "epoch": 4.578739044387899,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00031563209561570913,
      "loss": 4.1855,
      "mean_token_accuracy": 0.2786505088210106,
      "num_tokens": 12909329.0,
      "step": 16195
    },
    {
      "entropy": 4.601930809020996,
      "epoch": 4.580152671755725,
      "grad_norm": 1.65625,
      "learning_rate": 0.00031553092186115616,
      "loss": 4.4426,
      "mean_token_accuracy": 0.25930742770433424,
      "num_tokens": 12913378.0,
      "step": 16200
    },
    {
      "entropy": 4.643219470977783,
      "epoch": 4.581566299123551,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003154297396358911,
      "loss": 4.3661,
      "mean_token_accuracy": 0.27701577693223955,
      "num_tokens": 12917436.0,
      "step": 16205
    },
    {
      "entropy": 4.653262662887573,
      "epoch": 4.582979926491377,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003153285489610605,
      "loss": 4.3196,
      "mean_token_accuracy": 0.2773186400532722,
      "num_tokens": 12921648.0,
      "step": 16210
    },
    {
      "entropy": 4.686037063598633,
      "epoch": 4.584393553859202,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003152273498578126,
      "loss": 4.3894,
      "mean_token_accuracy": 0.2673725888133049,
      "num_tokens": 12925865.0,
      "step": 16215
    },
    {
      "entropy": 4.5330445766448975,
      "epoch": 4.585807181227029,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003151261423472973,
      "loss": 4.1727,
      "mean_token_accuracy": 0.29568320512771606,
      "num_tokens": 12929838.0,
      "step": 16220
    },
    {
      "entropy": 4.579984998703003,
      "epoch": 4.5872208085948545,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00031502492645066636,
      "loss": 4.3144,
      "mean_token_accuracy": 0.26929624378681183,
      "num_tokens": 12933851.0,
      "step": 16225
    },
    {
      "entropy": 4.635414981842041,
      "epoch": 4.58863443596268,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003149237021890734,
      "loss": 4.3834,
      "mean_token_accuracy": 0.26617604345083234,
      "num_tokens": 12938078.0,
      "step": 16230
    },
    {
      "entropy": 4.637782049179077,
      "epoch": 4.5900480633305065,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031482246958367346,
      "loss": 4.3342,
      "mean_token_accuracy": 0.2732896596193314,
      "num_tokens": 12941722.0,
      "step": 16235
    },
    {
      "entropy": 4.640135955810547,
      "epoch": 4.591461690698332,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031472122865562363,
      "loss": 4.4476,
      "mean_token_accuracy": 0.2589869245886803,
      "num_tokens": 12945642.0,
      "step": 16240
    },
    {
      "entropy": 4.7090253829956055,
      "epoch": 4.592875318066158,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0003146199794260827,
      "loss": 4.4275,
      "mean_token_accuracy": 0.26446429938077926,
      "num_tokens": 12949656.0,
      "step": 16245
    },
    {
      "entropy": 4.6690239906311035,
      "epoch": 4.594288945433983,
      "grad_norm": 1.640625,
      "learning_rate": 0.00031451872191621107,
      "loss": 4.4063,
      "mean_token_accuracy": 0.25611686408519746,
      "num_tokens": 12953854.0,
      "step": 16250
    },
    {
      "entropy": 4.821361017227173,
      "epoch": 4.59570257280181,
      "grad_norm": 1.875,
      "learning_rate": 0.0003144174561471708,
      "loss": 4.4164,
      "mean_token_accuracy": 0.25017520487308503,
      "num_tokens": 12957664.0,
      "step": 16255
    },
    {
      "entropy": 4.740560817718506,
      "epoch": 4.597116200169635,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003143161821401261,
      "loss": 4.4794,
      "mean_token_accuracy": 0.25151596665382386,
      "num_tokens": 12961542.0,
      "step": 16260
    },
    {
      "entropy": 4.686863374710083,
      "epoch": 4.598529827537461,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0003142148999162424,
      "loss": 4.356,
      "mean_token_accuracy": 0.265330608189106,
      "num_tokens": 12965224.0,
      "step": 16265
    },
    {
      "entropy": 4.625082969665527,
      "epoch": 4.599943454905287,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0003141136094966871,
      "loss": 4.373,
      "mean_token_accuracy": 0.26099874824285507,
      "num_tokens": 12968984.0,
      "step": 16270
    },
    {
      "entropy": 4.529994630813599,
      "epoch": 4.601357082273113,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003140123109026293,
      "loss": 4.3949,
      "mean_token_accuracy": 0.2702192410826683,
      "num_tokens": 12972959.0,
      "step": 16275
    },
    {
      "entropy": 4.645225286483765,
      "epoch": 4.6027707096409385,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00031391100415523984,
      "loss": 4.3161,
      "mean_token_accuracy": 0.27117006480693817,
      "num_tokens": 12976937.0,
      "step": 16280
    },
    {
      "entropy": 4.699813461303711,
      "epoch": 4.604184337008764,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000313809689275691,
      "loss": 4.3764,
      "mean_token_accuracy": 0.2727006822824478,
      "num_tokens": 12980969.0,
      "step": 16285
    },
    {
      "entropy": 4.724130010604858,
      "epoch": 4.6055979643765905,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003137083662851572,
      "loss": 4.3607,
      "mean_token_accuracy": 0.27018954753875735,
      "num_tokens": 12985017.0,
      "step": 16290
    },
    {
      "entropy": 4.629456472396851,
      "epoch": 4.607011591744416,
      "grad_norm": 1.703125,
      "learning_rate": 0.00031360703520481414,
      "loss": 4.3388,
      "mean_token_accuracy": 0.2669545769691467,
      "num_tokens": 12988718.0,
      "step": 16295
    },
    {
      "entropy": 4.687915086746216,
      "epoch": 4.608425219112242,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031350569605583943,
      "loss": 4.3881,
      "mean_token_accuracy": 0.2585668295621872,
      "num_tokens": 12992869.0,
      "step": 16300
    },
    {
      "entropy": 4.760725212097168,
      "epoch": 4.609838846480068,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003134043488594124,
      "loss": 4.5148,
      "mean_token_accuracy": 0.2573800429701805,
      "num_tokens": 12996777.0,
      "step": 16305
    },
    {
      "entropy": 4.687014770507813,
      "epoch": 4.611252473847894,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003133029936367141,
      "loss": 4.3668,
      "mean_token_accuracy": 0.26187960654497144,
      "num_tokens": 13000985.0,
      "step": 16310
    },
    {
      "entropy": 4.654061365127563,
      "epoch": 4.612666101215719,
      "grad_norm": 1.515625,
      "learning_rate": 0.0003132016304089268,
      "loss": 4.4281,
      "mean_token_accuracy": 0.2576258808374405,
      "num_tokens": 13005130.0,
      "step": 16315
    },
    {
      "entropy": 4.538539028167724,
      "epoch": 4.614079728583546,
      "grad_norm": 1.5625,
      "learning_rate": 0.000313100259197235,
      "loss": 4.2745,
      "mean_token_accuracy": 0.27330050617456436,
      "num_tokens": 13009234.0,
      "step": 16320
    },
    {
      "entropy": 4.602560424804688,
      "epoch": 4.615493355951371,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00031299888002282467,
      "loss": 4.2844,
      "mean_token_accuracy": 0.2615279108285904,
      "num_tokens": 13013256.0,
      "step": 16325
    },
    {
      "entropy": 4.681260871887207,
      "epoch": 4.616906983319197,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003128974929068834,
      "loss": 4.4195,
      "mean_token_accuracy": 0.2613119289278984,
      "num_tokens": 13017203.0,
      "step": 16330
    },
    {
      "entropy": 4.69005651473999,
      "epoch": 4.618320610687023,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00031279609787060036,
      "loss": 4.3935,
      "mean_token_accuracy": 0.2599942982196808,
      "num_tokens": 13021182.0,
      "step": 16335
    },
    {
      "entropy": 4.656219053268432,
      "epoch": 4.619734238054849,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003126946949351668,
      "loss": 4.3086,
      "mean_token_accuracy": 0.2669693425297737,
      "num_tokens": 13025329.0,
      "step": 16340
    },
    {
      "entropy": 4.649556016921997,
      "epoch": 4.6211478654226745,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003125932841217749,
      "loss": 4.3792,
      "mean_token_accuracy": 0.2729396492242813,
      "num_tokens": 13029787.0,
      "step": 16345
    },
    {
      "entropy": 4.627897787094116,
      "epoch": 4.6225614927905,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003124918654516192,
      "loss": 4.3517,
      "mean_token_accuracy": 0.2651043996214867,
      "num_tokens": 13033977.0,
      "step": 16350
    },
    {
      "entropy": 4.7114544868469235,
      "epoch": 4.623975120158327,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0003123904389458954,
      "loss": 4.3879,
      "mean_token_accuracy": 0.25620434433221817,
      "num_tokens": 13037639.0,
      "step": 16355
    },
    {
      "entropy": 4.706451559066773,
      "epoch": 4.625388747526152,
      "grad_norm": 1.75,
      "learning_rate": 0.0003122890046258011,
      "loss": 4.3711,
      "mean_token_accuracy": 0.26248732805252073,
      "num_tokens": 13041399.0,
      "step": 16360
    },
    {
      "entropy": 4.6298846244812015,
      "epoch": 4.626802374893978,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031218756251253537,
      "loss": 4.2913,
      "mean_token_accuracy": 0.2755704566836357,
      "num_tokens": 13045392.0,
      "step": 16365
    },
    {
      "entropy": 4.565517330169678,
      "epoch": 4.628216002261803,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031208611262729914,
      "loss": 4.3171,
      "mean_token_accuracy": 0.27611341774463655,
      "num_tokens": 13049197.0,
      "step": 16370
    },
    {
      "entropy": 4.52693510055542,
      "epoch": 4.62962962962963,
      "grad_norm": 1.71875,
      "learning_rate": 0.00031198465499129456,
      "loss": 4.3875,
      "mean_token_accuracy": 0.26303531527519225,
      "num_tokens": 13052992.0,
      "step": 16375
    },
    {
      "entropy": 4.557310342788696,
      "epoch": 4.631043256997455,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003118831896257259,
      "loss": 4.3291,
      "mean_token_accuracy": 0.2821709766983986,
      "num_tokens": 13056877.0,
      "step": 16380
    },
    {
      "entropy": 4.6307065963745115,
      "epoch": 4.632456884365281,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0003117817165517987,
      "loss": 4.3852,
      "mean_token_accuracy": 0.26940789371728896,
      "num_tokens": 13061108.0,
      "step": 16385
    },
    {
      "entropy": 4.685773658752441,
      "epoch": 4.633870511733107,
      "grad_norm": 1.484375,
      "learning_rate": 0.0003116802357907201,
      "loss": 4.342,
      "mean_token_accuracy": 0.27113235294818877,
      "num_tokens": 13065272.0,
      "step": 16390
    },
    {
      "entropy": 4.7389908790588375,
      "epoch": 4.635284139100933,
      "grad_norm": 1.625,
      "learning_rate": 0.00031157874736369916,
      "loss": 4.426,
      "mean_token_accuracy": 0.2559656694531441,
      "num_tokens": 13069151.0,
      "step": 16395
    },
    {
      "entropy": 4.822448682785034,
      "epoch": 4.6366977664687585,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003114772512919462,
      "loss": 4.4699,
      "mean_token_accuracy": 0.24808957129716874,
      "num_tokens": 13073362.0,
      "step": 16400
    },
    {
      "entropy": 4.712142276763916,
      "epoch": 4.638111393836585,
      "grad_norm": 1.765625,
      "learning_rate": 0.00031137574759667317,
      "loss": 4.4009,
      "mean_token_accuracy": 0.2586738631129265,
      "num_tokens": 13076987.0,
      "step": 16405
    },
    {
      "entropy": 4.5596373081207275,
      "epoch": 4.639525021204411,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00031127423629909395,
      "loss": 4.3327,
      "mean_token_accuracy": 0.26518011689186094,
      "num_tokens": 13081123.0,
      "step": 16410
    },
    {
      "entropy": 4.602100038528443,
      "epoch": 4.640938648572236,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00031117271742042365,
      "loss": 4.4043,
      "mean_token_accuracy": 0.26046515852212904,
      "num_tokens": 13085466.0,
      "step": 16415
    },
    {
      "entropy": 4.690690612792968,
      "epoch": 4.642352275940063,
      "grad_norm": 1.7265625,
      "learning_rate": 0.000311071190981879,
      "loss": 4.4482,
      "mean_token_accuracy": 0.2576863572001457,
      "num_tokens": 13089318.0,
      "step": 16420
    },
    {
      "entropy": 4.704412508010864,
      "epoch": 4.643765903307888,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00031096965700467855,
      "loss": 4.4149,
      "mean_token_accuracy": 0.2666538029909134,
      "num_tokens": 13093197.0,
      "step": 16425
    },
    {
      "entropy": 4.716766357421875,
      "epoch": 4.645179530675714,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00031086811551004224,
      "loss": 4.4097,
      "mean_token_accuracy": 0.26606289595365523,
      "num_tokens": 13097269.0,
      "step": 16430
    },
    {
      "entropy": 4.707798147201538,
      "epoch": 4.646593158043539,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0003107665665191916,
      "loss": 4.456,
      "mean_token_accuracy": 0.25798822194337845,
      "num_tokens": 13101457.0,
      "step": 16435
    },
    {
      "entropy": 4.726184415817261,
      "epoch": 4.648006785411366,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0003106650100533497,
      "loss": 4.5556,
      "mean_token_accuracy": 0.2443240076303482,
      "num_tokens": 13105649.0,
      "step": 16440
    },
    {
      "entropy": 4.665063142776489,
      "epoch": 4.649420412779191,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003105634461337412,
      "loss": 4.3334,
      "mean_token_accuracy": 0.2709833011031151,
      "num_tokens": 13109627.0,
      "step": 16445
    },
    {
      "entropy": 4.663022422790528,
      "epoch": 4.650834040147017,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003104618747815925,
      "loss": 4.3775,
      "mean_token_accuracy": 0.27554216980934143,
      "num_tokens": 13113688.0,
      "step": 16450
    },
    {
      "entropy": 4.613866996765137,
      "epoch": 4.652247667514843,
      "grad_norm": 1.671875,
      "learning_rate": 0.00031036029601813115,
      "loss": 4.376,
      "mean_token_accuracy": 0.2683763220906258,
      "num_tokens": 13117801.0,
      "step": 16455
    },
    {
      "entropy": 4.6529100894927975,
      "epoch": 4.653661294882669,
      "grad_norm": 1.65625,
      "learning_rate": 0.00031025870986458665,
      "loss": 4.3961,
      "mean_token_accuracy": 0.26831570267677307,
      "num_tokens": 13121902.0,
      "step": 16460
    },
    {
      "entropy": 4.76927638053894,
      "epoch": 4.655074922250495,
      "grad_norm": 1.75,
      "learning_rate": 0.00031015711634218977,
      "loss": 4.5214,
      "mean_token_accuracy": 0.2510710388422012,
      "num_tokens": 13126032.0,
      "step": 16465
    },
    {
      "entropy": 4.568165445327759,
      "epoch": 4.65648854961832,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00031005551547217303,
      "loss": 4.297,
      "mean_token_accuracy": 0.27141509354114535,
      "num_tokens": 13129929.0,
      "step": 16470
    },
    {
      "entropy": 4.593163347244262,
      "epoch": 4.657902176986147,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00030995390727577026,
      "loss": 4.3234,
      "mean_token_accuracy": 0.2743699744343758,
      "num_tokens": 13134307.0,
      "step": 16475
    },
    {
      "entropy": 4.616500043869019,
      "epoch": 4.659315804353972,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003098522917742171,
      "loss": 4.3762,
      "mean_token_accuracy": 0.26523832976818085,
      "num_tokens": 13138387.0,
      "step": 16480
    },
    {
      "entropy": 4.611096954345703,
      "epoch": 4.660729431721798,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00030975066898875043,
      "loss": 4.2903,
      "mean_token_accuracy": 0.27117238342761996,
      "num_tokens": 13142579.0,
      "step": 16485
    },
    {
      "entropy": 4.6581139087677,
      "epoch": 4.662143059089624,
      "grad_norm": 1.640625,
      "learning_rate": 0.0003096490389406089,
      "loss": 4.4025,
      "mean_token_accuracy": 0.2633488267660141,
      "num_tokens": 13146608.0,
      "step": 16490
    },
    {
      "entropy": 4.6335601806640625,
      "epoch": 4.66355668645745,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00030954740165103244,
      "loss": 4.3629,
      "mean_token_accuracy": 0.26636006981134414,
      "num_tokens": 13150628.0,
      "step": 16495
    },
    {
      "entropy": 4.621443843841552,
      "epoch": 4.664970313825275,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00030944575714126265,
      "loss": 4.3801,
      "mean_token_accuracy": 0.26223423331975937,
      "num_tokens": 13154576.0,
      "step": 16500
    },
    {
      "entropy": 4.6497985363006595,
      "epoch": 4.666383941193102,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003093441054325426,
      "loss": 4.4229,
      "mean_token_accuracy": 0.2720169350504875,
      "num_tokens": 13158342.0,
      "step": 16505
    },
    {
      "entropy": 4.710640478134155,
      "epoch": 4.667797568560927,
      "grad_norm": 1.71875,
      "learning_rate": 0.00030924244654611697,
      "loss": 4.4189,
      "mean_token_accuracy": 0.25889414846897124,
      "num_tokens": 13162395.0,
      "step": 16510
    },
    {
      "entropy": 4.622379446029663,
      "epoch": 4.669211195928753,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0003091407805032317,
      "loss": 4.4218,
      "mean_token_accuracy": 0.2684641107916832,
      "num_tokens": 13166662.0,
      "step": 16515
    },
    {
      "entropy": 4.695582580566406,
      "epoch": 4.6706248232965795,
      "grad_norm": 1.625,
      "learning_rate": 0.0003090391073251345,
      "loss": 4.4029,
      "mean_token_accuracy": 0.25636703222990037,
      "num_tokens": 13170496.0,
      "step": 16520
    },
    {
      "entropy": 4.637559509277343,
      "epoch": 4.672038450664405,
      "grad_norm": 1.765625,
      "learning_rate": 0.00030893742703307444,
      "loss": 4.3485,
      "mean_token_accuracy": 0.27687677890062334,
      "num_tokens": 13174298.0,
      "step": 16525
    },
    {
      "entropy": 4.624889612197876,
      "epoch": 4.673452078032231,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003088357396483018,
      "loss": 4.4211,
      "mean_token_accuracy": 0.26233103424310683,
      "num_tokens": 13178513.0,
      "step": 16530
    },
    {
      "entropy": 4.663642835617066,
      "epoch": 4.674865705400056,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00030873404519206885,
      "loss": 4.4107,
      "mean_token_accuracy": 0.25719706416130067,
      "num_tokens": 13182332.0,
      "step": 16535
    },
    {
      "entropy": 4.658498287200928,
      "epoch": 4.676279332767883,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003086323436856291,
      "loss": 4.3756,
      "mean_token_accuracy": 0.27469484955072404,
      "num_tokens": 13186286.0,
      "step": 16540
    },
    {
      "entropy": 4.647323513031006,
      "epoch": 4.677692960135708,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003085306351502375,
      "loss": 4.3523,
      "mean_token_accuracy": 0.2658663034439087,
      "num_tokens": 13189956.0,
      "step": 16545
    },
    {
      "entropy": 4.506886959075928,
      "epoch": 4.679106587503534,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003084289196071505,
      "loss": 4.1775,
      "mean_token_accuracy": 0.2951959252357483,
      "num_tokens": 13193733.0,
      "step": 16550
    },
    {
      "entropy": 4.6298808574676515,
      "epoch": 4.68052021487136,
      "grad_norm": 1.703125,
      "learning_rate": 0.00030832719707762603,
      "loss": 4.4636,
      "mean_token_accuracy": 0.2527745798230171,
      "num_tokens": 13197887.0,
      "step": 16555
    },
    {
      "entropy": 4.7162621974945065,
      "epoch": 4.681933842239186,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00030822546758292336,
      "loss": 4.3998,
      "mean_token_accuracy": 0.2583673134446144,
      "num_tokens": 13201413.0,
      "step": 16560
    },
    {
      "entropy": 4.637453603744507,
      "epoch": 4.683347469607011,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003081237311443035,
      "loss": 4.3894,
      "mean_token_accuracy": 0.26616528779268267,
      "num_tokens": 13205318.0,
      "step": 16565
    },
    {
      "entropy": 4.66531023979187,
      "epoch": 4.684761096974837,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003080219877830287,
      "loss": 4.3159,
      "mean_token_accuracy": 0.2794434025883675,
      "num_tokens": 13209199.0,
      "step": 16570
    },
    {
      "entropy": 4.486631393432617,
      "epoch": 4.6861747243426635,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00030792023752036264,
      "loss": 4.152,
      "mean_token_accuracy": 0.2830336093902588,
      "num_tokens": 13213101.0,
      "step": 16575
    },
    {
      "entropy": 4.640239953994751,
      "epoch": 4.687588351710489,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00030781848037757036,
      "loss": 4.4216,
      "mean_token_accuracy": 0.25933948159217834,
      "num_tokens": 13216879.0,
      "step": 16580
    },
    {
      "entropy": 4.545825099945068,
      "epoch": 4.689001979078315,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003077167163759188,
      "loss": 4.2134,
      "mean_token_accuracy": 0.2825061440467834,
      "num_tokens": 13220788.0,
      "step": 16585
    },
    {
      "entropy": 4.638560485839844,
      "epoch": 4.690415606446141,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0003076149455366758,
      "loss": 4.521,
      "mean_token_accuracy": 0.25922065824270246,
      "num_tokens": 13224929.0,
      "step": 16590
    },
    {
      "entropy": 4.668572807312012,
      "epoch": 4.691829233813967,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0003075131678811109,
      "loss": 4.3725,
      "mean_token_accuracy": 0.2713120341300964,
      "num_tokens": 13228967.0,
      "step": 16595
    },
    {
      "entropy": 4.733926391601562,
      "epoch": 4.693242861181792,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003074113834304949,
      "loss": 4.3369,
      "mean_token_accuracy": 0.264599597454071,
      "num_tokens": 13233057.0,
      "step": 16600
    },
    {
      "entropy": 4.550281763076782,
      "epoch": 4.694656488549619,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003073095922061002,
      "loss": 4.2449,
      "mean_token_accuracy": 0.27882263362407683,
      "num_tokens": 13236816.0,
      "step": 16605
    },
    {
      "entropy": 4.619332504272461,
      "epoch": 4.696070115917444,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00030720779422920046,
      "loss": 4.4456,
      "mean_token_accuracy": 0.24621639102697374,
      "num_tokens": 13240793.0,
      "step": 16610
    },
    {
      "entropy": 4.738232755661011,
      "epoch": 4.69748374328527,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00030710598952107105,
      "loss": 4.5175,
      "mean_token_accuracy": 0.2620479166507721,
      "num_tokens": 13244892.0,
      "step": 16615
    },
    {
      "entropy": 4.70944619178772,
      "epoch": 4.698897370653096,
      "grad_norm": 1.578125,
      "learning_rate": 0.0003070041781029883,
      "loss": 4.3788,
      "mean_token_accuracy": 0.27586438208818437,
      "num_tokens": 13248603.0,
      "step": 16620
    },
    {
      "entropy": 4.623185300827027,
      "epoch": 4.700310998020922,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00030690235999623016,
      "loss": 4.3253,
      "mean_token_accuracy": 0.28567795604467394,
      "num_tokens": 13252533.0,
      "step": 16625
    },
    {
      "entropy": 4.695318269729614,
      "epoch": 4.7017246253887475,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00030680053522207616,
      "loss": 4.4229,
      "mean_token_accuracy": 0.257570119202137,
      "num_tokens": 13256847.0,
      "step": 16630
    },
    {
      "entropy": 4.7637194156646725,
      "epoch": 4.703138252756573,
      "grad_norm": 1.609375,
      "learning_rate": 0.0003066987038018069,
      "loss": 4.4859,
      "mean_token_accuracy": 0.2549035683274269,
      "num_tokens": 13261220.0,
      "step": 16635
    },
    {
      "entropy": 4.696369266510009,
      "epoch": 4.7045518801243995,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003065968657567045,
      "loss": 4.3443,
      "mean_token_accuracy": 0.2667214125394821,
      "num_tokens": 13264653.0,
      "step": 16640
    },
    {
      "entropy": 4.705402374267578,
      "epoch": 4.705965507492225,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00030649502110805264,
      "loss": 4.3704,
      "mean_token_accuracy": 0.25567996948957444,
      "num_tokens": 13268755.0,
      "step": 16645
    },
    {
      "entropy": 4.589551019668579,
      "epoch": 4.707379134860051,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00030639316987713614,
      "loss": 4.2755,
      "mean_token_accuracy": 0.27202213257551194,
      "num_tokens": 13272857.0,
      "step": 16650
    },
    {
      "entropy": 4.626046371459961,
      "epoch": 4.708792762227876,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00030629131208524126,
      "loss": 4.3344,
      "mean_token_accuracy": 0.2753797575831413,
      "num_tokens": 13276852.0,
      "step": 16655
    },
    {
      "entropy": 4.714873361587524,
      "epoch": 4.710206389595703,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003061894477536556,
      "loss": 4.3942,
      "mean_token_accuracy": 0.2572348952293396,
      "num_tokens": 13280767.0,
      "step": 16660
    },
    {
      "entropy": 4.646268033981324,
      "epoch": 4.711620016963528,
      "grad_norm": 1.5625,
      "learning_rate": 0.0003060875769036683,
      "loss": 4.3144,
      "mean_token_accuracy": 0.28072679191827776,
      "num_tokens": 13284572.0,
      "step": 16665
    },
    {
      "entropy": 4.693643093109131,
      "epoch": 4.713033644331354,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003059856995565695,
      "loss": 4.4601,
      "mean_token_accuracy": 0.25954222977161406,
      "num_tokens": 13288500.0,
      "step": 16670
    },
    {
      "entropy": 4.687622356414795,
      "epoch": 4.71444727169918,
      "grad_norm": 1.625,
      "learning_rate": 0.00030588381573365127,
      "loss": 4.4406,
      "mean_token_accuracy": 0.2575272262096405,
      "num_tokens": 13292503.0,
      "step": 16675
    },
    {
      "entropy": 4.68555965423584,
      "epoch": 4.715860899067006,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003057819254562065,
      "loss": 4.3564,
      "mean_token_accuracy": 0.26468232125043867,
      "num_tokens": 13296530.0,
      "step": 16680
    },
    {
      "entropy": 4.591109943389893,
      "epoch": 4.7172745264348315,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003056800287455296,
      "loss": 4.3021,
      "mean_token_accuracy": 0.2733533412218094,
      "num_tokens": 13300472.0,
      "step": 16685
    },
    {
      "entropy": 4.595946598052978,
      "epoch": 4.718688153802658,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00030557812562291645,
      "loss": 4.2145,
      "mean_token_accuracy": 0.279855777323246,
      "num_tokens": 13304673.0,
      "step": 16690
    },
    {
      "entropy": 4.581054210662842,
      "epoch": 4.7201017811704835,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003054762161096642,
      "loss": 4.27,
      "mean_token_accuracy": 0.2727120116353035,
      "num_tokens": 13308561.0,
      "step": 16695
    },
    {
      "entropy": 4.614017724990845,
      "epoch": 4.721515408538309,
      "grad_norm": 1.5,
      "learning_rate": 0.00030537430022707114,
      "loss": 4.2915,
      "mean_token_accuracy": 0.27828367203474047,
      "num_tokens": 13312439.0,
      "step": 16700
    },
    {
      "entropy": 4.541008949279785,
      "epoch": 4.7229290359061356,
      "grad_norm": 1.671875,
      "learning_rate": 0.00030527237799643725,
      "loss": 4.3066,
      "mean_token_accuracy": 0.2695333167910576,
      "num_tokens": 13316184.0,
      "step": 16705
    },
    {
      "entropy": 4.607213735580444,
      "epoch": 4.724342663273961,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00030517044943906345,
      "loss": 4.2861,
      "mean_token_accuracy": 0.281087663769722,
      "num_tokens": 13319822.0,
      "step": 16710
    },
    {
      "entropy": 4.750605726242066,
      "epoch": 4.725756290641787,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003050685145762524,
      "loss": 4.4898,
      "mean_token_accuracy": 0.2469516932964325,
      "num_tokens": 13324019.0,
      "step": 16715
    },
    {
      "entropy": 4.714883327484131,
      "epoch": 4.727169918009612,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00030496657342930767,
      "loss": 4.3959,
      "mean_token_accuracy": 0.25984962433576586,
      "num_tokens": 13328003.0,
      "step": 16720
    },
    {
      "entropy": 4.674465179443359,
      "epoch": 4.728583545377439,
      "grad_norm": 1.71875,
      "learning_rate": 0.0003048646260195344,
      "loss": 4.344,
      "mean_token_accuracy": 0.27217655181884765,
      "num_tokens": 13331971.0,
      "step": 16725
    },
    {
      "entropy": 4.6249918937683105,
      "epoch": 4.729997172745264,
      "grad_norm": 1.6796875,
      "learning_rate": 0.000304762672368239,
      "loss": 4.3808,
      "mean_token_accuracy": 0.2677623823285103,
      "num_tokens": 13336064.0,
      "step": 16730
    },
    {
      "entropy": 4.71162166595459,
      "epoch": 4.73141080011309,
      "grad_norm": 1.6875,
      "learning_rate": 0.0003046607124967291,
      "loss": 4.4356,
      "mean_token_accuracy": 0.2485654816031456,
      "num_tokens": 13339572.0,
      "step": 16735
    },
    {
      "entropy": 4.6694354057312015,
      "epoch": 4.732824427480916,
      "grad_norm": 1.546875,
      "learning_rate": 0.0003045587464263137,
      "loss": 4.3952,
      "mean_token_accuracy": 0.2612585097551346,
      "num_tokens": 13343619.0,
      "step": 16740
    },
    {
      "entropy": 4.624593305587768,
      "epoch": 4.734238054848742,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00030445677417830316,
      "loss": 4.4143,
      "mean_token_accuracy": 0.2666842833161354,
      "num_tokens": 13348005.0,
      "step": 16745
    },
    {
      "entropy": 4.64473819732666,
      "epoch": 4.7356516822165675,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00030435479577400887,
      "loss": 4.3924,
      "mean_token_accuracy": 0.261576221883297,
      "num_tokens": 13351935.0,
      "step": 16750
    },
    {
      "entropy": 4.625226879119873,
      "epoch": 4.737065309584393,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0003042528112347438,
      "loss": 4.3511,
      "mean_token_accuracy": 0.2735592395067215,
      "num_tokens": 13355835.0,
      "step": 16755
    },
    {
      "entropy": 4.701547002792358,
      "epoch": 4.7384789369522196,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00030415082058182216,
      "loss": 4.4127,
      "mean_token_accuracy": 0.25295298397541044,
      "num_tokens": 13359766.0,
      "step": 16760
    },
    {
      "entropy": 4.638183116912842,
      "epoch": 4.739892564320045,
      "grad_norm": 1.625,
      "learning_rate": 0.0003040488238365592,
      "loss": 4.3507,
      "mean_token_accuracy": 0.26722771525382993,
      "num_tokens": 13363978.0,
      "step": 16765
    },
    {
      "entropy": 4.666762113571167,
      "epoch": 4.741306191687871,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0003039468210202717,
      "loss": 4.2973,
      "mean_token_accuracy": 0.27501087784767153,
      "num_tokens": 13368138.0,
      "step": 16770
    },
    {
      "entropy": 4.575671911239624,
      "epoch": 4.742719819055697,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003038448121542777,
      "loss": 4.3835,
      "mean_token_accuracy": 0.2782520577311516,
      "num_tokens": 13372245.0,
      "step": 16775
    },
    {
      "entropy": 4.551513957977295,
      "epoch": 4.744133446423523,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003037427972598962,
      "loss": 4.279,
      "mean_token_accuracy": 0.27850799560546874,
      "num_tokens": 13376285.0,
      "step": 16780
    },
    {
      "entropy": 4.642720937728882,
      "epoch": 4.745547073791348,
      "grad_norm": 1.609375,
      "learning_rate": 0.00030364077635844784,
      "loss": 4.4213,
      "mean_token_accuracy": 0.26817235052585603,
      "num_tokens": 13380287.0,
      "step": 16785
    },
    {
      "entropy": 4.696526336669922,
      "epoch": 4.746960701159175,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003035387494712543,
      "loss": 4.353,
      "mean_token_accuracy": 0.2682086542248726,
      "num_tokens": 13384485.0,
      "step": 16790
    },
    {
      "entropy": 4.7148560047149655,
      "epoch": 4.748374328527,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00030343671661963866,
      "loss": 4.385,
      "mean_token_accuracy": 0.2661318451166153,
      "num_tokens": 13388274.0,
      "step": 16795
    },
    {
      "entropy": 4.707017803192139,
      "epoch": 4.749787955894826,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0003033346778249252,
      "loss": 4.4663,
      "mean_token_accuracy": 0.2573892965912819,
      "num_tokens": 13391918.0,
      "step": 16800
    },
    {
      "entropy": 4.6423011302948,
      "epoch": 4.751201583262652,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003032326331084391,
      "loss": 4.3055,
      "mean_token_accuracy": 0.2778298482298851,
      "num_tokens": 13395972.0,
      "step": 16805
    },
    {
      "entropy": 4.594090270996094,
      "epoch": 4.752615210630478,
      "grad_norm": 1.65625,
      "learning_rate": 0.00030313058249150743,
      "loss": 4.3175,
      "mean_token_accuracy": 0.2710348069667816,
      "num_tokens": 13400038.0,
      "step": 16810
    },
    {
      "entropy": 4.648648262023926,
      "epoch": 4.754028837998304,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0003030285259954578,
      "loss": 4.4171,
      "mean_token_accuracy": 0.2581301912665367,
      "num_tokens": 13403993.0,
      "step": 16815
    },
    {
      "entropy": 4.638686990737915,
      "epoch": 4.755442465366129,
      "grad_norm": 1.78125,
      "learning_rate": 0.0003029264636416197,
      "loss": 4.3352,
      "mean_token_accuracy": 0.27248377799987794,
      "num_tokens": 13407673.0,
      "step": 16820
    },
    {
      "entropy": 4.620300769805908,
      "epoch": 4.756856092733956,
      "grad_norm": 1.640625,
      "learning_rate": 0.00030282439545132337,
      "loss": 4.1973,
      "mean_token_accuracy": 0.27797997891902926,
      "num_tokens": 13411725.0,
      "step": 16825
    },
    {
      "entropy": 4.644570064544678,
      "epoch": 4.758269720101781,
      "grad_norm": 1.59375,
      "learning_rate": 0.00030272232144590034,
      "loss": 4.4304,
      "mean_token_accuracy": 0.2662405177950859,
      "num_tokens": 13415895.0,
      "step": 16830
    },
    {
      "entropy": 4.675257396697998,
      "epoch": 4.759683347469607,
      "grad_norm": 1.8125,
      "learning_rate": 0.00030262024164668365,
      "loss": 4.3247,
      "mean_token_accuracy": 0.27960584312677383,
      "num_tokens": 13419767.0,
      "step": 16835
    },
    {
      "entropy": 4.704999923706055,
      "epoch": 4.761096974837433,
      "grad_norm": 1.609375,
      "learning_rate": 0.00030251815607500724,
      "loss": 4.4059,
      "mean_token_accuracy": 0.26073853522539137,
      "num_tokens": 13423693.0,
      "step": 16840
    },
    {
      "entropy": 4.681882238388061,
      "epoch": 4.762510602205259,
      "grad_norm": 1.671875,
      "learning_rate": 0.0003024160647522064,
      "loss": 4.3476,
      "mean_token_accuracy": 0.27402387708425524,
      "num_tokens": 13427690.0,
      "step": 16845
    },
    {
      "entropy": 4.5993781089782715,
      "epoch": 4.763924229573084,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00030231396769961756,
      "loss": 4.4103,
      "mean_token_accuracy": 0.2672385036945343,
      "num_tokens": 13431914.0,
      "step": 16850
    },
    {
      "entropy": 4.699969816207886,
      "epoch": 4.76533785694091,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0003022118649385783,
      "loss": 4.3888,
      "mean_token_accuracy": 0.2577453449368477,
      "num_tokens": 13436131.0,
      "step": 16855
    },
    {
      "entropy": 4.6269848346710205,
      "epoch": 4.766751484308736,
      "grad_norm": 1.625,
      "learning_rate": 0.00030210975649042765,
      "loss": 4.3818,
      "mean_token_accuracy": 0.27208405882120135,
      "num_tokens": 13440138.0,
      "step": 16860
    },
    {
      "entropy": 4.660031414031982,
      "epoch": 4.768165111676562,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00030200764237650553,
      "loss": 4.4074,
      "mean_token_accuracy": 0.27191009521484377,
      "num_tokens": 13444352.0,
      "step": 16865
    },
    {
      "entropy": 4.604122257232666,
      "epoch": 4.769578739044388,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0003019055226181531,
      "loss": 4.2566,
      "mean_token_accuracy": 0.2807096317410469,
      "num_tokens": 13448243.0,
      "step": 16870
    },
    {
      "entropy": 4.6470331192016605,
      "epoch": 4.770992366412214,
      "grad_norm": 1.546875,
      "learning_rate": 0.00030180339723671277,
      "loss": 4.3565,
      "mean_token_accuracy": 0.26498320400714875,
      "num_tokens": 13452435.0,
      "step": 16875
    },
    {
      "entropy": 4.627344369888306,
      "epoch": 4.77240599378004,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00030170126625352826,
      "loss": 4.4595,
      "mean_token_accuracy": 0.2579712450504303,
      "num_tokens": 13456482.0,
      "step": 16880
    },
    {
      "entropy": 4.587978267669678,
      "epoch": 4.773819621147865,
      "grad_norm": 1.65625,
      "learning_rate": 0.00030159912968994416,
      "loss": 4.3144,
      "mean_token_accuracy": 0.27625485211610795,
      "num_tokens": 13460232.0,
      "step": 16885
    },
    {
      "entropy": 4.6500543594360355,
      "epoch": 4.775233248515692,
      "grad_norm": 1.6875,
      "learning_rate": 0.00030149698756730647,
      "loss": 4.3968,
      "mean_token_accuracy": 0.2669374644756317,
      "num_tokens": 13464056.0,
      "step": 16890
    },
    {
      "entropy": 4.651838684082032,
      "epoch": 4.776646875883517,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00030139483990696225,
      "loss": 4.3552,
      "mean_token_accuracy": 0.2722385570406914,
      "num_tokens": 13467923.0,
      "step": 16895
    },
    {
      "entropy": 4.719180345535278,
      "epoch": 4.778060503251343,
      "grad_norm": 1.734375,
      "learning_rate": 0.0003012926867302597,
      "loss": 4.398,
      "mean_token_accuracy": 0.26541535556316376,
      "num_tokens": 13471742.0,
      "step": 16900
    },
    {
      "entropy": 4.648122930526734,
      "epoch": 4.779474130619169,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0003011905280585482,
      "loss": 4.4128,
      "mean_token_accuracy": 0.26626057624816896,
      "num_tokens": 13475679.0,
      "step": 16905
    },
    {
      "entropy": 4.565210247039795,
      "epoch": 4.780887757986995,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0003010883639131783,
      "loss": 4.2584,
      "mean_token_accuracy": 0.26753906458616256,
      "num_tokens": 13479655.0,
      "step": 16910
    },
    {
      "entropy": 4.62536129951477,
      "epoch": 4.78230138535482,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00030098619431550167,
      "loss": 4.4438,
      "mean_token_accuracy": 0.2520436316728592,
      "num_tokens": 13483533.0,
      "step": 16915
    },
    {
      "entropy": 4.530517578125,
      "epoch": 4.783715012722646,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00030088401928687116,
      "loss": 4.3694,
      "mean_token_accuracy": 0.26120499670505526,
      "num_tokens": 13487404.0,
      "step": 16920
    },
    {
      "entropy": 4.51376166343689,
      "epoch": 4.7851286400904725,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00030078183884864063,
      "loss": 4.3442,
      "mean_token_accuracy": 0.2666648641228676,
      "num_tokens": 13491081.0,
      "step": 16925
    },
    {
      "entropy": 4.650524568557739,
      "epoch": 4.786542267458298,
      "grad_norm": 1.609375,
      "learning_rate": 0.00030067965302216524,
      "loss": 4.3306,
      "mean_token_accuracy": 0.2745926812291145,
      "num_tokens": 13495356.0,
      "step": 16930
    },
    {
      "entropy": 4.622806024551392,
      "epoch": 4.787955894826124,
      "grad_norm": 1.640625,
      "learning_rate": 0.00030057746182880106,
      "loss": 4.4253,
      "mean_token_accuracy": 0.2596185848116875,
      "num_tokens": 13499116.0,
      "step": 16935
    },
    {
      "entropy": 4.677014780044556,
      "epoch": 4.789369522193949,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00030047526528990563,
      "loss": 4.3532,
      "mean_token_accuracy": 0.26344441473484037,
      "num_tokens": 13502745.0,
      "step": 16940
    },
    {
      "entropy": 4.62804946899414,
      "epoch": 4.790783149561776,
      "grad_norm": 1.703125,
      "learning_rate": 0.0003003730634268373,
      "loss": 4.3767,
      "mean_token_accuracy": 0.2720178559422493,
      "num_tokens": 13506541.0,
      "step": 16945
    },
    {
      "entropy": 4.636971139907837,
      "epoch": 4.792196776929601,
      "grad_norm": 1.65625,
      "learning_rate": 0.0003002708562609556,
      "loss": 4.3372,
      "mean_token_accuracy": 0.26888758689165115,
      "num_tokens": 13510701.0,
      "step": 16950
    },
    {
      "entropy": 4.633968162536621,
      "epoch": 4.793610404297427,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00030016864381362125,
      "loss": 4.3616,
      "mean_token_accuracy": 0.2725736141204834,
      "num_tokens": 13514989.0,
      "step": 16955
    },
    {
      "entropy": 4.590221118927002,
      "epoch": 4.795024031665253,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00030006642610619594,
      "loss": 4.3083,
      "mean_token_accuracy": 0.27587707340717316,
      "num_tokens": 13519088.0,
      "step": 16960
    },
    {
      "entropy": 4.616307878494263,
      "epoch": 4.796437659033079,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0002999642031600425,
      "loss": 4.3884,
      "mean_token_accuracy": 0.2654386192560196,
      "num_tokens": 13522992.0,
      "step": 16965
    },
    {
      "entropy": 4.635094404220581,
      "epoch": 4.797851286400904,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002998619749965251,
      "loss": 4.3346,
      "mean_token_accuracy": 0.25688099265098574,
      "num_tokens": 13527082.0,
      "step": 16970
    },
    {
      "entropy": 4.6677570819854735,
      "epoch": 4.799264913768731,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002997597416370085,
      "loss": 4.3224,
      "mean_token_accuracy": 0.26733817160129547,
      "num_tokens": 13530941.0,
      "step": 16975
    },
    {
      "entropy": 4.623213434219361,
      "epoch": 4.8006785411365565,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002996575031028592,
      "loss": 4.3654,
      "mean_token_accuracy": 0.2699998363852501,
      "num_tokens": 13535352.0,
      "step": 16980
    },
    {
      "entropy": 4.6293353080749515,
      "epoch": 4.802092168504382,
      "grad_norm": 1.671875,
      "learning_rate": 0.00029955525941544416,
      "loss": 4.3442,
      "mean_token_accuracy": 0.2755414456129074,
      "num_tokens": 13539339.0,
      "step": 16985
    },
    {
      "entropy": 4.6356117725372314,
      "epoch": 4.8035057958722085,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002994530105961317,
      "loss": 4.412,
      "mean_token_accuracy": 0.2556716799736023,
      "num_tokens": 13543216.0,
      "step": 16990
    },
    {
      "entropy": 4.737662744522095,
      "epoch": 4.804919423240034,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002993507566662913,
      "loss": 4.3467,
      "mean_token_accuracy": 0.27066956013441085,
      "num_tokens": 13547125.0,
      "step": 16995
    },
    {
      "entropy": 4.74843807220459,
      "epoch": 4.80633305060786,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00029924849764729333,
      "loss": 4.3357,
      "mean_token_accuracy": 0.2747945189476013,
      "num_tokens": 13551522.0,
      "step": 17000
    },
    {
      "epoch": 4.80633305060786,
      "eval_entropy": 4.507879226280391,
      "eval_loss": 4.984694957733154,
      "eval_mean_token_accuracy": 0.24906854301395356,
      "eval_num_tokens": 13551522.0,
      "eval_runtime": 6.6278,
      "eval_samples_per_second": 1702.53,
      "eval_steps_per_second": 212.892,
      "step": 17000
    },
    {
      "entropy": 4.662420845031738,
      "epoch": 4.807746677975685,
      "grad_norm": 1.90625,
      "learning_rate": 0.0002991462335605092,
      "loss": 4.4423,
      "mean_token_accuracy": 0.26048055440187456,
      "num_tokens": 13555145.0,
      "step": 17005
    },
    {
      "entropy": 4.653487014770508,
      "epoch": 4.809160305343512,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002990439644273117,
      "loss": 4.4598,
      "mean_token_accuracy": 0.24753400087356567,
      "num_tokens": 13559521.0,
      "step": 17010
    },
    {
      "entropy": 4.5827367305755615,
      "epoch": 4.810573932711337,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002989416902690743,
      "loss": 4.3277,
      "mean_token_accuracy": 0.2716978803277016,
      "num_tokens": 13563289.0,
      "step": 17015
    },
    {
      "entropy": 4.647923231124878,
      "epoch": 4.811987560079163,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002988394111071716,
      "loss": 4.3018,
      "mean_token_accuracy": 0.27684334218502044,
      "num_tokens": 13567316.0,
      "step": 17020
    },
    {
      "entropy": 4.758372116088867,
      "epoch": 4.813401187446989,
      "grad_norm": 1.515625,
      "learning_rate": 0.0002987371269629795,
      "loss": 4.4104,
      "mean_token_accuracy": 0.26345213055610656,
      "num_tokens": 13571251.0,
      "step": 17025
    },
    {
      "entropy": 4.582005977630615,
      "epoch": 4.814814814814815,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00029863483785787454,
      "loss": 4.2813,
      "mean_token_accuracy": 0.2687725484371185,
      "num_tokens": 13575278.0,
      "step": 17030
    },
    {
      "entropy": 4.609024906158448,
      "epoch": 4.8162284421826405,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002985325438132347,
      "loss": 4.4383,
      "mean_token_accuracy": 0.26044579446315763,
      "num_tokens": 13579240.0,
      "step": 17035
    },
    {
      "entropy": 4.583941888809204,
      "epoch": 4.817642069550466,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00029843024485043867,
      "loss": 4.3462,
      "mean_token_accuracy": 0.28053483068943025,
      "num_tokens": 13583363.0,
      "step": 17040
    },
    {
      "entropy": 4.570259475708008,
      "epoch": 4.8190556969182925,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0002983279409908664,
      "loss": 4.3257,
      "mean_token_accuracy": 0.2842549428343773,
      "num_tokens": 13587432.0,
      "step": 17045
    },
    {
      "entropy": 4.635831546783447,
      "epoch": 4.820469324286118,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002982256322558988,
      "loss": 4.2755,
      "mean_token_accuracy": 0.2728837087750435,
      "num_tokens": 13591017.0,
      "step": 17050
    },
    {
      "entropy": 4.69610276222229,
      "epoch": 4.821882951653944,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00029812331866691747,
      "loss": 4.2938,
      "mean_token_accuracy": 0.2654881998896599,
      "num_tokens": 13594874.0,
      "step": 17055
    },
    {
      "entropy": 4.6435059070587155,
      "epoch": 4.82329657902177,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002980210002453057,
      "loss": 4.4422,
      "mean_token_accuracy": 0.26374017298221586,
      "num_tokens": 13598839.0,
      "step": 17060
    },
    {
      "entropy": 4.607900667190552,
      "epoch": 4.824710206389596,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002979186770124472,
      "loss": 4.4367,
      "mean_token_accuracy": 0.2626284509897232,
      "num_tokens": 13602953.0,
      "step": 17065
    },
    {
      "entropy": 4.600216484069824,
      "epoch": 4.826123833757421,
      "grad_norm": 1.59375,
      "learning_rate": 0.00029781634898972694,
      "loss": 4.3229,
      "mean_token_accuracy": 0.2784839630126953,
      "num_tokens": 13607081.0,
      "step": 17070
    },
    {
      "entropy": 4.598723554611206,
      "epoch": 4.827537461125248,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002977140161985308,
      "loss": 4.4113,
      "mean_token_accuracy": 0.26113034784793854,
      "num_tokens": 13611242.0,
      "step": 17075
    },
    {
      "entropy": 4.6343833923339846,
      "epoch": 4.828951088493073,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00029761167866024577,
      "loss": 4.3952,
      "mean_token_accuracy": 0.2603229284286499,
      "num_tokens": 13615257.0,
      "step": 17080
    },
    {
      "entropy": 4.731088066101075,
      "epoch": 4.830364715860899,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00029750933639625964,
      "loss": 4.4575,
      "mean_token_accuracy": 0.2524557963013649,
      "num_tokens": 13619517.0,
      "step": 17085
    },
    {
      "entropy": 4.648014497756958,
      "epoch": 4.831778343228725,
      "grad_norm": 1.515625,
      "learning_rate": 0.00029740698942796144,
      "loss": 4.328,
      "mean_token_accuracy": 0.27787732928991316,
      "num_tokens": 13623721.0,
      "step": 17090
    },
    {
      "entropy": 4.633738613128662,
      "epoch": 4.833191970596551,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00029730463777674106,
      "loss": 4.3746,
      "mean_token_accuracy": 0.2690957561135292,
      "num_tokens": 13627599.0,
      "step": 17095
    },
    {
      "entropy": 4.57444281578064,
      "epoch": 4.8346055979643765,
      "grad_norm": 1.65625,
      "learning_rate": 0.00029720228146398926,
      "loss": 4.3321,
      "mean_token_accuracy": 0.2736136496067047,
      "num_tokens": 13631464.0,
      "step": 17100
    },
    {
      "entropy": 4.657678651809692,
      "epoch": 4.836019225332202,
      "grad_norm": 1.703125,
      "learning_rate": 0.00029709992051109795,
      "loss": 4.4258,
      "mean_token_accuracy": 0.26399451643228533,
      "num_tokens": 13635551.0,
      "step": 17105
    },
    {
      "entropy": 4.713652086257935,
      "epoch": 4.8374328527000285,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002969975549394599,
      "loss": 4.4362,
      "mean_token_accuracy": 0.2578902363777161,
      "num_tokens": 13639277.0,
      "step": 17110
    },
    {
      "entropy": 4.6141579151153564,
      "epoch": 4.838846480067854,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002968951847704689,
      "loss": 4.3014,
      "mean_token_accuracy": 0.27058072984218595,
      "num_tokens": 13643528.0,
      "step": 17115
    },
    {
      "entropy": 4.6469684600830075,
      "epoch": 4.84026010743568,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00029679281002551974,
      "loss": 4.3715,
      "mean_token_accuracy": 0.26994224935770034,
      "num_tokens": 13647674.0,
      "step": 17120
    },
    {
      "entropy": 4.650054597854615,
      "epoch": 4.841673734803506,
      "grad_norm": 1.59375,
      "learning_rate": 0.000296690430726008,
      "loss": 4.4184,
      "mean_token_accuracy": 0.2572953328490257,
      "num_tokens": 13651530.0,
      "step": 17125
    },
    {
      "entropy": 4.609944534301758,
      "epoch": 4.843087362171332,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0002965880468933303,
      "loss": 4.4221,
      "mean_token_accuracy": 0.25846058279275896,
      "num_tokens": 13655691.0,
      "step": 17130
    },
    {
      "entropy": 4.6357497215271,
      "epoch": 4.844500989539157,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00029648565854888446,
      "loss": 4.2843,
      "mean_token_accuracy": 0.2830148160457611,
      "num_tokens": 13659622.0,
      "step": 17135
    },
    {
      "entropy": 4.660069227218628,
      "epoch": 4.845914616906983,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002963832657140687,
      "loss": 4.3464,
      "mean_token_accuracy": 0.26215028017759323,
      "num_tokens": 13663622.0,
      "step": 17140
    },
    {
      "entropy": 4.587862491607666,
      "epoch": 4.847328244274809,
      "grad_norm": 1.609375,
      "learning_rate": 0.00029628086841028265,
      "loss": 4.3124,
      "mean_token_accuracy": 0.2739256232976913,
      "num_tokens": 13667879.0,
      "step": 17145
    },
    {
      "entropy": 4.656843137741089,
      "epoch": 4.848741871642635,
      "grad_norm": 1.59375,
      "learning_rate": 0.00029617846665892675,
      "loss": 4.3975,
      "mean_token_accuracy": 0.2642351359128952,
      "num_tokens": 13672040.0,
      "step": 17150
    },
    {
      "entropy": 4.553271532058716,
      "epoch": 4.8501554990104605,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00029607606048140226,
      "loss": 4.2163,
      "mean_token_accuracy": 0.2793445959687233,
      "num_tokens": 13676148.0,
      "step": 17155
    },
    {
      "entropy": 4.611855363845825,
      "epoch": 4.851569126378287,
      "grad_norm": 1.671875,
      "learning_rate": 0.00029597364989911146,
      "loss": 4.3539,
      "mean_token_accuracy": 0.2685681626200676,
      "num_tokens": 13680167.0,
      "step": 17160
    },
    {
      "entropy": 4.700084638595581,
      "epoch": 4.8529827537461125,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002958712349334575,
      "loss": 4.4561,
      "mean_token_accuracy": 0.25947524309158326,
      "num_tokens": 13684073.0,
      "step": 17165
    },
    {
      "entropy": 4.702335500717163,
      "epoch": 4.854396381113938,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00029576881560584454,
      "loss": 4.3717,
      "mean_token_accuracy": 0.26463101208209994,
      "num_tokens": 13688190.0,
      "step": 17170
    },
    {
      "entropy": 4.634279680252075,
      "epoch": 4.855810008481765,
      "grad_norm": 1.703125,
      "learning_rate": 0.00029566639193767756,
      "loss": 4.363,
      "mean_token_accuracy": 0.27274217307567594,
      "num_tokens": 13692241.0,
      "step": 17175
    },
    {
      "entropy": 4.6750610828399655,
      "epoch": 4.85722363584959,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002955639639503624,
      "loss": 4.3832,
      "mean_token_accuracy": 0.2573415517807007,
      "num_tokens": 13696355.0,
      "step": 17180
    },
    {
      "entropy": 4.599871730804443,
      "epoch": 4.858637263217416,
      "grad_norm": 1.640625,
      "learning_rate": 0.00029546153166530603,
      "loss": 4.29,
      "mean_token_accuracy": 0.26142925918102267,
      "num_tokens": 13700494.0,
      "step": 17185
    },
    {
      "entropy": 4.706183433532715,
      "epoch": 4.860050890585242,
      "grad_norm": 1.6953125,
      "learning_rate": 0.000295359095103916,
      "loss": 4.3428,
      "mean_token_accuracy": 0.26500159651041033,
      "num_tokens": 13704346.0,
      "step": 17190
    },
    {
      "entropy": 4.735514497756958,
      "epoch": 4.861464517953068,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000295256654287601,
      "loss": 4.4424,
      "mean_token_accuracy": 0.25283278822898864,
      "num_tokens": 13708284.0,
      "step": 17195
    },
    {
      "entropy": 4.6674792766571045,
      "epoch": 4.862878145320893,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002951542092377706,
      "loss": 4.3532,
      "mean_token_accuracy": 0.26659985035657885,
      "num_tokens": 13712143.0,
      "step": 17200
    },
    {
      "entropy": 4.705288028717041,
      "epoch": 4.864291772688719,
      "grad_norm": 1.515625,
      "learning_rate": 0.00029505175997583503,
      "loss": 4.3877,
      "mean_token_accuracy": 0.26072197407484055,
      "num_tokens": 13716470.0,
      "step": 17205
    },
    {
      "entropy": 4.574871921539307,
      "epoch": 4.865705400056545,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00029494930652320555,
      "loss": 4.2698,
      "mean_token_accuracy": 0.27754402011632917,
      "num_tokens": 13720643.0,
      "step": 17210
    },
    {
      "entropy": 4.699667644500733,
      "epoch": 4.867119027424371,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002948468489012945,
      "loss": 4.3383,
      "mean_token_accuracy": 0.26346936225891116,
      "num_tokens": 13724807.0,
      "step": 17215
    },
    {
      "entropy": 4.609403944015503,
      "epoch": 4.8685326547921965,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002947443871315146,
      "loss": 4.3367,
      "mean_token_accuracy": 0.2640977144241333,
      "num_tokens": 13728642.0,
      "step": 17220
    },
    {
      "entropy": 4.637300205230713,
      "epoch": 4.869946282160022,
      "grad_norm": 1.578125,
      "learning_rate": 0.00029464192123528006,
      "loss": 4.3801,
      "mean_token_accuracy": 0.27004282623529435,
      "num_tokens": 13732723.0,
      "step": 17225
    },
    {
      "entropy": 4.715240383148194,
      "epoch": 4.871359909527849,
      "grad_norm": 1.671875,
      "learning_rate": 0.00029453945123400537,
      "loss": 4.5277,
      "mean_token_accuracy": 0.2524471148848534,
      "num_tokens": 13736695.0,
      "step": 17230
    },
    {
      "entropy": 4.692420387268067,
      "epoch": 4.872773536895674,
      "grad_norm": 1.6875,
      "learning_rate": 0.00029443697714910623,
      "loss": 4.3425,
      "mean_token_accuracy": 0.26700054556131364,
      "num_tokens": 13740646.0,
      "step": 17235
    },
    {
      "entropy": 4.560957670211792,
      "epoch": 4.8741871642635,
      "grad_norm": 1.578125,
      "learning_rate": 0.000294334499001999,
      "loss": 4.3109,
      "mean_token_accuracy": 0.271545498073101,
      "num_tokens": 13744527.0,
      "step": 17240
    },
    {
      "entropy": 4.684900903701783,
      "epoch": 4.875600791631326,
      "grad_norm": 1.609375,
      "learning_rate": 0.00029423201681410107,
      "loss": 4.5023,
      "mean_token_accuracy": 0.25237357169389724,
      "num_tokens": 13748586.0,
      "step": 17245
    },
    {
      "entropy": 4.558684945106506,
      "epoch": 4.877014418999152,
      "grad_norm": 1.484375,
      "learning_rate": 0.0002941295306068305,
      "loss": 4.3404,
      "mean_token_accuracy": 0.2679005518555641,
      "num_tokens": 13752508.0,
      "step": 17250
    },
    {
      "entropy": 4.701856279373169,
      "epoch": 4.878428046366977,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002940270404016064,
      "loss": 4.4135,
      "mean_token_accuracy": 0.2686366319656372,
      "num_tokens": 13756417.0,
      "step": 17255
    },
    {
      "entropy": 4.643270683288574,
      "epoch": 4.879841673734804,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002939245462198485,
      "loss": 4.3566,
      "mean_token_accuracy": 0.26327947378158567,
      "num_tokens": 13760648.0,
      "step": 17260
    },
    {
      "entropy": 4.651649904251099,
      "epoch": 4.881255301102629,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00029382204808297755,
      "loss": 4.294,
      "mean_token_accuracy": 0.26741882264614103,
      "num_tokens": 13764526.0,
      "step": 17265
    },
    {
      "entropy": 4.599102592468261,
      "epoch": 4.882668928470455,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002937195460124148,
      "loss": 4.3628,
      "mean_token_accuracy": 0.2562838762998581,
      "num_tokens": 13767943.0,
      "step": 17270
    },
    {
      "entropy": 4.648179531097412,
      "epoch": 4.884082555838281,
      "grad_norm": 1.796875,
      "learning_rate": 0.0002936170400295827,
      "loss": 4.3315,
      "mean_token_accuracy": 0.2716938093304634,
      "num_tokens": 13772220.0,
      "step": 17275
    },
    {
      "entropy": 4.7043657302856445,
      "epoch": 4.885496183206107,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0002935145301559044,
      "loss": 4.4294,
      "mean_token_accuracy": 0.2519528016448021,
      "num_tokens": 13776029.0,
      "step": 17280
    },
    {
      "entropy": 4.6400152206420895,
      "epoch": 4.886909810573933,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0002934120164128038,
      "loss": 4.4332,
      "mean_token_accuracy": 0.25731891244649885,
      "num_tokens": 13780343.0,
      "step": 17285
    },
    {
      "entropy": 4.660840129852295,
      "epoch": 4.888323437941758,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002933094988217057,
      "loss": 4.4028,
      "mean_token_accuracy": 0.2587141692638397,
      "num_tokens": 13784586.0,
      "step": 17290
    },
    {
      "entropy": 4.645528554916382,
      "epoch": 4.889737065309585,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00029320697740403543,
      "loss": 4.3469,
      "mean_token_accuracy": 0.26822112798690795,
      "num_tokens": 13788715.0,
      "step": 17295
    },
    {
      "entropy": 4.599263620376587,
      "epoch": 4.89115069267741,
      "grad_norm": 1.671875,
      "learning_rate": 0.00029310445218121953,
      "loss": 4.3678,
      "mean_token_accuracy": 0.27996574640274047,
      "num_tokens": 13792508.0,
      "step": 17300
    },
    {
      "entropy": 4.633902072906494,
      "epoch": 4.892564320045236,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002930019231746851,
      "loss": 4.3791,
      "mean_token_accuracy": 0.26065263599157334,
      "num_tokens": 13796169.0,
      "step": 17305
    },
    {
      "entropy": 4.625073957443237,
      "epoch": 4.893977947413062,
      "grad_norm": 1.53125,
      "learning_rate": 0.00029289939040585993,
      "loss": 4.4224,
      "mean_token_accuracy": 0.2746116518974304,
      "num_tokens": 13800413.0,
      "step": 17310
    },
    {
      "entropy": 4.690625095367432,
      "epoch": 4.895391574780888,
      "grad_norm": 1.6875,
      "learning_rate": 0.000292796853896173,
      "loss": 4.4269,
      "mean_token_accuracy": 0.261276350915432,
      "num_tokens": 13804642.0,
      "step": 17315
    },
    {
      "entropy": 4.653651475906372,
      "epoch": 4.896805202148713,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002926943136670536,
      "loss": 4.3664,
      "mean_token_accuracy": 0.2732554867863655,
      "num_tokens": 13808764.0,
      "step": 17320
    },
    {
      "entropy": 4.658253622055054,
      "epoch": 4.898218829516539,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002925917697399321,
      "loss": 4.3779,
      "mean_token_accuracy": 0.27158187329769135,
      "num_tokens": 13812816.0,
      "step": 17325
    },
    {
      "entropy": 4.540979242324829,
      "epoch": 4.899632456884365,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002924892221362394,
      "loss": 4.3206,
      "mean_token_accuracy": 0.2625752776861191,
      "num_tokens": 13816903.0,
      "step": 17330
    },
    {
      "entropy": 4.58760290145874,
      "epoch": 4.901046084252191,
      "grad_norm": 1.703125,
      "learning_rate": 0.00029238667087740756,
      "loss": 4.3192,
      "mean_token_accuracy": 0.26830507069826126,
      "num_tokens": 13820723.0,
      "step": 17335
    },
    {
      "entropy": 4.694156694412231,
      "epoch": 4.902459711620017,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002922841159848689,
      "loss": 4.4756,
      "mean_token_accuracy": 0.24976342022418976,
      "num_tokens": 13824590.0,
      "step": 17340
    },
    {
      "entropy": 4.708672094345093,
      "epoch": 4.903873338987843,
      "grad_norm": 1.65625,
      "learning_rate": 0.000292181557480057,
      "loss": 4.4978,
      "mean_token_accuracy": 0.2412676140666008,
      "num_tokens": 13828831.0,
      "step": 17345
    },
    {
      "entropy": 4.55072135925293,
      "epoch": 4.905286966355669,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002920789953844058,
      "loss": 4.3383,
      "mean_token_accuracy": 0.2761520937085152,
      "num_tokens": 13832519.0,
      "step": 17350
    },
    {
      "entropy": 4.613760852813721,
      "epoch": 4.906700593723494,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002919764297193501,
      "loss": 4.367,
      "mean_token_accuracy": 0.2682879686355591,
      "num_tokens": 13836494.0,
      "step": 17355
    },
    {
      "entropy": 4.650279808044433,
      "epoch": 4.908114221091321,
      "grad_norm": 1.734375,
      "learning_rate": 0.00029187386050632565,
      "loss": 4.3907,
      "mean_token_accuracy": 0.2720949038863182,
      "num_tokens": 13840298.0,
      "step": 17360
    },
    {
      "entropy": 4.643546772003174,
      "epoch": 4.909527848459146,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002917712877667687,
      "loss": 4.4247,
      "mean_token_accuracy": 0.26323710978031156,
      "num_tokens": 13844203.0,
      "step": 17365
    },
    {
      "entropy": 4.619313716888428,
      "epoch": 4.910941475826972,
      "grad_norm": 1.875,
      "learning_rate": 0.0002916687115221162,
      "loss": 4.3662,
      "mean_token_accuracy": 0.27232060134410857,
      "num_tokens": 13847924.0,
      "step": 17370
    },
    {
      "entropy": 4.648451900482177,
      "epoch": 4.912355103194798,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002915661317938062,
      "loss": 4.3731,
      "mean_token_accuracy": 0.2627526342868805,
      "num_tokens": 13851573.0,
      "step": 17375
    },
    {
      "entropy": 4.701597261428833,
      "epoch": 4.913768730562624,
      "grad_norm": 1.640625,
      "learning_rate": 0.00029146354860327706,
      "loss": 4.3986,
      "mean_token_accuracy": 0.2575708657503128,
      "num_tokens": 13855737.0,
      "step": 17380
    },
    {
      "entropy": 4.714762592315674,
      "epoch": 4.915182357930449,
      "grad_norm": 1.59375,
      "learning_rate": 0.00029136096197196807,
      "loss": 4.4918,
      "mean_token_accuracy": 0.24637792855501175,
      "num_tokens": 13859937.0,
      "step": 17385
    },
    {
      "entropy": 4.627321386337281,
      "epoch": 4.916595985298275,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00029125837192131923,
      "loss": 4.3671,
      "mean_token_accuracy": 0.26661708354949953,
      "num_tokens": 13863840.0,
      "step": 17390
    },
    {
      "entropy": 4.630975770950317,
      "epoch": 4.9180096126661015,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002911557784727711,
      "loss": 4.3852,
      "mean_token_accuracy": 0.2751364454627037,
      "num_tokens": 13867619.0,
      "step": 17395
    },
    {
      "entropy": 4.567185640335083,
      "epoch": 4.919423240033927,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00029105318164776524,
      "loss": 4.2917,
      "mean_token_accuracy": 0.2785364031791687,
      "num_tokens": 13871555.0,
      "step": 17400
    },
    {
      "entropy": 4.611128091812134,
      "epoch": 4.920836867401753,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002909505814677437,
      "loss": 4.439,
      "mean_token_accuracy": 0.2646558165550232,
      "num_tokens": 13875629.0,
      "step": 17405
    },
    {
      "entropy": 4.591842365264893,
      "epoch": 4.922250494769579,
      "grad_norm": 1.890625,
      "learning_rate": 0.00029084797795414925,
      "loss": 4.257,
      "mean_token_accuracy": 0.2929065451025963,
      "num_tokens": 13879441.0,
      "step": 17410
    },
    {
      "entropy": 4.7575901508331295,
      "epoch": 4.923664122137405,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002907453711284254,
      "loss": 4.4252,
      "mean_token_accuracy": 0.26412821710109713,
      "num_tokens": 13883500.0,
      "step": 17415
    },
    {
      "entropy": 4.660673522949219,
      "epoch": 4.92507774950523,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00029064276101201634,
      "loss": 4.3533,
      "mean_token_accuracy": 0.26880827695131304,
      "num_tokens": 13887489.0,
      "step": 17420
    },
    {
      "entropy": 4.705656623840332,
      "epoch": 4.926491376873056,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000290540147626367,
      "loss": 4.3783,
      "mean_token_accuracy": 0.26742522418498993,
      "num_tokens": 13891430.0,
      "step": 17425
    },
    {
      "entropy": 4.641457986831665,
      "epoch": 4.927905004240882,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00029043753099292286,
      "loss": 4.4293,
      "mean_token_accuracy": 0.26377545297145844,
      "num_tokens": 13895656.0,
      "step": 17430
    },
    {
      "entropy": 4.605193090438843,
      "epoch": 4.929318631608708,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002903349111331302,
      "loss": 4.3199,
      "mean_token_accuracy": 0.27568828463554385,
      "num_tokens": 13899440.0,
      "step": 17435
    },
    {
      "entropy": 4.6377135753631595,
      "epoch": 4.9307322589765334,
      "grad_norm": 1.640625,
      "learning_rate": 0.000290232288068436,
      "loss": 4.386,
      "mean_token_accuracy": 0.2740935891866684,
      "num_tokens": 13903529.0,
      "step": 17440
    },
    {
      "entropy": 4.753320217132568,
      "epoch": 4.93214588634436,
      "grad_norm": 1.703125,
      "learning_rate": 0.00029012966182028773,
      "loss": 4.4731,
      "mean_token_accuracy": 0.2634014695882797,
      "num_tokens": 13907676.0,
      "step": 17445
    },
    {
      "entropy": 4.729758262634277,
      "epoch": 4.9335595137121855,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00029002703241013376,
      "loss": 4.456,
      "mean_token_accuracy": 0.2569182485342026,
      "num_tokens": 13911773.0,
      "step": 17450
    },
    {
      "entropy": 4.644197225570679,
      "epoch": 4.934973141080011,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00028992439985942286,
      "loss": 4.371,
      "mean_token_accuracy": 0.27112300097942355,
      "num_tokens": 13915877.0,
      "step": 17455
    },
    {
      "entropy": 4.615805768966675,
      "epoch": 4.9363867684478375,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002898217641896047,
      "loss": 4.4062,
      "mean_token_accuracy": 0.2611717417836189,
      "num_tokens": 13919695.0,
      "step": 17460
    },
    {
      "entropy": 4.637508964538574,
      "epoch": 4.937800395815663,
      "grad_norm": 1.859375,
      "learning_rate": 0.0002897191254221295,
      "loss": 4.4224,
      "mean_token_accuracy": 0.25554819256067274,
      "num_tokens": 13923430.0,
      "step": 17465
    },
    {
      "entropy": 4.68900957107544,
      "epoch": 4.939214023183489,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002896164835784481,
      "loss": 4.3416,
      "mean_token_accuracy": 0.26223977506160734,
      "num_tokens": 13927511.0,
      "step": 17470
    },
    {
      "entropy": 4.649257326126099,
      "epoch": 4.940627650551315,
      "grad_norm": 1.640625,
      "learning_rate": 0.00028951383868001213,
      "loss": 4.4607,
      "mean_token_accuracy": 0.2682576537132263,
      "num_tokens": 13931272.0,
      "step": 17475
    },
    {
      "entropy": 4.666018629074097,
      "epoch": 4.942041277919141,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002894111907482735,
      "loss": 4.386,
      "mean_token_accuracy": 0.2683520704507828,
      "num_tokens": 13935399.0,
      "step": 17480
    },
    {
      "entropy": 4.752054643630982,
      "epoch": 4.943454905286966,
      "grad_norm": 1.75,
      "learning_rate": 0.0002893085398046851,
      "loss": 4.4018,
      "mean_token_accuracy": 0.27069261223077773,
      "num_tokens": 13939205.0,
      "step": 17485
    },
    {
      "entropy": 4.604634141921997,
      "epoch": 4.944868532654792,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002892058858707005,
      "loss": 4.3312,
      "mean_token_accuracy": 0.26671389788389205,
      "num_tokens": 13943265.0,
      "step": 17490
    },
    {
      "entropy": 4.613538789749145,
      "epoch": 4.946282160022618,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002891032289677734,
      "loss": 4.3746,
      "mean_token_accuracy": 0.2627345934510231,
      "num_tokens": 13947354.0,
      "step": 17495
    },
    {
      "entropy": 4.58404221534729,
      "epoch": 4.947695787390444,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0002890005691173588,
      "loss": 4.301,
      "mean_token_accuracy": 0.27311893850564956,
      "num_tokens": 13951464.0,
      "step": 17500
    },
    {
      "entropy": 4.622881126403809,
      "epoch": 4.9491094147582695,
      "grad_norm": 1.546875,
      "learning_rate": 0.0002888979063409118,
      "loss": 4.3548,
      "mean_token_accuracy": 0.26803555637598037,
      "num_tokens": 13955671.0,
      "step": 17505
    },
    {
      "entropy": 4.519285202026367,
      "epoch": 4.950523042126095,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00028879524065988834,
      "loss": 4.2732,
      "mean_token_accuracy": 0.28535034209489823,
      "num_tokens": 13959607.0,
      "step": 17510
    },
    {
      "entropy": 4.588117170333862,
      "epoch": 4.9519366694939215,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002886925720957448,
      "loss": 4.251,
      "mean_token_accuracy": 0.27143285423517227,
      "num_tokens": 13963783.0,
      "step": 17515
    },
    {
      "entropy": 4.661261510848999,
      "epoch": 4.953350296861747,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00028858990066993844,
      "loss": 4.3447,
      "mean_token_accuracy": 0.2624423950910568,
      "num_tokens": 13968034.0,
      "step": 17520
    },
    {
      "entropy": 4.635093784332275,
      "epoch": 4.954763924229573,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00028848722640392683,
      "loss": 4.3392,
      "mean_token_accuracy": 0.26818606108427046,
      "num_tokens": 13971618.0,
      "step": 17525
    },
    {
      "entropy": 4.600364017486572,
      "epoch": 4.956177551597399,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00028838454931916837,
      "loss": 4.3906,
      "mean_token_accuracy": 0.2623980164527893,
      "num_tokens": 13975991.0,
      "step": 17530
    },
    {
      "entropy": 4.590821361541748,
      "epoch": 4.957591178965225,
      "grad_norm": 1.578125,
      "learning_rate": 0.00028828186943712184,
      "loss": 4.3303,
      "mean_token_accuracy": 0.26593889594078063,
      "num_tokens": 13980065.0,
      "step": 17535
    },
    {
      "entropy": 4.659433078765869,
      "epoch": 4.95900480633305,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00028817918677924675,
      "loss": 4.4166,
      "mean_token_accuracy": 0.2705103918910027,
      "num_tokens": 13983866.0,
      "step": 17540
    },
    {
      "entropy": 4.656587934494018,
      "epoch": 4.960418433700877,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002880765013670031,
      "loss": 4.3966,
      "mean_token_accuracy": 0.25667043030261993,
      "num_tokens": 13987795.0,
      "step": 17545
    },
    {
      "entropy": 4.633965492248535,
      "epoch": 4.961832061068702,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002879738132218515,
      "loss": 4.3259,
      "mean_token_accuracy": 0.2770823687314987,
      "num_tokens": 13991484.0,
      "step": 17550
    },
    {
      "entropy": 4.555609273910522,
      "epoch": 4.963245688436528,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00028787112236525317,
      "loss": 4.2794,
      "mean_token_accuracy": 0.2719076618552208,
      "num_tokens": 13995210.0,
      "step": 17555
    },
    {
      "entropy": 4.6143590927124025,
      "epoch": 4.964659315804354,
      "grad_norm": 1.515625,
      "learning_rate": 0.00028776842881866985,
      "loss": 4.3982,
      "mean_token_accuracy": 0.260223975777626,
      "num_tokens": 13999417.0,
      "step": 17560
    },
    {
      "entropy": 4.6553192138671875,
      "epoch": 4.96607294317218,
      "grad_norm": 1.53125,
      "learning_rate": 0.00028766573260356396,
      "loss": 4.405,
      "mean_token_accuracy": 0.265083347260952,
      "num_tokens": 14003667.0,
      "step": 17565
    },
    {
      "entropy": 4.6726165294647215,
      "epoch": 4.9674865705400055,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002875630337413982,
      "loss": 4.3362,
      "mean_token_accuracy": 0.27129075229167937,
      "num_tokens": 14007862.0,
      "step": 17570
    },
    {
      "entropy": 4.678282737731934,
      "epoch": 4.968900197907832,
      "grad_norm": 1.59375,
      "learning_rate": 0.00028746033225363597,
      "loss": 4.3337,
      "mean_token_accuracy": 0.2686383157968521,
      "num_tokens": 14012019.0,
      "step": 17575
    },
    {
      "entropy": 4.662152814865112,
      "epoch": 4.970313825275658,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00028735762816174156,
      "loss": 4.307,
      "mean_token_accuracy": 0.2693633049726486,
      "num_tokens": 14015914.0,
      "step": 17580
    },
    {
      "entropy": 4.594103908538818,
      "epoch": 4.971727452643483,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002872549214871791,
      "loss": 4.3904,
      "mean_token_accuracy": 0.26960517168045045,
      "num_tokens": 14020082.0,
      "step": 17585
    },
    {
      "entropy": 4.5735537052154545,
      "epoch": 4.973141080011309,
      "grad_norm": 1.640625,
      "learning_rate": 0.00028715221225141397,
      "loss": 4.3428,
      "mean_token_accuracy": 0.26914130598306657,
      "num_tokens": 14023949.0,
      "step": 17590
    },
    {
      "entropy": 4.602140855789185,
      "epoch": 4.974554707379135,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002870495004759116,
      "loss": 4.3112,
      "mean_token_accuracy": 0.2758776113390923,
      "num_tokens": 14028017.0,
      "step": 17595
    },
    {
      "entropy": 4.685879230499268,
      "epoch": 4.975968334746961,
      "grad_norm": 1.71875,
      "learning_rate": 0.00028694678618213816,
      "loss": 4.4539,
      "mean_token_accuracy": 0.26337245255708697,
      "num_tokens": 14031722.0,
      "step": 17600
    },
    {
      "entropy": 4.768280553817749,
      "epoch": 4.977381962114786,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00028684406939156017,
      "loss": 4.4576,
      "mean_token_accuracy": 0.2497575804591179,
      "num_tokens": 14035889.0,
      "step": 17605
    },
    {
      "entropy": 4.759493350982666,
      "epoch": 4.978795589482612,
      "grad_norm": 1.78125,
      "learning_rate": 0.00028674135012564503,
      "loss": 4.4687,
      "mean_token_accuracy": 0.2507967472076416,
      "num_tokens": 14039534.0,
      "step": 17610
    },
    {
      "entropy": 4.578010892868042,
      "epoch": 4.980209216850438,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002866386284058602,
      "loss": 4.2995,
      "mean_token_accuracy": 0.27462128251791,
      "num_tokens": 14043757.0,
      "step": 17615
    },
    {
      "entropy": 4.50666561126709,
      "epoch": 4.981622844218264,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00028653590425367414,
      "loss": 4.2718,
      "mean_token_accuracy": 0.2788428246974945,
      "num_tokens": 14047851.0,
      "step": 17620
    },
    {
      "entropy": 4.642121505737305,
      "epoch": 4.9830364715860895,
      "grad_norm": 1.734375,
      "learning_rate": 0.00028643317769055536,
      "loss": 4.4358,
      "mean_token_accuracy": 0.26547839790582656,
      "num_tokens": 14051824.0,
      "step": 17625
    },
    {
      "entropy": 4.733785963058471,
      "epoch": 4.984450098953916,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002863304487379732,
      "loss": 4.3971,
      "mean_token_accuracy": 0.2592454344034195,
      "num_tokens": 14055917.0,
      "step": 17630
    },
    {
      "entropy": 4.699671888351441,
      "epoch": 4.985863726321742,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002862277174173972,
      "loss": 4.4519,
      "mean_token_accuracy": 0.25982857346534727,
      "num_tokens": 14059826.0,
      "step": 17635
    },
    {
      "entropy": 4.604775333404541,
      "epoch": 4.987277353689567,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002861249837502978,
      "loss": 4.3173,
      "mean_token_accuracy": 0.2755934104323387,
      "num_tokens": 14063753.0,
      "step": 17640
    },
    {
      "entropy": 4.649540901184082,
      "epoch": 4.988690981057394,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002860222477581455,
      "loss": 4.4075,
      "mean_token_accuracy": 0.26268425285816194,
      "num_tokens": 14067618.0,
      "step": 17645
    },
    {
      "entropy": 4.633990049362183,
      "epoch": 4.990104608425219,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002859195094624117,
      "loss": 4.3955,
      "mean_token_accuracy": 0.25973164439201357,
      "num_tokens": 14071443.0,
      "step": 17650
    },
    {
      "entropy": 4.640814924240113,
      "epoch": 4.991518235793045,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002858167688845679,
      "loss": 4.4004,
      "mean_token_accuracy": 0.2667288690805435,
      "num_tokens": 14075061.0,
      "step": 17655
    },
    {
      "entropy": 4.659849739074707,
      "epoch": 4.992931863160871,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002857140260460863,
      "loss": 4.4337,
      "mean_token_accuracy": 0.2700193628668785,
      "num_tokens": 14078841.0,
      "step": 17660
    },
    {
      "entropy": 4.602811336517334,
      "epoch": 4.994345490528697,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002856112809684394,
      "loss": 4.3099,
      "mean_token_accuracy": 0.2809151828289032,
      "num_tokens": 14082525.0,
      "step": 17665
    },
    {
      "entropy": 4.788411569595337,
      "epoch": 4.995759117896522,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002855085336731005,
      "loss": 4.497,
      "mean_token_accuracy": 0.25621832609176637,
      "num_tokens": 14086542.0,
      "step": 17670
    },
    {
      "entropy": 4.603898000717163,
      "epoch": 4.997172745264348,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002854057841815431,
      "loss": 4.359,
      "mean_token_accuracy": 0.2720381736755371,
      "num_tokens": 14090632.0,
      "step": 17675
    },
    {
      "entropy": 4.61132493019104,
      "epoch": 4.998586372632174,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00028530303251524103,
      "loss": 4.3074,
      "mean_token_accuracy": 0.27113581001758574,
      "num_tokens": 14094788.0,
      "step": 17680
    },
    {
      "entropy": 4.648876523971557,
      "epoch": 5.0,
      "grad_norm": 2.0625,
      "learning_rate": 0.00028520027869566897,
      "loss": 4.4445,
      "mean_token_accuracy": 0.2594813764095306,
      "num_tokens": 14098625.0,
      "step": 17685
    },
    {
      "entropy": 4.681940650939941,
      "epoch": 5.001413627367826,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002850975227443017,
      "loss": 4.2186,
      "mean_token_accuracy": 0.2729207083582878,
      "num_tokens": 14102910.0,
      "step": 17690
    },
    {
      "entropy": 4.637896299362183,
      "epoch": 5.002827254735652,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002849947646826146,
      "loss": 4.2148,
      "mean_token_accuracy": 0.28600748926401137,
      "num_tokens": 14106927.0,
      "step": 17695
    },
    {
      "entropy": 4.620762491226197,
      "epoch": 5.004240882103478,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002848920045320834,
      "loss": 4.1706,
      "mean_token_accuracy": 0.2819751903414726,
      "num_tokens": 14110777.0,
      "step": 17700
    },
    {
      "entropy": 4.655347394943237,
      "epoch": 5.005654509471303,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002847892423141846,
      "loss": 4.256,
      "mean_token_accuracy": 0.2726939290761948,
      "num_tokens": 14114753.0,
      "step": 17705
    },
    {
      "entropy": 4.629832124710083,
      "epoch": 5.007068136839129,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00028468647805039455,
      "loss": 4.1838,
      "mean_token_accuracy": 0.27807227373123167,
      "num_tokens": 14118957.0,
      "step": 17710
    },
    {
      "entropy": 4.574212741851807,
      "epoch": 5.008481764206955,
      "grad_norm": 1.703125,
      "learning_rate": 0.00028458371176219056,
      "loss": 4.1471,
      "mean_token_accuracy": 0.28320263624191283,
      "num_tokens": 14122918.0,
      "step": 17715
    },
    {
      "entropy": 4.697670984268188,
      "epoch": 5.009895391574781,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002844809434710501,
      "loss": 4.3676,
      "mean_token_accuracy": 0.26281843036413194,
      "num_tokens": 14126912.0,
      "step": 17720
    },
    {
      "entropy": 4.744941234588623,
      "epoch": 5.011309018942606,
      "grad_norm": 1.609375,
      "learning_rate": 0.00028437817319845103,
      "loss": 4.3351,
      "mean_token_accuracy": 0.2678901106119156,
      "num_tokens": 14130924.0,
      "step": 17725
    },
    {
      "entropy": 4.714976739883423,
      "epoch": 5.012722646310433,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00028427540096587174,
      "loss": 4.2945,
      "mean_token_accuracy": 0.27422787845134733,
      "num_tokens": 14135169.0,
      "step": 17730
    },
    {
      "entropy": 4.763168382644653,
      "epoch": 5.014136273678258,
      "grad_norm": 1.71875,
      "learning_rate": 0.00028417262679479105,
      "loss": 4.3047,
      "mean_token_accuracy": 0.2694909140467644,
      "num_tokens": 14138869.0,
      "step": 17735
    },
    {
      "entropy": 4.629745626449585,
      "epoch": 5.015549901046084,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00028406985070668806,
      "loss": 4.2417,
      "mean_token_accuracy": 0.27756265699863436,
      "num_tokens": 14142696.0,
      "step": 17740
    },
    {
      "entropy": 4.605633687973023,
      "epoch": 5.0169635284139105,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002839670727230424,
      "loss": 4.1547,
      "mean_token_accuracy": 0.2927120178937912,
      "num_tokens": 14146480.0,
      "step": 17745
    },
    {
      "entropy": 4.582312679290771,
      "epoch": 5.018377155781736,
      "grad_norm": 1.625,
      "learning_rate": 0.0002838642928653339,
      "loss": 4.1751,
      "mean_token_accuracy": 0.2844663098454475,
      "num_tokens": 14150423.0,
      "step": 17750
    },
    {
      "entropy": 4.643700218200683,
      "epoch": 5.019790783149562,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00028376151115504315,
      "loss": 4.2586,
      "mean_token_accuracy": 0.26429236531257627,
      "num_tokens": 14154501.0,
      "step": 17755
    },
    {
      "entropy": 4.680150651931763,
      "epoch": 5.021204410517387,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00028365872761365065,
      "loss": 4.2201,
      "mean_token_accuracy": 0.26573477387428285,
      "num_tokens": 14158371.0,
      "step": 17760
    },
    {
      "entropy": 4.6918616771698,
      "epoch": 5.022618037885214,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00028355594226263766,
      "loss": 4.2776,
      "mean_token_accuracy": 0.27032297998666766,
      "num_tokens": 14162486.0,
      "step": 17765
    },
    {
      "entropy": 4.644112205505371,
      "epoch": 5.024031665253039,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00028345315512348553,
      "loss": 4.2393,
      "mean_token_accuracy": 0.2768447771668434,
      "num_tokens": 14166503.0,
      "step": 17770
    },
    {
      "entropy": 4.683706951141358,
      "epoch": 5.025445292620865,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00028335036621767633,
      "loss": 4.3003,
      "mean_token_accuracy": 0.26455442011356356,
      "num_tokens": 14170596.0,
      "step": 17775
    },
    {
      "entropy": 4.647518062591553,
      "epoch": 5.026858919988691,
      "grad_norm": 1.59375,
      "learning_rate": 0.00028324757556669227,
      "loss": 4.1487,
      "mean_token_accuracy": 0.28049067705869674,
      "num_tokens": 14174630.0,
      "step": 17780
    },
    {
      "entropy": 4.651795721054077,
      "epoch": 5.028272547356517,
      "grad_norm": 1.828125,
      "learning_rate": 0.00028314478319201595,
      "loss": 4.2413,
      "mean_token_accuracy": 0.2745653823018074,
      "num_tokens": 14178348.0,
      "step": 17785
    },
    {
      "entropy": 4.67831859588623,
      "epoch": 5.029686174724342,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002830419891151302,
      "loss": 4.2463,
      "mean_token_accuracy": 0.2630108520388603,
      "num_tokens": 14182479.0,
      "step": 17790
    },
    {
      "entropy": 4.583390903472901,
      "epoch": 5.031099802092169,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002829391933575185,
      "loss": 4.1097,
      "mean_token_accuracy": 0.28394579887390137,
      "num_tokens": 14186215.0,
      "step": 17795
    },
    {
      "entropy": 4.600494289398194,
      "epoch": 5.0325134294599945,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00028283639594066444,
      "loss": 4.2152,
      "mean_token_accuracy": 0.28342385590076447,
      "num_tokens": 14190030.0,
      "step": 17800
    },
    {
      "entropy": 4.567687845230102,
      "epoch": 5.03392705682782,
      "grad_norm": 1.65625,
      "learning_rate": 0.00028273359688605216,
      "loss": 4.1452,
      "mean_token_accuracy": 0.2843063324689865,
      "num_tokens": 14193949.0,
      "step": 17805
    },
    {
      "entropy": 4.55774416923523,
      "epoch": 5.035340684195646,
      "grad_norm": 1.7109375,
      "learning_rate": 0.000282630796215166,
      "loss": 4.2113,
      "mean_token_accuracy": 0.28773525059223176,
      "num_tokens": 14197745.0,
      "step": 17810
    },
    {
      "entropy": 4.591823673248291,
      "epoch": 5.036754311563472,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002825279939494905,
      "loss": 4.1351,
      "mean_token_accuracy": 0.2811704933643341,
      "num_tokens": 14201815.0,
      "step": 17815
    },
    {
      "entropy": 4.7307781219482425,
      "epoch": 5.038167938931298,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0002824251901105108,
      "loss": 4.3837,
      "mean_token_accuracy": 0.2576375171542168,
      "num_tokens": 14206261.0,
      "step": 17820
    },
    {
      "entropy": 4.645414113998413,
      "epoch": 5.039581566299123,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002823223847197124,
      "loss": 4.2141,
      "mean_token_accuracy": 0.2730750903487206,
      "num_tokens": 14210120.0,
      "step": 17825
    },
    {
      "entropy": 4.599419164657593,
      "epoch": 5.04099519366695,
      "grad_norm": 1.703125,
      "learning_rate": 0.00028221957779858074,
      "loss": 4.2078,
      "mean_token_accuracy": 0.281390979886055,
      "num_tokens": 14214057.0,
      "step": 17830
    },
    {
      "entropy": 4.567046642303467,
      "epoch": 5.042408821034775,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002821167693686021,
      "loss": 4.1779,
      "mean_token_accuracy": 0.2911563217639923,
      "num_tokens": 14218170.0,
      "step": 17835
    },
    {
      "entropy": 4.505235004425049,
      "epoch": 5.043822448402601,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00028201395945126265,
      "loss": 4.0215,
      "mean_token_accuracy": 0.30649341344833375,
      "num_tokens": 14221807.0,
      "step": 17840
    },
    {
      "entropy": 4.580894804000854,
      "epoch": 5.045236075770427,
      "grad_norm": 1.796875,
      "learning_rate": 0.000281911148068049,
      "loss": 4.2017,
      "mean_token_accuracy": 0.2790472790598869,
      "num_tokens": 14225625.0,
      "step": 17845
    },
    {
      "entropy": 4.555585050582886,
      "epoch": 5.046649703138253,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00028180833524044807,
      "loss": 4.1163,
      "mean_token_accuracy": 0.28526702523231506,
      "num_tokens": 14229322.0,
      "step": 17850
    },
    {
      "entropy": 4.624291706085205,
      "epoch": 5.0480633305060785,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002817055209899473,
      "loss": 4.2579,
      "mean_token_accuracy": 0.271120610833168,
      "num_tokens": 14233254.0,
      "step": 17855
    },
    {
      "entropy": 4.651657152175903,
      "epoch": 5.049476957873904,
      "grad_norm": 1.96875,
      "learning_rate": 0.00028160270533803406,
      "loss": 4.2859,
      "mean_token_accuracy": 0.2754980280995369,
      "num_tokens": 14237391.0,
      "step": 17860
    },
    {
      "entropy": 4.60790982246399,
      "epoch": 5.0508905852417305,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0002814998883061963,
      "loss": 4.1297,
      "mean_token_accuracy": 0.28592145144939424,
      "num_tokens": 14241172.0,
      "step": 17865
    },
    {
      "entropy": 4.639704036712646,
      "epoch": 5.052304212609556,
      "grad_norm": 1.71875,
      "learning_rate": 0.000281397069915922,
      "loss": 4.2636,
      "mean_token_accuracy": 0.2726334139704704,
      "num_tokens": 14244930.0,
      "step": 17870
    },
    {
      "entropy": 4.5957639694213865,
      "epoch": 5.053717839977382,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002812942501886997,
      "loss": 4.1582,
      "mean_token_accuracy": 0.2712148755788803,
      "num_tokens": 14248868.0,
      "step": 17875
    },
    {
      "entropy": 4.660074186325073,
      "epoch": 5.055131467345208,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002811914291460179,
      "loss": 4.28,
      "mean_token_accuracy": 0.2654839098453522,
      "num_tokens": 14253036.0,
      "step": 17880
    },
    {
      "entropy": 4.727026510238647,
      "epoch": 5.056545094713034,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002810886068093658,
      "loss": 4.2814,
      "mean_token_accuracy": 0.26966695934534074,
      "num_tokens": 14257260.0,
      "step": 17885
    },
    {
      "entropy": 4.688440608978271,
      "epoch": 5.057958722080859,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00028098578320023257,
      "loss": 4.3284,
      "mean_token_accuracy": 0.27380492687225344,
      "num_tokens": 14261515.0,
      "step": 17890
    },
    {
      "entropy": 4.616662693023682,
      "epoch": 5.059372349448686,
      "grad_norm": 1.65625,
      "learning_rate": 0.00028088295834010756,
      "loss": 4.1725,
      "mean_token_accuracy": 0.27269008457660676,
      "num_tokens": 14265292.0,
      "step": 17895
    },
    {
      "entropy": 4.649060297012329,
      "epoch": 5.060785976816511,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002807801322504807,
      "loss": 4.1971,
      "mean_token_accuracy": 0.2775640174746513,
      "num_tokens": 14269172.0,
      "step": 17900
    },
    {
      "entropy": 4.480349159240722,
      "epoch": 5.062199604184337,
      "grad_norm": 1.9921875,
      "learning_rate": 0.000280677304952842,
      "loss": 4.0518,
      "mean_token_accuracy": 0.29249016642570497,
      "num_tokens": 14272957.0,
      "step": 17905
    },
    {
      "entropy": 4.593162727355957,
      "epoch": 5.0636132315521625,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002805744764686816,
      "loss": 4.2565,
      "mean_token_accuracy": 0.2796078145503998,
      "num_tokens": 14277071.0,
      "step": 17910
    },
    {
      "entropy": 4.586597204208374,
      "epoch": 5.065026858919989,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00028047164681949013,
      "loss": 4.1179,
      "mean_token_accuracy": 0.2919781655073166,
      "num_tokens": 14280948.0,
      "step": 17915
    },
    {
      "entropy": 4.613659000396728,
      "epoch": 5.0664404862878145,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002803688160267582,
      "loss": 4.2234,
      "mean_token_accuracy": 0.2763011157512665,
      "num_tokens": 14284865.0,
      "step": 17920
    },
    {
      "entropy": 4.517279434204101,
      "epoch": 5.06785411365564,
      "grad_norm": 1.765625,
      "learning_rate": 0.000280265984111977,
      "loss": 4.1422,
      "mean_token_accuracy": 0.2819807693362236,
      "num_tokens": 14288764.0,
      "step": 17925
    },
    {
      "entropy": 4.541666269302368,
      "epoch": 5.0692677410234666,
      "grad_norm": 1.84375,
      "learning_rate": 0.0002801631510966377,
      "loss": 4.1669,
      "mean_token_accuracy": 0.2794782042503357,
      "num_tokens": 14292700.0,
      "step": 17930
    },
    {
      "entropy": 4.641516923904419,
      "epoch": 5.070681368391292,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00028006031700223175,
      "loss": 4.1904,
      "mean_token_accuracy": 0.27206475138664243,
      "num_tokens": 14296564.0,
      "step": 17935
    },
    {
      "entropy": 4.670997190475464,
      "epoch": 5.072094995759118,
      "grad_norm": 1.65625,
      "learning_rate": 0.00027995748185025084,
      "loss": 4.2805,
      "mean_token_accuracy": 0.26715395152568816,
      "num_tokens": 14300438.0,
      "step": 17940
    },
    {
      "entropy": 4.635912656784058,
      "epoch": 5.073508623126944,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00027985464566218686,
      "loss": 4.3321,
      "mean_token_accuracy": 0.26580086946487425,
      "num_tokens": 14304604.0,
      "step": 17945
    },
    {
      "entropy": 4.624765300750733,
      "epoch": 5.07492225049477,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002797518084595319,
      "loss": 4.2383,
      "mean_token_accuracy": 0.27349206507205964,
      "num_tokens": 14308629.0,
      "step": 17950
    },
    {
      "entropy": 4.657820415496826,
      "epoch": 5.076335877862595,
      "grad_norm": 1.9140625,
      "learning_rate": 0.0002796489702637784,
      "loss": 4.3326,
      "mean_token_accuracy": 0.2687433287501335,
      "num_tokens": 14312573.0,
      "step": 17955
    },
    {
      "entropy": 4.606703996658325,
      "epoch": 5.077749505230421,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002795461310964189,
      "loss": 4.1274,
      "mean_token_accuracy": 0.29115923792123793,
      "num_tokens": 14316762.0,
      "step": 17960
    },
    {
      "entropy": 4.54243426322937,
      "epoch": 5.079163132598247,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002794432909789461,
      "loss": 4.1554,
      "mean_token_accuracy": 0.28753942251205444,
      "num_tokens": 14321010.0,
      "step": 17965
    },
    {
      "entropy": 4.528362655639649,
      "epoch": 5.080576759966073,
      "grad_norm": 1.6015625,
      "learning_rate": 0.000279340449932853,
      "loss": 4.199,
      "mean_token_accuracy": 0.2858507752418518,
      "num_tokens": 14325093.0,
      "step": 17970
    },
    {
      "entropy": 4.604999256134033,
      "epoch": 5.0819903873338985,
      "grad_norm": 1.7890625,
      "learning_rate": 0.00027923760797963267,
      "loss": 4.1851,
      "mean_token_accuracy": 0.2842082679271698,
      "num_tokens": 14328950.0,
      "step": 17975
    },
    {
      "entropy": 4.612251281738281,
      "epoch": 5.083404014701725,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00027913476514077843,
      "loss": 4.2173,
      "mean_token_accuracy": 0.27711353451013565,
      "num_tokens": 14332999.0,
      "step": 17980
    },
    {
      "entropy": 4.622544765472412,
      "epoch": 5.084817642069551,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002790319214377838,
      "loss": 4.1798,
      "mean_token_accuracy": 0.2904620379209518,
      "num_tokens": 14337010.0,
      "step": 17985
    },
    {
      "entropy": 4.606207370758057,
      "epoch": 5.086231269437376,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00027892907689214264,
      "loss": 4.2008,
      "mean_token_accuracy": 0.281675785779953,
      "num_tokens": 14341049.0,
      "step": 17990
    },
    {
      "entropy": 4.599209833145141,
      "epoch": 5.087644896805202,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002788262315253487,
      "loss": 4.2336,
      "mean_token_accuracy": 0.2817929744720459,
      "num_tokens": 14345212.0,
      "step": 17995
    },
    {
      "entropy": 4.6215372562408445,
      "epoch": 5.089058524173028,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00027872338535889603,
      "loss": 4.2314,
      "mean_token_accuracy": 0.28000919073820113,
      "num_tokens": 14348961.0,
      "step": 18000
    },
    {
      "epoch": 5.089058524173028,
      "eval_entropy": 4.462487761303518,
      "eval_loss": 4.972434043884277,
      "eval_mean_token_accuracy": 0.25110239974367626,
      "eval_num_tokens": 14348961.0,
      "eval_runtime": 6.6142,
      "eval_samples_per_second": 1706.033,
      "eval_steps_per_second": 213.33,
      "step": 18000
    },
    {
      "entropy": 4.622426509857178,
      "epoch": 5.090472151540854,
      "grad_norm": 1.59375,
      "learning_rate": 0.00027862053841427875,
      "loss": 4.2169,
      "mean_token_accuracy": 0.27534106373786926,
      "num_tokens": 14352976.0,
      "step": 18005
    },
    {
      "entropy": 4.644236278533936,
      "epoch": 5.091885778908679,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00027851769071299146,
      "loss": 4.2018,
      "mean_token_accuracy": 0.27857192158699035,
      "num_tokens": 14356979.0,
      "step": 18010
    },
    {
      "entropy": 4.667098903656006,
      "epoch": 5.093299406276506,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002784148422765284,
      "loss": 4.3187,
      "mean_token_accuracy": 0.2697510153055191,
      "num_tokens": 14361057.0,
      "step": 18015
    },
    {
      "entropy": 4.66478796005249,
      "epoch": 5.094713033644331,
      "grad_norm": 1.65625,
      "learning_rate": 0.00027831199312638457,
      "loss": 4.2908,
      "mean_token_accuracy": 0.2676544815301895,
      "num_tokens": 14365117.0,
      "step": 18020
    },
    {
      "entropy": 4.614839315414429,
      "epoch": 5.096126661012157,
      "grad_norm": 1.796875,
      "learning_rate": 0.00027820914328405456,
      "loss": 4.1979,
      "mean_token_accuracy": 0.2805491715669632,
      "num_tokens": 14369319.0,
      "step": 18025
    },
    {
      "entropy": 4.607352066040039,
      "epoch": 5.097540288379983,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00027810629277103357,
      "loss": 4.0953,
      "mean_token_accuracy": 0.2890956774353981,
      "num_tokens": 14372957.0,
      "step": 18030
    },
    {
      "entropy": 4.603690910339355,
      "epoch": 5.098953915747809,
      "grad_norm": 1.625,
      "learning_rate": 0.0002780034416088165,
      "loss": 4.1623,
      "mean_token_accuracy": 0.28051307052373886,
      "num_tokens": 14376750.0,
      "step": 18035
    },
    {
      "entropy": 4.639180994033813,
      "epoch": 5.100367543115635,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002779005898188987,
      "loss": 4.2273,
      "mean_token_accuracy": 0.2664508432149887,
      "num_tokens": 14380594.0,
      "step": 18040
    },
    {
      "entropy": 4.6327508926391605,
      "epoch": 5.10178117048346,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00027779773742277555,
      "loss": 4.2344,
      "mean_token_accuracy": 0.2702344208955765,
      "num_tokens": 14384595.0,
      "step": 18045
    },
    {
      "entropy": 4.592365741729736,
      "epoch": 5.103194797851287,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002776948844419427,
      "loss": 4.1983,
      "mean_token_accuracy": 0.2797009602189064,
      "num_tokens": 14388417.0,
      "step": 18050
    },
    {
      "entropy": 4.6856608390808105,
      "epoch": 5.104608425219112,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002775920308978955,
      "loss": 4.1807,
      "mean_token_accuracy": 0.2833316132426262,
      "num_tokens": 14392319.0,
      "step": 18055
    },
    {
      "entropy": 4.584122610092163,
      "epoch": 5.106022052586938,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00027748917681213005,
      "loss": 4.1903,
      "mean_token_accuracy": 0.27874227464199064,
      "num_tokens": 14395979.0,
      "step": 18060
    },
    {
      "entropy": 4.705148315429687,
      "epoch": 5.107435679954764,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00027738632220614183,
      "loss": 4.2809,
      "mean_token_accuracy": 0.2679614141583443,
      "num_tokens": 14400255.0,
      "step": 18065
    },
    {
      "entropy": 4.602199077606201,
      "epoch": 5.10884930732259,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0002772834671014272,
      "loss": 4.2751,
      "mean_token_accuracy": 0.27031530290842054,
      "num_tokens": 14404600.0,
      "step": 18070
    },
    {
      "entropy": 4.602102565765381,
      "epoch": 5.110262934690415,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000277180611519482,
      "loss": 4.2539,
      "mean_token_accuracy": 0.2729628786444664,
      "num_tokens": 14408737.0,
      "step": 18075
    },
    {
      "entropy": 4.682097387313843,
      "epoch": 5.111676562058242,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002770777554818025,
      "loss": 4.4046,
      "mean_token_accuracy": 0.26826545000076296,
      "num_tokens": 14413086.0,
      "step": 18080
    },
    {
      "entropy": 4.595051717758179,
      "epoch": 5.113090189426067,
      "grad_norm": 1.625,
      "learning_rate": 0.0002769748990098849,
      "loss": 4.1903,
      "mean_token_accuracy": 0.28434537947177885,
      "num_tokens": 14416960.0,
      "step": 18085
    },
    {
      "entropy": 4.636078548431397,
      "epoch": 5.114503816793893,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002768720421252258,
      "loss": 4.1382,
      "mean_token_accuracy": 0.2847408652305603,
      "num_tokens": 14420992.0,
      "step": 18090
    },
    {
      "entropy": 4.661658716201782,
      "epoch": 5.115917444161719,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00027676918484932124,
      "loss": 4.3378,
      "mean_token_accuracy": 0.26819100230932236,
      "num_tokens": 14425345.0,
      "step": 18095
    },
    {
      "entropy": 4.578392219543457,
      "epoch": 5.117331071529545,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002766663272036681,
      "loss": 4.1565,
      "mean_token_accuracy": 0.28541643768548963,
      "num_tokens": 14429247.0,
      "step": 18100
    },
    {
      "entropy": 4.625102615356445,
      "epoch": 5.118744698897371,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00027656346920976286,
      "loss": 4.1922,
      "mean_token_accuracy": 0.27022860646247865,
      "num_tokens": 14433145.0,
      "step": 18105
    },
    {
      "entropy": 4.672838258743286,
      "epoch": 5.120158326265196,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002764606108891022,
      "loss": 4.2711,
      "mean_token_accuracy": 0.26948482543230057,
      "num_tokens": 14437303.0,
      "step": 18110
    },
    {
      "entropy": 4.667522811889649,
      "epoch": 5.121571953633023,
      "grad_norm": 1.734375,
      "learning_rate": 0.00027635775226318287,
      "loss": 4.3656,
      "mean_token_accuracy": 0.2555467501282692,
      "num_tokens": 14441343.0,
      "step": 18115
    },
    {
      "entropy": 4.567931222915649,
      "epoch": 5.122985581000848,
      "grad_norm": 1.6875,
      "learning_rate": 0.00027625489335350177,
      "loss": 4.1649,
      "mean_token_accuracy": 0.2738897934556007,
      "num_tokens": 14445405.0,
      "step": 18120
    },
    {
      "entropy": 4.6652360439300535,
      "epoch": 5.124399208368674,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002761520341815557,
      "loss": 4.2075,
      "mean_token_accuracy": 0.28114925473928454,
      "num_tokens": 14449246.0,
      "step": 18125
    },
    {
      "entropy": 4.643034839630127,
      "epoch": 5.1258128357365,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002760491747688416,
      "loss": 4.2426,
      "mean_token_accuracy": 0.26908164024353026,
      "num_tokens": 14453570.0,
      "step": 18130
    },
    {
      "entropy": 4.596049499511719,
      "epoch": 5.127226463104326,
      "grad_norm": 1.5625,
      "learning_rate": 0.00027594631513685656,
      "loss": 4.1243,
      "mean_token_accuracy": 0.29086785912513735,
      "num_tokens": 14457572.0,
      "step": 18135
    },
    {
      "entropy": 4.62825117111206,
      "epoch": 5.128640090472151,
      "grad_norm": 1.671875,
      "learning_rate": 0.00027584345530709737,
      "loss": 4.1511,
      "mean_token_accuracy": 0.2820554256439209,
      "num_tokens": 14461624.0,
      "step": 18140
    },
    {
      "entropy": 4.650904798507691,
      "epoch": 5.130053717839977,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002757405953010613,
      "loss": 4.3324,
      "mean_token_accuracy": 0.2676112487912178,
      "num_tokens": 14465793.0,
      "step": 18145
    },
    {
      "entropy": 4.457570600509643,
      "epoch": 5.1314673452078035,
      "grad_norm": 1.796875,
      "learning_rate": 0.0002756377351402454,
      "loss": 4.0839,
      "mean_token_accuracy": 0.29158367216587067,
      "num_tokens": 14469361.0,
      "step": 18150
    },
    {
      "entropy": 4.628597164154053,
      "epoch": 5.132880972575629,
      "grad_norm": 1.796875,
      "learning_rate": 0.00027553487484614683,
      "loss": 4.3024,
      "mean_token_accuracy": 0.2640903204679489,
      "num_tokens": 14473220.0,
      "step": 18155
    },
    {
      "entropy": 4.614457654953003,
      "epoch": 5.134294599943455,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002754320144402626,
      "loss": 4.2226,
      "mean_token_accuracy": 0.26706048250198366,
      "num_tokens": 14477219.0,
      "step": 18160
    },
    {
      "entropy": 4.5794867992401125,
      "epoch": 5.135708227311281,
      "grad_norm": 1.671875,
      "learning_rate": 0.00027532915394409016,
      "loss": 4.1816,
      "mean_token_accuracy": 0.2868286997079849,
      "num_tokens": 14481230.0,
      "step": 18165
    },
    {
      "entropy": 4.67985463142395,
      "epoch": 5.137121854679107,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002752262933791265,
      "loss": 4.3097,
      "mean_token_accuracy": 0.27596009224653245,
      "num_tokens": 14485230.0,
      "step": 18170
    },
    {
      "entropy": 4.555800724029541,
      "epoch": 5.138535482046932,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0002751234327668689,
      "loss": 4.2019,
      "mean_token_accuracy": 0.28304718136787416,
      "num_tokens": 14489220.0,
      "step": 18175
    },
    {
      "entropy": 4.674853801727295,
      "epoch": 5.139949109414759,
      "grad_norm": 1.640625,
      "learning_rate": 0.00027502057212881476,
      "loss": 4.2865,
      "mean_token_accuracy": 0.2661323085427284,
      "num_tokens": 14493391.0,
      "step": 18180
    },
    {
      "entropy": 4.568795967102051,
      "epoch": 5.141362736782584,
      "grad_norm": 1.6484375,
      "learning_rate": 0.000274917711486461,
      "loss": 4.0669,
      "mean_token_accuracy": 0.29125612378120425,
      "num_tokens": 14497593.0,
      "step": 18185
    },
    {
      "entropy": 4.557483339309693,
      "epoch": 5.14277636415041,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00027481485086130515,
      "loss": 4.1818,
      "mean_token_accuracy": 0.27145916223526,
      "num_tokens": 14501181.0,
      "step": 18190
    },
    {
      "entropy": 4.608134984970093,
      "epoch": 5.144189991518235,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002747119902748443,
      "loss": 4.2762,
      "mean_token_accuracy": 0.26403513103723525,
      "num_tokens": 14505063.0,
      "step": 18195
    },
    {
      "entropy": 4.662157726287842,
      "epoch": 5.145603618886062,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0002746091297485756,
      "loss": 4.3004,
      "mean_token_accuracy": 0.2676041230559349,
      "num_tokens": 14508929.0,
      "step": 18200
    },
    {
      "entropy": 4.598190450668335,
      "epoch": 5.1470172462538875,
      "grad_norm": 1.625,
      "learning_rate": 0.00027450626930399647,
      "loss": 4.2104,
      "mean_token_accuracy": 0.2749990612268448,
      "num_tokens": 14512875.0,
      "step": 18205
    },
    {
      "entropy": 4.5629970073699955,
      "epoch": 5.148430873621713,
      "grad_norm": 1.609375,
      "learning_rate": 0.000274403408962604,
      "loss": 4.1595,
      "mean_token_accuracy": 0.28572010397911074,
      "num_tokens": 14517327.0,
      "step": 18210
    },
    {
      "entropy": 4.67906174659729,
      "epoch": 5.1498445009895395,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002743005487458954,
      "loss": 4.3376,
      "mean_token_accuracy": 0.2638292297720909,
      "num_tokens": 14521607.0,
      "step": 18215
    },
    {
      "entropy": 4.546378755569458,
      "epoch": 5.151258128357365,
      "grad_norm": 1.640625,
      "learning_rate": 0.00027419768867536766,
      "loss": 4.0682,
      "mean_token_accuracy": 0.28990632444620135,
      "num_tokens": 14525501.0,
      "step": 18220
    },
    {
      "entropy": 4.614981842041016,
      "epoch": 5.152671755725191,
      "grad_norm": 1.796875,
      "learning_rate": 0.0002740948287725181,
      "loss": 4.2126,
      "mean_token_accuracy": 0.2858993232250214,
      "num_tokens": 14529312.0,
      "step": 18225
    },
    {
      "entropy": 4.517384481430054,
      "epoch": 5.154085383093017,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00027399196905884376,
      "loss": 4.1169,
      "mean_token_accuracy": 0.28115067481994627,
      "num_tokens": 14533324.0,
      "step": 18230
    },
    {
      "entropy": 4.564222526550293,
      "epoch": 5.155499010460843,
      "grad_norm": 1.65625,
      "learning_rate": 0.00027388910955584165,
      "loss": 4.1934,
      "mean_token_accuracy": 0.2764231815934181,
      "num_tokens": 14537468.0,
      "step": 18235
    },
    {
      "entropy": 4.601641416549683,
      "epoch": 5.156912637828668,
      "grad_norm": 1.65625,
      "learning_rate": 0.00027378625028500874,
      "loss": 4.1882,
      "mean_token_accuracy": 0.26902192980051043,
      "num_tokens": 14541719.0,
      "step": 18240
    },
    {
      "entropy": 4.689247465133667,
      "epoch": 5.158326265196494,
      "grad_norm": 1.734375,
      "learning_rate": 0.000273683391267842,
      "loss": 4.284,
      "mean_token_accuracy": 0.2633858650922775,
      "num_tokens": 14546007.0,
      "step": 18245
    },
    {
      "entropy": 4.750868463516236,
      "epoch": 5.15973989256432,
      "grad_norm": 1.71875,
      "learning_rate": 0.00027358053252583836,
      "loss": 4.3154,
      "mean_token_accuracy": 0.266945880651474,
      "num_tokens": 14550307.0,
      "step": 18250
    },
    {
      "entropy": 4.621160173416138,
      "epoch": 5.161153519932146,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00027347767408049465,
      "loss": 4.2374,
      "mean_token_accuracy": 0.27726053446531296,
      "num_tokens": 14554279.0,
      "step": 18255
    },
    {
      "entropy": 4.574586820602417,
      "epoch": 5.1625671472999715,
      "grad_norm": 1.875,
      "learning_rate": 0.0002733748159533077,
      "loss": 4.1822,
      "mean_token_accuracy": 0.2718992680311203,
      "num_tokens": 14558217.0,
      "step": 18260
    },
    {
      "entropy": 4.663677072525024,
      "epoch": 5.163980774667798,
      "grad_norm": 1.6640625,
      "learning_rate": 0.000273271958165774,
      "loss": 4.4041,
      "mean_token_accuracy": 0.24852942526340485,
      "num_tokens": 14562437.0,
      "step": 18265
    },
    {
      "entropy": 4.628178977966309,
      "epoch": 5.1653944020356235,
      "grad_norm": 1.6875,
      "learning_rate": 0.00027316910073939054,
      "loss": 4.2506,
      "mean_token_accuracy": 0.2729086816310883,
      "num_tokens": 14566402.0,
      "step": 18270
    },
    {
      "entropy": 4.636946439743042,
      "epoch": 5.166808029403449,
      "grad_norm": 1.609375,
      "learning_rate": 0.00027306624369565364,
      "loss": 4.233,
      "mean_token_accuracy": 0.2820824235677719,
      "num_tokens": 14570331.0,
      "step": 18275
    },
    {
      "entropy": 4.674766254425049,
      "epoch": 5.168221656771275,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002729633870560599,
      "loss": 4.2301,
      "mean_token_accuracy": 0.27533732652664183,
      "num_tokens": 14574152.0,
      "step": 18280
    },
    {
      "entropy": 4.554713487625122,
      "epoch": 5.169635284139101,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002728605308421055,
      "loss": 4.1029,
      "mean_token_accuracy": 0.28770689219236373,
      "num_tokens": 14578169.0,
      "step": 18285
    },
    {
      "entropy": 4.673542881011963,
      "epoch": 5.171048911506927,
      "grad_norm": 1.6875,
      "learning_rate": 0.00027275767507528706,
      "loss": 4.3181,
      "mean_token_accuracy": 0.2688418492674828,
      "num_tokens": 14582329.0,
      "step": 18290
    },
    {
      "entropy": 4.644868326187134,
      "epoch": 5.172462538874752,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00027265481977710064,
      "loss": 4.2999,
      "mean_token_accuracy": 0.2675821021199226,
      "num_tokens": 14586529.0,
      "step": 18295
    },
    {
      "entropy": 4.703611993789673,
      "epoch": 5.173876166242579,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002725519649690424,
      "loss": 4.3118,
      "mean_token_accuracy": 0.2683951273560524,
      "num_tokens": 14590430.0,
      "step": 18300
    },
    {
      "entropy": 4.746325349807739,
      "epoch": 5.175289793610404,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002724491106726083,
      "loss": 4.3288,
      "mean_token_accuracy": 0.25284790694713594,
      "num_tokens": 14594291.0,
      "step": 18305
    },
    {
      "entropy": 4.591769504547119,
      "epoch": 5.17670342097823,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002723462569092943,
      "loss": 4.1898,
      "mean_token_accuracy": 0.28290189802646637,
      "num_tokens": 14598150.0,
      "step": 18310
    },
    {
      "entropy": 4.708440542221069,
      "epoch": 5.178117048346056,
      "grad_norm": 1.59375,
      "learning_rate": 0.00027224340370059626,
      "loss": 4.3289,
      "mean_token_accuracy": 0.2689444959163666,
      "num_tokens": 14602208.0,
      "step": 18315
    },
    {
      "entropy": 4.573243808746338,
      "epoch": 5.179530675713882,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002721405510680098,
      "loss": 4.1936,
      "mean_token_accuracy": 0.27735846787691115,
      "num_tokens": 14606250.0,
      "step": 18320
    },
    {
      "entropy": 4.71888689994812,
      "epoch": 5.1809443030817075,
      "grad_norm": 1.640625,
      "learning_rate": 0.00027203769903303034,
      "loss": 4.2783,
      "mean_token_accuracy": 0.2736815869808197,
      "num_tokens": 14610074.0,
      "step": 18325
    },
    {
      "entropy": 4.65406608581543,
      "epoch": 5.182357930449533,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00027193484761715354,
      "loss": 4.1627,
      "mean_token_accuracy": 0.279283432662487,
      "num_tokens": 14614146.0,
      "step": 18330
    },
    {
      "entropy": 4.626205396652222,
      "epoch": 5.1837715578173595,
      "grad_norm": 1.734375,
      "learning_rate": 0.0002718319968418747,
      "loss": 4.2701,
      "mean_token_accuracy": 0.2753400236368179,
      "num_tokens": 14618239.0,
      "step": 18335
    },
    {
      "entropy": 4.603142499923706,
      "epoch": 5.185185185185185,
      "grad_norm": 1.71875,
      "learning_rate": 0.00027172914672868895,
      "loss": 4.1792,
      "mean_token_accuracy": 0.2760543435811996,
      "num_tokens": 14622400.0,
      "step": 18340
    },
    {
      "entropy": 4.652845048904419,
      "epoch": 5.186598812553011,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002716262972990912,
      "loss": 4.2476,
      "mean_token_accuracy": 0.27162855714559553,
      "num_tokens": 14626392.0,
      "step": 18345
    },
    {
      "entropy": 4.6030449867248535,
      "epoch": 5.188012439920837,
      "grad_norm": 1.703125,
      "learning_rate": 0.00027152344857457643,
      "loss": 4.2257,
      "mean_token_accuracy": 0.28411724269390104,
      "num_tokens": 14630558.0,
      "step": 18350
    },
    {
      "entropy": 4.576498651504517,
      "epoch": 5.189426067288663,
      "grad_norm": 1.703125,
      "learning_rate": 0.00027142060057663955,
      "loss": 4.1392,
      "mean_token_accuracy": 0.27743361592292787,
      "num_tokens": 14634450.0,
      "step": 18355
    },
    {
      "entropy": 4.609848976135254,
      "epoch": 5.190839694656488,
      "grad_norm": 1.734375,
      "learning_rate": 0.00027131775332677496,
      "loss": 4.2265,
      "mean_token_accuracy": 0.28096762150526045,
      "num_tokens": 14638177.0,
      "step": 18360
    },
    {
      "entropy": 4.593127822875976,
      "epoch": 5.192253322024315,
      "grad_norm": 1.65625,
      "learning_rate": 0.00027121490684647707,
      "loss": 4.2301,
      "mean_token_accuracy": 0.2799376532435417,
      "num_tokens": 14642098.0,
      "step": 18365
    },
    {
      "entropy": 4.5574421882629395,
      "epoch": 5.19366694939214,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00027111206115724035,
      "loss": 4.2461,
      "mean_token_accuracy": 0.279080955684185,
      "num_tokens": 14646031.0,
      "step": 18370
    },
    {
      "entropy": 4.5973968505859375,
      "epoch": 5.195080576759966,
      "grad_norm": 1.765625,
      "learning_rate": 0.00027100921628055874,
      "loss": 4.1972,
      "mean_token_accuracy": 0.27076782286167145,
      "num_tokens": 14650253.0,
      "step": 18375
    },
    {
      "entropy": 4.658344888687134,
      "epoch": 5.1964942041277915,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002709063722379263,
      "loss": 4.2627,
      "mean_token_accuracy": 0.27380903661251066,
      "num_tokens": 14654229.0,
      "step": 18380
    },
    {
      "entropy": 4.6030689716339115,
      "epoch": 5.197907831495618,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002708035290508367,
      "loss": 4.2145,
      "mean_token_accuracy": 0.28216633200645447,
      "num_tokens": 14658239.0,
      "step": 18385
    },
    {
      "entropy": 4.633175420761108,
      "epoch": 5.1993214588634435,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00027070068674078356,
      "loss": 4.1902,
      "mean_token_accuracy": 0.28344414234161375,
      "num_tokens": 14662280.0,
      "step": 18390
    },
    {
      "entropy": 4.673092889785766,
      "epoch": 5.200735086231269,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002705978453292604,
      "loss": 4.2995,
      "mean_token_accuracy": 0.2824153542518616,
      "num_tokens": 14666160.0,
      "step": 18395
    },
    {
      "entropy": 4.631181526184082,
      "epoch": 5.202148713599096,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002704950048377603,
      "loss": 4.0901,
      "mean_token_accuracy": 0.28358401358127594,
      "num_tokens": 14670121.0,
      "step": 18400
    },
    {
      "entropy": 4.703688049316407,
      "epoch": 5.203562340966921,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00027039216528777625,
      "loss": 4.2546,
      "mean_token_accuracy": 0.26633368581533434,
      "num_tokens": 14674193.0,
      "step": 18405
    },
    {
      "entropy": 4.592945766448975,
      "epoch": 5.204975968334747,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00027028932670080123,
      "loss": 4.1569,
      "mean_token_accuracy": 0.28913151323795316,
      "num_tokens": 14678012.0,
      "step": 18410
    },
    {
      "entropy": 4.728510808944702,
      "epoch": 5.206389595702573,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002701864890983278,
      "loss": 4.2832,
      "mean_token_accuracy": 0.2715243771672249,
      "num_tokens": 14682137.0,
      "step": 18415
    },
    {
      "entropy": 4.591843175888061,
      "epoch": 5.207803223070399,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00027008365250184846,
      "loss": 4.2041,
      "mean_token_accuracy": 0.28303227871656417,
      "num_tokens": 14686290.0,
      "step": 18420
    },
    {
      "entropy": 4.646112298965454,
      "epoch": 5.209216850438224,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002699808169328552,
      "loss": 4.1664,
      "mean_token_accuracy": 0.2743216946721077,
      "num_tokens": 14690152.0,
      "step": 18425
    },
    {
      "entropy": 4.570328903198242,
      "epoch": 5.21063047780605,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0002698779824128402,
      "loss": 4.1733,
      "mean_token_accuracy": 0.2766466587781906,
      "num_tokens": 14694072.0,
      "step": 18430
    },
    {
      "entropy": 4.70737943649292,
      "epoch": 5.212044105173876,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002697751489632953,
      "loss": 4.2658,
      "mean_token_accuracy": 0.27266574949026107,
      "num_tokens": 14698069.0,
      "step": 18435
    },
    {
      "entropy": 4.663687372207642,
      "epoch": 5.213457732541702,
      "grad_norm": 1.6875,
      "learning_rate": 0.00026967231660571196,
      "loss": 4.2018,
      "mean_token_accuracy": 0.29125762581825254,
      "num_tokens": 14702036.0,
      "step": 18440
    },
    {
      "entropy": 4.662603855133057,
      "epoch": 5.2148713599095275,
      "grad_norm": 1.75,
      "learning_rate": 0.0002695694853615813,
      "loss": 4.253,
      "mean_token_accuracy": 0.2822985857725143,
      "num_tokens": 14705984.0,
      "step": 18445
    },
    {
      "entropy": 4.62885684967041,
      "epoch": 5.216284987277354,
      "grad_norm": 1.671875,
      "learning_rate": 0.00026946665525239477,
      "loss": 4.1326,
      "mean_token_accuracy": 0.29303757548332215,
      "num_tokens": 14710224.0,
      "step": 18450
    },
    {
      "entropy": 4.655870532989502,
      "epoch": 5.21769861464518,
      "grad_norm": 1.703125,
      "learning_rate": 0.00026936382629964313,
      "loss": 4.3046,
      "mean_token_accuracy": 0.26066340804100036,
      "num_tokens": 14714236.0,
      "step": 18455
    },
    {
      "entropy": 4.626061248779297,
      "epoch": 5.219112242013005,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002692609985248169,
      "loss": 4.1142,
      "mean_token_accuracy": 0.2891070291399956,
      "num_tokens": 14718310.0,
      "step": 18460
    },
    {
      "entropy": 4.677752065658569,
      "epoch": 5.220525869380832,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002691581719494064,
      "loss": 4.2072,
      "mean_token_accuracy": 0.279716756939888,
      "num_tokens": 14722076.0,
      "step": 18465
    },
    {
      "entropy": 4.60687780380249,
      "epoch": 5.221939496748657,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002690553465949019,
      "loss": 4.2418,
      "mean_token_accuracy": 0.2797772720456123,
      "num_tokens": 14726003.0,
      "step": 18470
    },
    {
      "entropy": 4.675338649749756,
      "epoch": 5.223353124116483,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002689525224827932,
      "loss": 4.2647,
      "mean_token_accuracy": 0.27072785794734955,
      "num_tokens": 14729859.0,
      "step": 18475
    },
    {
      "entropy": 4.5966472148895265,
      "epoch": 5.224766751484308,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002688496996345699,
      "loss": 4.2032,
      "mean_token_accuracy": 0.26913378983736036,
      "num_tokens": 14734011.0,
      "step": 18480
    },
    {
      "entropy": 4.692791175842285,
      "epoch": 5.226180378852135,
      "grad_norm": 1.625,
      "learning_rate": 0.00026874687807172134,
      "loss": 4.2316,
      "mean_token_accuracy": 0.27154673635959625,
      "num_tokens": 14737866.0,
      "step": 18485
    },
    {
      "entropy": 4.727388906478882,
      "epoch": 5.22759400621996,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0002686440578157366,
      "loss": 4.2774,
      "mean_token_accuracy": 0.2619865283370018,
      "num_tokens": 14741762.0,
      "step": 18490
    },
    {
      "entropy": 4.71250958442688,
      "epoch": 5.229007633587786,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002685412388881045,
      "loss": 4.1966,
      "mean_token_accuracy": 0.28462516963481904,
      "num_tokens": 14745936.0,
      "step": 18495
    },
    {
      "entropy": 4.672093391418457,
      "epoch": 5.230421260955612,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002684384213103135,
      "loss": 4.2632,
      "mean_token_accuracy": 0.2812881335616112,
      "num_tokens": 14750044.0,
      "step": 18500
    },
    {
      "entropy": 4.55256199836731,
      "epoch": 5.231834888323438,
      "grad_norm": 1.75,
      "learning_rate": 0.0002683356051038518,
      "loss": 4.0896,
      "mean_token_accuracy": 0.29661971479654314,
      "num_tokens": 14753796.0,
      "step": 18505
    },
    {
      "entropy": 4.573391628265381,
      "epoch": 5.233248515691264,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00026823279029020755,
      "loss": 4.2405,
      "mean_token_accuracy": 0.275679449737072,
      "num_tokens": 14757861.0,
      "step": 18510
    },
    {
      "entropy": 4.511469316482544,
      "epoch": 5.23466214305909,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002681299768908682,
      "loss": 4.1916,
      "mean_token_accuracy": 0.2777367129921913,
      "num_tokens": 14761624.0,
      "step": 18515
    },
    {
      "entropy": 4.644030475616455,
      "epoch": 5.236075770426916,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0002680271649273213,
      "loss": 4.2153,
      "mean_token_accuracy": 0.27331595718860624,
      "num_tokens": 14765590.0,
      "step": 18520
    },
    {
      "entropy": 4.660901308059692,
      "epoch": 5.237489397794741,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002679243544210537,
      "loss": 4.2087,
      "mean_token_accuracy": 0.2706322863698006,
      "num_tokens": 14769655.0,
      "step": 18525
    },
    {
      "entropy": 4.690081691741943,
      "epoch": 5.238903025162567,
      "grad_norm": 1.84375,
      "learning_rate": 0.0002678215453935523,
      "loss": 4.241,
      "mean_token_accuracy": 0.28086629807949065,
      "num_tokens": 14773696.0,
      "step": 18530
    },
    {
      "entropy": 4.632171535491944,
      "epoch": 5.240316652530393,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002677187378663035,
      "loss": 4.2079,
      "mean_token_accuracy": 0.2811109319329262,
      "num_tokens": 14777654.0,
      "step": 18535
    },
    {
      "entropy": 4.633652782440185,
      "epoch": 5.241730279898219,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002676159318607935,
      "loss": 4.2534,
      "mean_token_accuracy": 0.2746643215417862,
      "num_tokens": 14781736.0,
      "step": 18540
    },
    {
      "entropy": 4.636980772018433,
      "epoch": 5.243143907266044,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002675131273985081,
      "loss": 4.233,
      "mean_token_accuracy": 0.2708143562078476,
      "num_tokens": 14785645.0,
      "step": 18545
    },
    {
      "entropy": 4.7187553405761715,
      "epoch": 5.244557534633871,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00026741032450093273,
      "loss": 4.3996,
      "mean_token_accuracy": 0.2557079210877419,
      "num_tokens": 14789820.0,
      "step": 18550
    },
    {
      "entropy": 4.722614860534668,
      "epoch": 5.245971162001696,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00026730752318955266,
      "loss": 4.2593,
      "mean_token_accuracy": 0.2720838934183121,
      "num_tokens": 14793630.0,
      "step": 18555
    },
    {
      "entropy": 4.69912281036377,
      "epoch": 5.247384789369522,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00026720472348585266,
      "loss": 4.303,
      "mean_token_accuracy": 0.2673646673560143,
      "num_tokens": 14797803.0,
      "step": 18560
    },
    {
      "entropy": 4.624587631225586,
      "epoch": 5.248798416737348,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00026710192541131724,
      "loss": 4.151,
      "mean_token_accuracy": 0.28393644392490386,
      "num_tokens": 14801862.0,
      "step": 18565
    },
    {
      "entropy": 4.647853517532349,
      "epoch": 5.250212044105174,
      "grad_norm": 1.78125,
      "learning_rate": 0.00026699912898743046,
      "loss": 4.2276,
      "mean_token_accuracy": 0.27198379635810854,
      "num_tokens": 14806003.0,
      "step": 18570
    },
    {
      "entropy": 4.600250005722046,
      "epoch": 5.251625671473,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002668963342356764,
      "loss": 4.0979,
      "mean_token_accuracy": 0.28832162022590635,
      "num_tokens": 14809650.0,
      "step": 18575
    },
    {
      "entropy": 4.701062202453613,
      "epoch": 5.253039298840825,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00026679354117753836,
      "loss": 4.243,
      "mean_token_accuracy": 0.2655442088842392,
      "num_tokens": 14813410.0,
      "step": 18580
    },
    {
      "entropy": 4.712422275543213,
      "epoch": 5.254452926208652,
      "grad_norm": 1.828125,
      "learning_rate": 0.00026669074983449943,
      "loss": 4.2833,
      "mean_token_accuracy": 0.2708080753684044,
      "num_tokens": 14817383.0,
      "step": 18585
    },
    {
      "entropy": 4.605118227005005,
      "epoch": 5.255866553576477,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00026658796022804225,
      "loss": 4.1636,
      "mean_token_accuracy": 0.2790886893868446,
      "num_tokens": 14821191.0,
      "step": 18590
    },
    {
      "entropy": 4.627947044372559,
      "epoch": 5.257280180944303,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00026648517237964957,
      "loss": 4.1608,
      "mean_token_accuracy": 0.2807081341743469,
      "num_tokens": 14825171.0,
      "step": 18595
    },
    {
      "entropy": 4.631296110153198,
      "epoch": 5.258693808312129,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00026638238631080313,
      "loss": 4.2291,
      "mean_token_accuracy": 0.2772295907139778,
      "num_tokens": 14829359.0,
      "step": 18600
    },
    {
      "entropy": 4.6267928123474125,
      "epoch": 5.260107435679955,
      "grad_norm": 1.625,
      "learning_rate": 0.00026627960204298476,
      "loss": 4.2387,
      "mean_token_accuracy": 0.2717928856611252,
      "num_tokens": 14833199.0,
      "step": 18605
    },
    {
      "entropy": 4.53325572013855,
      "epoch": 5.2615210630477804,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002661768195976755,
      "loss": 4.0858,
      "mean_token_accuracy": 0.2916944414377213,
      "num_tokens": 14836953.0,
      "step": 18610
    },
    {
      "entropy": 4.608791303634644,
      "epoch": 5.262934690415607,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002660740389963565,
      "loss": 4.2462,
      "mean_token_accuracy": 0.27025809437036513,
      "num_tokens": 14840861.0,
      "step": 18615
    },
    {
      "entropy": 4.642741775512695,
      "epoch": 5.2643483177834325,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00026597126026050807,
      "loss": 4.2762,
      "mean_token_accuracy": 0.2764749780297279,
      "num_tokens": 14844766.0,
      "step": 18620
    },
    {
      "entropy": 4.563149642944336,
      "epoch": 5.265761945151258,
      "grad_norm": 1.796875,
      "learning_rate": 0.00026586848341161043,
      "loss": 4.1878,
      "mean_token_accuracy": 0.2792450666427612,
      "num_tokens": 14848645.0,
      "step": 18625
    },
    {
      "entropy": 4.670603275299072,
      "epoch": 5.267175572519084,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002657657084711432,
      "loss": 4.219,
      "mean_token_accuracy": 0.2804061099886894,
      "num_tokens": 14852504.0,
      "step": 18630
    },
    {
      "entropy": 4.608864641189575,
      "epoch": 5.26858919988691,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002656629354605859,
      "loss": 4.2228,
      "mean_token_accuracy": 0.2806055650115013,
      "num_tokens": 14856617.0,
      "step": 18635
    },
    {
      "entropy": 4.587740564346314,
      "epoch": 5.270002827254736,
      "grad_norm": 1.625,
      "learning_rate": 0.0002655601644014173,
      "loss": 4.1645,
      "mean_token_accuracy": 0.29293579459190366,
      "num_tokens": 14860410.0,
      "step": 18640
    },
    {
      "entropy": 4.725992202758789,
      "epoch": 5.271416454622561,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00026545739531511607,
      "loss": 4.3162,
      "mean_token_accuracy": 0.2622071847319603,
      "num_tokens": 14864439.0,
      "step": 18645
    },
    {
      "entropy": 4.690218114852906,
      "epoch": 5.272830081990388,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002653546282231601,
      "loss": 4.2566,
      "mean_token_accuracy": 0.2678606897592545,
      "num_tokens": 14869031.0,
      "step": 18650
    },
    {
      "entropy": 4.671438217163086,
      "epoch": 5.274243709358213,
      "grad_norm": 1.765625,
      "learning_rate": 0.00026525186314702725,
      "loss": 4.2078,
      "mean_token_accuracy": 0.28286716938018797,
      "num_tokens": 14872977.0,
      "step": 18655
    },
    {
      "entropy": 4.6507518768310545,
      "epoch": 5.275657336726039,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002651491001081947,
      "loss": 4.226,
      "mean_token_accuracy": 0.27823057770729065,
      "num_tokens": 14877385.0,
      "step": 18660
    },
    {
      "entropy": 4.581309604644775,
      "epoch": 5.2770709640938644,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00026504633912813937,
      "loss": 4.1753,
      "mean_token_accuracy": 0.28859920501708985,
      "num_tokens": 14881467.0,
      "step": 18665
    },
    {
      "entropy": 4.580193996429443,
      "epoch": 5.278484591461691,
      "grad_norm": 1.703125,
      "learning_rate": 0.00026494358022833745,
      "loss": 4.1377,
      "mean_token_accuracy": 0.2820078134536743,
      "num_tokens": 14885431.0,
      "step": 18670
    },
    {
      "entropy": 4.602320909500122,
      "epoch": 5.2798982188295165,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026484082343026524,
      "loss": 4.1985,
      "mean_token_accuracy": 0.2842243745923042,
      "num_tokens": 14889650.0,
      "step": 18675
    },
    {
      "entropy": 4.58087797164917,
      "epoch": 5.281311846197342,
      "grad_norm": 1.640625,
      "learning_rate": 0.00026473806875539806,
      "loss": 4.1815,
      "mean_token_accuracy": 0.2822600841522217,
      "num_tokens": 14893968.0,
      "step": 18680
    },
    {
      "entropy": 4.6389039039611815,
      "epoch": 5.2827254735651685,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002646353162252111,
      "loss": 4.2746,
      "mean_token_accuracy": 0.2711835041642189,
      "num_tokens": 14898066.0,
      "step": 18685
    },
    {
      "entropy": 4.586738157272339,
      "epoch": 5.284139100932994,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002645325658611788,
      "loss": 4.1327,
      "mean_token_accuracy": 0.276435324549675,
      "num_tokens": 14901976.0,
      "step": 18690
    },
    {
      "entropy": 4.633464670181274,
      "epoch": 5.28555272830082,
      "grad_norm": 1.640625,
      "learning_rate": 0.00026442981768477563,
      "loss": 4.2544,
      "mean_token_accuracy": 0.2773670494556427,
      "num_tokens": 14906159.0,
      "step": 18695
    },
    {
      "entropy": 4.631740140914917,
      "epoch": 5.286966355668646,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002643270717174752,
      "loss": 4.1552,
      "mean_token_accuracy": 0.2795482039451599,
      "num_tokens": 14909818.0,
      "step": 18700
    },
    {
      "entropy": 4.563196754455566,
      "epoch": 5.288379983036472,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002642243279807508,
      "loss": 4.211,
      "mean_token_accuracy": 0.28608210384845734,
      "num_tokens": 14913882.0,
      "step": 18705
    },
    {
      "entropy": 4.618905782699585,
      "epoch": 5.289793610404297,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002641215864960751,
      "loss": 4.3227,
      "mean_token_accuracy": 0.2632347777485847,
      "num_tokens": 14917667.0,
      "step": 18710
    },
    {
      "entropy": 4.646334075927735,
      "epoch": 5.291207237772123,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00026401884728492064,
      "loss": 4.2361,
      "mean_token_accuracy": 0.27219184935092927,
      "num_tokens": 14921602.0,
      "step": 18715
    },
    {
      "entropy": 4.649905157089234,
      "epoch": 5.292620865139949,
      "grad_norm": 1.71875,
      "learning_rate": 0.00026391611036875916,
      "loss": 4.3004,
      "mean_token_accuracy": 0.26781432181596754,
      "num_tokens": 14925341.0,
      "step": 18720
    },
    {
      "entropy": 4.673179960250854,
      "epoch": 5.294034492507775,
      "grad_norm": 1.609375,
      "learning_rate": 0.00026381337576906214,
      "loss": 4.2257,
      "mean_token_accuracy": 0.2805128574371338,
      "num_tokens": 14929468.0,
      "step": 18725
    },
    {
      "entropy": 4.623729801177978,
      "epoch": 5.2954481198756005,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002637106435073003,
      "loss": 4.2585,
      "mean_token_accuracy": 0.2691301226615906,
      "num_tokens": 14933386.0,
      "step": 18730
    },
    {
      "entropy": 4.595997142791748,
      "epoch": 5.296861747243427,
      "grad_norm": 1.734375,
      "learning_rate": 0.00026360791360494426,
      "loss": 4.3206,
      "mean_token_accuracy": 0.26249667406082156,
      "num_tokens": 14937407.0,
      "step": 18735
    },
    {
      "entropy": 4.6018718719482425,
      "epoch": 5.2982753746112525,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002635051860834638,
      "loss": 4.299,
      "mean_token_accuracy": 0.2677106007933617,
      "num_tokens": 14941791.0,
      "step": 18740
    },
    {
      "entropy": 4.567838907241821,
      "epoch": 5.299689001979078,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002634024609643283,
      "loss": 4.1502,
      "mean_token_accuracy": 0.28991774320602415,
      "num_tokens": 14945572.0,
      "step": 18745
    },
    {
      "entropy": 4.650605630874634,
      "epoch": 5.301102629346905,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002632997382690068,
      "loss": 4.2616,
      "mean_token_accuracy": 0.2745348662137985,
      "num_tokens": 14949868.0,
      "step": 18750
    },
    {
      "entropy": 4.590937948226928,
      "epoch": 5.30251625671473,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002631970180189676,
      "loss": 4.1928,
      "mean_token_accuracy": 0.28098822236061094,
      "num_tokens": 14953930.0,
      "step": 18755
    },
    {
      "entropy": 4.6213561534881595,
      "epoch": 5.303929884082556,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0002630943002356787,
      "loss": 4.1746,
      "mean_token_accuracy": 0.2907917559146881,
      "num_tokens": 14958254.0,
      "step": 18760
    },
    {
      "entropy": 4.743972206115723,
      "epoch": 5.305343511450381,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00026299158494060747,
      "loss": 4.429,
      "mean_token_accuracy": 0.2536447778344154,
      "num_tokens": 14962356.0,
      "step": 18765
    },
    {
      "entropy": 4.626659488677978,
      "epoch": 5.306757138818208,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002628888721552206,
      "loss": 4.2224,
      "mean_token_accuracy": 0.27343734055757524,
      "num_tokens": 14966211.0,
      "step": 18770
    },
    {
      "entropy": 4.632144165039063,
      "epoch": 5.308170766186033,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00026278616190098463,
      "loss": 4.1973,
      "mean_token_accuracy": 0.27903459817171095,
      "num_tokens": 14970018.0,
      "step": 18775
    },
    {
      "entropy": 4.6532481670379635,
      "epoch": 5.309584393553859,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00026268345419936534,
      "loss": 4.2265,
      "mean_token_accuracy": 0.2669056043028831,
      "num_tokens": 14973511.0,
      "step": 18780
    },
    {
      "entropy": 4.6394013404846195,
      "epoch": 5.310998020921685,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00026258074907182795,
      "loss": 4.2685,
      "mean_token_accuracy": 0.270871165394783,
      "num_tokens": 14977336.0,
      "step": 18785
    },
    {
      "entropy": 4.5096845626831055,
      "epoch": 5.312411648289511,
      "grad_norm": 1.671875,
      "learning_rate": 0.00026247804653983716,
      "loss": 4.0615,
      "mean_token_accuracy": 0.29352526366710663,
      "num_tokens": 14981136.0,
      "step": 18790
    },
    {
      "entropy": 4.605981636047363,
      "epoch": 5.3138252756573365,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002623753466248572,
      "loss": 4.1708,
      "mean_token_accuracy": 0.27835207879543306,
      "num_tokens": 14984983.0,
      "step": 18795
    },
    {
      "entropy": 4.616169357299805,
      "epoch": 5.315238903025163,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00026227264934835175,
      "loss": 4.1808,
      "mean_token_accuracy": 0.2928533762693405,
      "num_tokens": 14988798.0,
      "step": 18800
    },
    {
      "entropy": 4.506228256225586,
      "epoch": 5.316652530392989,
      "grad_norm": 1.59375,
      "learning_rate": 0.0002621699547317839,
      "loss": 4.1301,
      "mean_token_accuracy": 0.28412469327449796,
      "num_tokens": 14992716.0,
      "step": 18805
    },
    {
      "entropy": 4.581400775909424,
      "epoch": 5.318066157760814,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00026206726279661606,
      "loss": 4.1726,
      "mean_token_accuracy": 0.2808115780353546,
      "num_tokens": 14996872.0,
      "step": 18810
    },
    {
      "entropy": 4.546552038192749,
      "epoch": 5.31947978512864,
      "grad_norm": 1.640625,
      "learning_rate": 0.00026196457356431047,
      "loss": 4.2235,
      "mean_token_accuracy": 0.26792534440755844,
      "num_tokens": 15000622.0,
      "step": 18815
    },
    {
      "entropy": 4.649037647247314,
      "epoch": 5.320893412496466,
      "grad_norm": 1.828125,
      "learning_rate": 0.00026186188705632835,
      "loss": 4.2988,
      "mean_token_accuracy": 0.2687313675880432,
      "num_tokens": 15004748.0,
      "step": 18820
    },
    {
      "entropy": 4.619935131072998,
      "epoch": 5.322307039864292,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002617592032941306,
      "loss": 4.1816,
      "mean_token_accuracy": 0.28145519345998765,
      "num_tokens": 15008605.0,
      "step": 18825
    },
    {
      "entropy": 4.626696681976318,
      "epoch": 5.323720667232117,
      "grad_norm": 1.75,
      "learning_rate": 0.0002616565222991775,
      "loss": 4.1867,
      "mean_token_accuracy": 0.2812833070755005,
      "num_tokens": 15012280.0,
      "step": 18830
    },
    {
      "entropy": 4.678423833847046,
      "epoch": 5.325134294599944,
      "grad_norm": 1.5859375,
      "learning_rate": 0.00026155384409292867,
      "loss": 4.2886,
      "mean_token_accuracy": 0.27542456388473513,
      "num_tokens": 15016659.0,
      "step": 18835
    },
    {
      "entropy": 4.5982053756713865,
      "epoch": 5.326547921967769,
      "grad_norm": 1.8046875,
      "learning_rate": 0.0002614511686968433,
      "loss": 4.234,
      "mean_token_accuracy": 0.2766795545816422,
      "num_tokens": 15020630.0,
      "step": 18840
    },
    {
      "entropy": 4.734406471252441,
      "epoch": 5.327961549335595,
      "grad_norm": 1.890625,
      "learning_rate": 0.00026134849613237987,
      "loss": 4.3808,
      "mean_token_accuracy": 0.2551599845290184,
      "num_tokens": 15024555.0,
      "step": 18845
    },
    {
      "entropy": 4.711378479003907,
      "epoch": 5.3293751767034205,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00026124582642099637,
      "loss": 4.2772,
      "mean_token_accuracy": 0.271517014503479,
      "num_tokens": 15028319.0,
      "step": 18850
    },
    {
      "entropy": 4.649682331085205,
      "epoch": 5.330788804071247,
      "grad_norm": 1.78125,
      "learning_rate": 0.00026114315958415,
      "loss": 4.2311,
      "mean_token_accuracy": 0.28240356892347335,
      "num_tokens": 15032517.0,
      "step": 18855
    },
    {
      "entropy": 4.675660848617554,
      "epoch": 5.332202431439073,
      "grad_norm": 1.671875,
      "learning_rate": 0.00026104049564329776,
      "loss": 4.2017,
      "mean_token_accuracy": 0.27231002151966094,
      "num_tokens": 15036696.0,
      "step": 18860
    },
    {
      "entropy": 4.634500217437744,
      "epoch": 5.333616058806898,
      "grad_norm": 1.8359375,
      "learning_rate": 0.0002609378346198955,
      "loss": 4.2577,
      "mean_token_accuracy": 0.26580379605293275,
      "num_tokens": 15040658.0,
      "step": 18865
    },
    {
      "entropy": 4.710043621063233,
      "epoch": 5.335029686174725,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002608351765353988,
      "loss": 4.2769,
      "mean_token_accuracy": 0.2752949848771095,
      "num_tokens": 15044845.0,
      "step": 18870
    },
    {
      "entropy": 4.532862710952759,
      "epoch": 5.33644331354255,
      "grad_norm": 1.65625,
      "learning_rate": 0.00026073252141126265,
      "loss": 4.1753,
      "mean_token_accuracy": 0.2978943303227425,
      "num_tokens": 15049220.0,
      "step": 18875
    },
    {
      "entropy": 4.499542236328125,
      "epoch": 5.337856940910376,
      "grad_norm": 1.65625,
      "learning_rate": 0.00026062986926894135,
      "loss": 4.1802,
      "mean_token_accuracy": 0.29202671349048615,
      "num_tokens": 15053787.0,
      "step": 18880
    },
    {
      "entropy": 4.5772247314453125,
      "epoch": 5.339270568278202,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0002605272201298884,
      "loss": 4.2111,
      "mean_token_accuracy": 0.2846470981836319,
      "num_tokens": 15057610.0,
      "step": 18885
    },
    {
      "entropy": 4.639570522308349,
      "epoch": 5.340684195646028,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00026042457401555713,
      "loss": 4.209,
      "mean_token_accuracy": 0.28194244801998136,
      "num_tokens": 15061428.0,
      "step": 18890
    },
    {
      "entropy": 4.754345846176148,
      "epoch": 5.342097823013853,
      "grad_norm": 1.703125,
      "learning_rate": 0.00026032193094739956,
      "loss": 4.4559,
      "mean_token_accuracy": 0.25426744371652604,
      "num_tokens": 15065664.0,
      "step": 18895
    },
    {
      "entropy": 4.674582481384277,
      "epoch": 5.34351145038168,
      "grad_norm": 1.71875,
      "learning_rate": 0.00026021929094686774,
      "loss": 4.2568,
      "mean_token_accuracy": 0.27333059161901474,
      "num_tokens": 15070239.0,
      "step": 18900
    },
    {
      "entropy": 4.685844945907593,
      "epoch": 5.344925077749505,
      "grad_norm": 1.5625,
      "learning_rate": 0.00026011665403541276,
      "loss": 4.3048,
      "mean_token_accuracy": 0.26664285361766815,
      "num_tokens": 15074454.0,
      "step": 18905
    },
    {
      "entropy": 4.629459095001221,
      "epoch": 5.346338705117331,
      "grad_norm": 1.65625,
      "learning_rate": 0.000260014020234485,
      "loss": 4.2541,
      "mean_token_accuracy": 0.27486573457717894,
      "num_tokens": 15078254.0,
      "step": 18910
    },
    {
      "entropy": 4.596324634552002,
      "epoch": 5.347752332485157,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00025991138956553433,
      "loss": 4.2016,
      "mean_token_accuracy": 0.28660893738269805,
      "num_tokens": 15082191.0,
      "step": 18915
    },
    {
      "entropy": 4.6127830982208256,
      "epoch": 5.349165959852983,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025980876205001,
      "loss": 4.1915,
      "mean_token_accuracy": 0.28312242925167086,
      "num_tokens": 15086468.0,
      "step": 18920
    },
    {
      "entropy": 4.658314323425293,
      "epoch": 5.350579587220809,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002597061377093604,
      "loss": 4.2584,
      "mean_token_accuracy": 0.27346752434968946,
      "num_tokens": 15090495.0,
      "step": 18925
    },
    {
      "entropy": 4.596587419509888,
      "epoch": 5.351993214588634,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002596035165650335,
      "loss": 4.2671,
      "mean_token_accuracy": 0.27864247411489484,
      "num_tokens": 15094547.0,
      "step": 18930
    },
    {
      "entropy": 4.643705749511719,
      "epoch": 5.353406841956461,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002595008986384763,
      "loss": 4.2936,
      "mean_token_accuracy": 0.271594925224781,
      "num_tokens": 15098574.0,
      "step": 18935
    },
    {
      "entropy": 4.599588394165039,
      "epoch": 5.354820469324286,
      "grad_norm": 1.6875,
      "learning_rate": 0.00025939828395113554,
      "loss": 4.2146,
      "mean_token_accuracy": 0.2775819316506386,
      "num_tokens": 15102568.0,
      "step": 18940
    },
    {
      "entropy": 4.602304697036743,
      "epoch": 5.356234096692112,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025929567252445696,
      "loss": 4.2208,
      "mean_token_accuracy": 0.2833869680762291,
      "num_tokens": 15106280.0,
      "step": 18945
    },
    {
      "entropy": 4.477961730957031,
      "epoch": 5.357647724059937,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00025919306437988576,
      "loss": 4.2097,
      "mean_token_accuracy": 0.29753144085407257,
      "num_tokens": 15110268.0,
      "step": 18950
    },
    {
      "entropy": 4.603532791137695,
      "epoch": 5.359061351427764,
      "grad_norm": 1.8125,
      "learning_rate": 0.00025909045953886623,
      "loss": 4.2276,
      "mean_token_accuracy": 0.27949305921792983,
      "num_tokens": 15113976.0,
      "step": 18955
    },
    {
      "entropy": 4.586815547943115,
      "epoch": 5.360474978795589,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0002589878580228423,
      "loss": 4.2212,
      "mean_token_accuracy": 0.27464399486780167,
      "num_tokens": 15117950.0,
      "step": 18960
    },
    {
      "entropy": 4.620389223098755,
      "epoch": 5.361888606163415,
      "grad_norm": 1.609375,
      "learning_rate": 0.0002588852598532571,
      "loss": 4.2894,
      "mean_token_accuracy": 0.2768131285905838,
      "num_tokens": 15121937.0,
      "step": 18965
    },
    {
      "entropy": 4.586650276184082,
      "epoch": 5.3633022335312415,
      "grad_norm": 1.6875,
      "learning_rate": 0.00025878266505155293,
      "loss": 4.2379,
      "mean_token_accuracy": 0.26970518976449964,
      "num_tokens": 15125890.0,
      "step": 18970
    },
    {
      "entropy": 4.710099077224731,
      "epoch": 5.364715860899067,
      "grad_norm": 2.0,
      "learning_rate": 0.0002586800736391714,
      "loss": 4.3752,
      "mean_token_accuracy": 0.25835520774126053,
      "num_tokens": 15129950.0,
      "step": 18975
    },
    {
      "entropy": 4.69536452293396,
      "epoch": 5.366129488266893,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002585774856375537,
      "loss": 4.2302,
      "mean_token_accuracy": 0.27157638370990755,
      "num_tokens": 15133962.0,
      "step": 18980
    },
    {
      "entropy": 4.551952743530274,
      "epoch": 5.367543115634719,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002584749010681398,
      "loss": 4.1492,
      "mean_token_accuracy": 0.2863867521286011,
      "num_tokens": 15138054.0,
      "step": 18985
    },
    {
      "entropy": 4.628730440139771,
      "epoch": 5.368956743002545,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002583723199523695,
      "loss": 4.2376,
      "mean_token_accuracy": 0.28358758687973024,
      "num_tokens": 15141947.0,
      "step": 18990
    },
    {
      "entropy": 4.575771713256836,
      "epoch": 5.37037037037037,
      "grad_norm": 1.765625,
      "learning_rate": 0.00025826974231168135,
      "loss": 4.2256,
      "mean_token_accuracy": 0.26644419580698014,
      "num_tokens": 15145872.0,
      "step": 18995
    },
    {
      "entropy": 4.645976161956787,
      "epoch": 5.371783997738196,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002581671681675137,
      "loss": 4.1807,
      "mean_token_accuracy": 0.2919409513473511,
      "num_tokens": 15149909.0,
      "step": 19000
    },
    {
      "epoch": 5.371783997738196,
      "eval_entropy": 4.478628210612513,
      "eval_loss": 4.9594197273254395,
      "eval_mean_token_accuracy": 0.25150526264282425,
      "eval_num_tokens": 15149909.0,
      "eval_runtime": 6.6203,
      "eval_samples_per_second": 1704.464,
      "eval_steps_per_second": 213.134,
      "step": 19000
    },
    {
      "entropy": 4.644647121429443,
      "epoch": 5.373197625106022,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025806459754130383,
      "loss": 4.2567,
      "mean_token_accuracy": 0.27349751740694045,
      "num_tokens": 15154024.0,
      "step": 19005
    },
    {
      "entropy": 4.6143567085266115,
      "epoch": 5.374611252473848,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002579620304544884,
      "loss": 4.2439,
      "mean_token_accuracy": 0.289864382147789,
      "num_tokens": 15157902.0,
      "step": 19010
    },
    {
      "entropy": 4.598811435699463,
      "epoch": 5.376024879841673,
      "grad_norm": 1.6875,
      "learning_rate": 0.00025785946692850317,
      "loss": 4.1081,
      "mean_token_accuracy": 0.2938228785991669,
      "num_tokens": 15161741.0,
      "step": 19015
    },
    {
      "entropy": 4.638845014572143,
      "epoch": 5.3774385072095,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002577569069847834,
      "loss": 4.1599,
      "mean_token_accuracy": 0.27967132031917574,
      "num_tokens": 15165936.0,
      "step": 19020
    },
    {
      "entropy": 4.562109327316284,
      "epoch": 5.3788521345773255,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002576543506447634,
      "loss": 4.1463,
      "mean_token_accuracy": 0.28714260160923005,
      "num_tokens": 15169593.0,
      "step": 19025
    },
    {
      "entropy": 4.709366750717163,
      "epoch": 5.380265761945151,
      "grad_norm": 1.796875,
      "learning_rate": 0.00025755179792987686,
      "loss": 4.3469,
      "mean_token_accuracy": 0.2589185982942581,
      "num_tokens": 15173553.0,
      "step": 19030
    },
    {
      "entropy": 4.680873441696167,
      "epoch": 5.3816793893129775,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0002574492488615566,
      "loss": 4.2823,
      "mean_token_accuracy": 0.26915482729673385,
      "num_tokens": 15177387.0,
      "step": 19035
    },
    {
      "entropy": 4.6759429454803465,
      "epoch": 5.383093016680803,
      "grad_norm": 1.65625,
      "learning_rate": 0.00025734670346123494,
      "loss": 4.2347,
      "mean_token_accuracy": 0.2724695816636086,
      "num_tokens": 15181465.0,
      "step": 19040
    },
    {
      "entropy": 4.712905788421631,
      "epoch": 5.384506644048629,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025724416175034305,
      "loss": 4.2744,
      "mean_token_accuracy": 0.266559274494648,
      "num_tokens": 15185161.0,
      "step": 19045
    },
    {
      "entropy": 4.532208824157715,
      "epoch": 5.385920271416454,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00025714162375031145,
      "loss": 4.1589,
      "mean_token_accuracy": 0.2781372606754303,
      "num_tokens": 15189117.0,
      "step": 19050
    },
    {
      "entropy": 4.667049694061279,
      "epoch": 5.387333898784281,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0002570390894825701,
      "loss": 4.3367,
      "mean_token_accuracy": 0.2629843771457672,
      "num_tokens": 15193161.0,
      "step": 19055
    },
    {
      "entropy": 4.479118156433105,
      "epoch": 5.388747526152106,
      "grad_norm": 1.6875,
      "learning_rate": 0.000256936558968548,
      "loss": 4.1305,
      "mean_token_accuracy": 0.29088871777057645,
      "num_tokens": 15197201.0,
      "step": 19060
    },
    {
      "entropy": 4.6705046653747555,
      "epoch": 5.390161153519932,
      "grad_norm": 1.796875,
      "learning_rate": 0.0002568340322296734,
      "loss": 4.2796,
      "mean_token_accuracy": 0.25762269347906114,
      "num_tokens": 15200895.0,
      "step": 19065
    },
    {
      "entropy": 4.591110420227051,
      "epoch": 5.391574780887758,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002567315092873737,
      "loss": 4.1412,
      "mean_token_accuracy": 0.28759763538837435,
      "num_tokens": 15204779.0,
      "step": 19070
    },
    {
      "entropy": 4.630869150161743,
      "epoch": 5.392988408255584,
      "grad_norm": 1.65625,
      "learning_rate": 0.00025662899016307557,
      "loss": 4.2277,
      "mean_token_accuracy": 0.27812871187925337,
      "num_tokens": 15208592.0,
      "step": 19075
    },
    {
      "entropy": 4.508808994293213,
      "epoch": 5.3944020356234095,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025652647487820493,
      "loss": 4.242,
      "mean_token_accuracy": 0.27717190980911255,
      "num_tokens": 15212389.0,
      "step": 19080
    },
    {
      "entropy": 4.555541086196899,
      "epoch": 5.395815662991236,
      "grad_norm": 1.9296875,
      "learning_rate": 0.0002564239634541868,
      "loss": 4.2021,
      "mean_token_accuracy": 0.2755569502711296,
      "num_tokens": 15216081.0,
      "step": 19085
    },
    {
      "entropy": 4.62540111541748,
      "epoch": 5.3972292903590615,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00025632145591244544,
      "loss": 4.2574,
      "mean_token_accuracy": 0.28093823194503786,
      "num_tokens": 15219873.0,
      "step": 19090
    },
    {
      "entropy": 4.687761449813843,
      "epoch": 5.398642917726887,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025621895227440425,
      "loss": 4.2941,
      "mean_token_accuracy": 0.2729255899786949,
      "num_tokens": 15223759.0,
      "step": 19095
    },
    {
      "entropy": 4.600138759613037,
      "epoch": 5.400056545094713,
      "grad_norm": 1.5625,
      "learning_rate": 0.0002561164525614859,
      "loss": 4.1326,
      "mean_token_accuracy": 0.2902753382921219,
      "num_tokens": 15227775.0,
      "step": 19100
    },
    {
      "entropy": 4.649198818206787,
      "epoch": 5.401470172462539,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002560139567951123,
      "loss": 4.3045,
      "mean_token_accuracy": 0.2675986588001251,
      "num_tokens": 15231520.0,
      "step": 19105
    },
    {
      "entropy": 4.576262903213501,
      "epoch": 5.402883799830365,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002559114649967043,
      "loss": 4.1577,
      "mean_token_accuracy": 0.2829198196530342,
      "num_tokens": 15235782.0,
      "step": 19110
    },
    {
      "entropy": 4.531535720825195,
      "epoch": 5.40429742719819,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00025580897718768196,
      "loss": 4.2453,
      "mean_token_accuracy": 0.28671201020479203,
      "num_tokens": 15239862.0,
      "step": 19115
    },
    {
      "entropy": 4.728608179092407,
      "epoch": 5.405711054566017,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002557064933894647,
      "loss": 4.3263,
      "mean_token_accuracy": 0.26747415363788607,
      "num_tokens": 15244223.0,
      "step": 19120
    },
    {
      "entropy": 4.717931699752808,
      "epoch": 5.407124681933842,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002556040136234711,
      "loss": 4.318,
      "mean_token_accuracy": 0.26174742430448533,
      "num_tokens": 15248123.0,
      "step": 19125
    },
    {
      "entropy": 4.680100822448731,
      "epoch": 5.408538309301668,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00025550153791111874,
      "loss": 4.3227,
      "mean_token_accuracy": 0.2601091593503952,
      "num_tokens": 15252007.0,
      "step": 19130
    },
    {
      "entropy": 4.649523782730102,
      "epoch": 5.4099519366694935,
      "grad_norm": 1.640625,
      "learning_rate": 0.00025539906627382433,
      "loss": 4.2739,
      "mean_token_accuracy": 0.269521526992321,
      "num_tokens": 15256205.0,
      "step": 19135
    },
    {
      "entropy": 4.6080155849456785,
      "epoch": 5.41136556403732,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00025529659873300384,
      "loss": 4.1937,
      "mean_token_accuracy": 0.2782732829451561,
      "num_tokens": 15259782.0,
      "step": 19140
    },
    {
      "entropy": 4.494810438156128,
      "epoch": 5.4127791914051455,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00025519413531007243,
      "loss": 4.1664,
      "mean_token_accuracy": 0.29185024946928023,
      "num_tokens": 15263805.0,
      "step": 19145
    },
    {
      "entropy": 4.579969263076782,
      "epoch": 5.414192818772971,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025509167602644423,
      "loss": 4.2223,
      "mean_token_accuracy": 0.2837143912911415,
      "num_tokens": 15267525.0,
      "step": 19150
    },
    {
      "entropy": 4.563865518569946,
      "epoch": 5.415606446140798,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00025498922090353267,
      "loss": 4.1929,
      "mean_token_accuracy": 0.2756590932607651,
      "num_tokens": 15271487.0,
      "step": 19155
    },
    {
      "entropy": 4.613961887359619,
      "epoch": 5.417020073508623,
      "grad_norm": 1.8359375,
      "learning_rate": 0.00025488676996275015,
      "loss": 4.2018,
      "mean_token_accuracy": 0.2795114114880562,
      "num_tokens": 15275298.0,
      "step": 19160
    },
    {
      "entropy": 4.643416452407837,
      "epoch": 5.418433700876449,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0002547843232255083,
      "loss": 4.2723,
      "mean_token_accuracy": 0.27032252699136733,
      "num_tokens": 15279349.0,
      "step": 19165
    },
    {
      "entropy": 4.646218729019165,
      "epoch": 5.419847328244275,
      "grad_norm": 1.8046875,
      "learning_rate": 0.00025468188071321794,
      "loss": 4.2701,
      "mean_token_accuracy": 0.2798605144023895,
      "num_tokens": 15283197.0,
      "step": 19170
    },
    {
      "entropy": 4.663592290878296,
      "epoch": 5.421260955612101,
      "grad_norm": 1.578125,
      "learning_rate": 0.00025457944244728893,
      "loss": 4.3613,
      "mean_token_accuracy": 0.26325341761112214,
      "num_tokens": 15287394.0,
      "step": 19175
    },
    {
      "entropy": 4.6631388664245605,
      "epoch": 5.422674582979926,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00025447700844913004,
      "loss": 4.2978,
      "mean_token_accuracy": 0.2654219761490822,
      "num_tokens": 15291232.0,
      "step": 19180
    },
    {
      "entropy": 4.659866046905518,
      "epoch": 5.424088210347753,
      "grad_norm": 1.59375,
      "learning_rate": 0.00025437457874014953,
      "loss": 4.1538,
      "mean_token_accuracy": 0.28383120745420454,
      "num_tokens": 15295390.0,
      "step": 19185
    },
    {
      "entropy": 4.674236202239991,
      "epoch": 5.425501837715578,
      "grad_norm": 1.640625,
      "learning_rate": 0.00025427215334175454,
      "loss": 4.277,
      "mean_token_accuracy": 0.2716983124613762,
      "num_tokens": 15299345.0,
      "step": 19190
    },
    {
      "entropy": 4.58827428817749,
      "epoch": 5.426915465083404,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002541697322753513,
      "loss": 4.3167,
      "mean_token_accuracy": 0.26557250022888185,
      "num_tokens": 15303009.0,
      "step": 19195
    },
    {
      "entropy": 4.563638305664062,
      "epoch": 5.4283290924512295,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002540673155623452,
      "loss": 4.1504,
      "mean_token_accuracy": 0.2789662808179855,
      "num_tokens": 15307192.0,
      "step": 19200
    },
    {
      "entropy": 4.58415904045105,
      "epoch": 5.429742719819056,
      "grad_norm": 1.78125,
      "learning_rate": 0.00025396490322414074,
      "loss": 4.2403,
      "mean_token_accuracy": 0.2881885826587677,
      "num_tokens": 15311147.0,
      "step": 19205
    },
    {
      "entropy": 4.72127251625061,
      "epoch": 5.431156347186882,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002538624952821414,
      "loss": 4.2725,
      "mean_token_accuracy": 0.27173391729593277,
      "num_tokens": 15315030.0,
      "step": 19210
    },
    {
      "entropy": 4.64974193572998,
      "epoch": 5.432569974554707,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00025376009175774985,
      "loss": 4.2295,
      "mean_token_accuracy": 0.27934877574443817,
      "num_tokens": 15319220.0,
      "step": 19215
    },
    {
      "entropy": 4.530527400970459,
      "epoch": 5.433983601922534,
      "grad_norm": 1.71875,
      "learning_rate": 0.0002536576926723678,
      "loss": 4.1061,
      "mean_token_accuracy": 0.27648999392986295,
      "num_tokens": 15323080.0,
      "step": 19220
    },
    {
      "entropy": 4.631598901748657,
      "epoch": 5.435397229290359,
      "grad_norm": 1.734375,
      "learning_rate": 0.00025355529804739593,
      "loss": 4.3094,
      "mean_token_accuracy": 0.270842470228672,
      "num_tokens": 15326983.0,
      "step": 19225
    },
    {
      "entropy": 4.624290227890015,
      "epoch": 5.436810856658185,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00025345290790423425,
      "loss": 4.2325,
      "mean_token_accuracy": 0.2717694744467735,
      "num_tokens": 15330900.0,
      "step": 19230
    },
    {
      "entropy": 4.664579439163208,
      "epoch": 5.43822448402601,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002533505222642815,
      "loss": 4.1979,
      "mean_token_accuracy": 0.28367589712142943,
      "num_tokens": 15334457.0,
      "step": 19235
    },
    {
      "entropy": 4.669460153579712,
      "epoch": 5.439638111393837,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00025324814114893566,
      "loss": 4.2561,
      "mean_token_accuracy": 0.27877930253744126,
      "num_tokens": 15338250.0,
      "step": 19240
    },
    {
      "entropy": 4.600064659118653,
      "epoch": 5.441051738761662,
      "grad_norm": 1.640625,
      "learning_rate": 0.00025314576457959395,
      "loss": 4.2648,
      "mean_token_accuracy": 0.2847629949450493,
      "num_tokens": 15342164.0,
      "step": 19245
    },
    {
      "entropy": 4.61215615272522,
      "epoch": 5.442465366129488,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00025304339257765224,
      "loss": 4.2586,
      "mean_token_accuracy": 0.27588938027620313,
      "num_tokens": 15346291.0,
      "step": 19250
    },
    {
      "entropy": 4.636011123657227,
      "epoch": 5.443878993497314,
      "grad_norm": 1.796875,
      "learning_rate": 0.00025294102516450575,
      "loss": 4.1851,
      "mean_token_accuracy": 0.27901550084352494,
      "num_tokens": 15350262.0,
      "step": 19255
    },
    {
      "entropy": 4.695978355407715,
      "epoch": 5.44529262086514,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002528386623615485,
      "loss": 4.3118,
      "mean_token_accuracy": 0.26561625152826307,
      "num_tokens": 15354053.0,
      "step": 19260
    },
    {
      "entropy": 4.6332361698150635,
      "epoch": 5.446706248232966,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00025273630419017385,
      "loss": 4.164,
      "mean_token_accuracy": 0.28656245470047,
      "num_tokens": 15357985.0,
      "step": 19265
    },
    {
      "entropy": 4.585815334320069,
      "epoch": 5.448119875600792,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00025263395067177393,
      "loss": 4.1487,
      "mean_token_accuracy": 0.2833580255508423,
      "num_tokens": 15361717.0,
      "step": 19270
    },
    {
      "entropy": 4.669905662536621,
      "epoch": 5.449533502968618,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00025253160182774005,
      "loss": 4.294,
      "mean_token_accuracy": 0.26854324638843535,
      "num_tokens": 15365802.0,
      "step": 19275
    },
    {
      "entropy": 4.541899728775024,
      "epoch": 5.450947130336443,
      "grad_norm": 1.6484375,
      "learning_rate": 0.0002524292576794623,
      "loss": 4.169,
      "mean_token_accuracy": 0.2756502881646156,
      "num_tokens": 15370154.0,
      "step": 19280
    },
    {
      "entropy": 4.694264984130859,
      "epoch": 5.452360757704269,
      "grad_norm": 1.8828125,
      "learning_rate": 0.00025232691824833033,
      "loss": 4.2244,
      "mean_token_accuracy": 0.2624236062169075,
      "num_tokens": 15374128.0,
      "step": 19285
    },
    {
      "entropy": 4.65144190788269,
      "epoch": 5.453774385072095,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002522245835557321,
      "loss": 4.2571,
      "mean_token_accuracy": 0.2665132641792297,
      "num_tokens": 15378120.0,
      "step": 19290
    },
    {
      "entropy": 4.736142158508301,
      "epoch": 5.455188012439921,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002521222536230552,
      "loss": 4.3253,
      "mean_token_accuracy": 0.25937038064002993,
      "num_tokens": 15382155.0,
      "step": 19295
    },
    {
      "entropy": 4.665150499343872,
      "epoch": 5.456601639807746,
      "grad_norm": 1.875,
      "learning_rate": 0.00025201992847168563,
      "loss": 4.2073,
      "mean_token_accuracy": 0.2789221122860909,
      "num_tokens": 15386114.0,
      "step": 19300
    },
    {
      "entropy": 4.6453368186950685,
      "epoch": 5.458015267175573,
      "grad_norm": 1.59375,
      "learning_rate": 0.00025191760812300903,
      "loss": 4.2663,
      "mean_token_accuracy": 0.27358233630657197,
      "num_tokens": 15389942.0,
      "step": 19305
    },
    {
      "entropy": 4.647691535949707,
      "epoch": 5.459428894543398,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002518152925984096,
      "loss": 4.2556,
      "mean_token_accuracy": 0.2787679523229599,
      "num_tokens": 15393997.0,
      "step": 19310
    },
    {
      "entropy": 4.631117820739746,
      "epoch": 5.460842521911224,
      "grad_norm": 1.734375,
      "learning_rate": 0.00025171298191927063,
      "loss": 4.2224,
      "mean_token_accuracy": 0.27944723069667815,
      "num_tokens": 15398047.0,
      "step": 19315
    },
    {
      "entropy": 4.694785833358765,
      "epoch": 5.4622561492790505,
      "grad_norm": 1.7578125,
      "learning_rate": 0.0002516106761069744,
      "loss": 4.3096,
      "mean_token_accuracy": 0.2690328165888786,
      "num_tokens": 15401989.0,
      "step": 19320
    },
    {
      "entropy": 4.600774192810059,
      "epoch": 5.463669776646876,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002515083751829023,
      "loss": 4.2165,
      "mean_token_accuracy": 0.29311950206756593,
      "num_tokens": 15405954.0,
      "step": 19325
    },
    {
      "entropy": 4.639438009262085,
      "epoch": 5.465083404014702,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002514060791684344,
      "loss": 4.2347,
      "mean_token_accuracy": 0.2766521587967873,
      "num_tokens": 15409976.0,
      "step": 19330
    },
    {
      "entropy": 4.580479574203491,
      "epoch": 5.466497031382527,
      "grad_norm": 1.8671875,
      "learning_rate": 0.0002513037880849502,
      "loss": 4.152,
      "mean_token_accuracy": 0.28289880603551865,
      "num_tokens": 15414125.0,
      "step": 19335
    },
    {
      "entropy": 4.6118323802948,
      "epoch": 5.467910658750354,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002512015019538276,
      "loss": 4.2815,
      "mean_token_accuracy": 0.2731777563691139,
      "num_tokens": 15417903.0,
      "step": 19340
    },
    {
      "entropy": 4.625623559951782,
      "epoch": 5.469324286118179,
      "grad_norm": 1.71875,
      "learning_rate": 0.000251099220796444,
      "loss": 4.2637,
      "mean_token_accuracy": 0.27459862977266314,
      "num_tokens": 15421939.0,
      "step": 19345
    },
    {
      "entropy": 4.682634925842285,
      "epoch": 5.470737913486005,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002509969446341755,
      "loss": 4.1702,
      "mean_token_accuracy": 0.2767213940620422,
      "num_tokens": 15425975.0,
      "step": 19350
    },
    {
      "entropy": 4.71600227355957,
      "epoch": 5.472151540853831,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002508946734883971,
      "loss": 4.2066,
      "mean_token_accuracy": 0.28790567219257357,
      "num_tokens": 15430138.0,
      "step": 19355
    },
    {
      "entropy": 4.644129514694214,
      "epoch": 5.473565168221657,
      "grad_norm": 1.8515625,
      "learning_rate": 0.00025079240738048287,
      "loss": 4.2281,
      "mean_token_accuracy": 0.28062531501054766,
      "num_tokens": 15433825.0,
      "step": 19360
    },
    {
      "entropy": 4.628208160400391,
      "epoch": 5.474978795589482,
      "grad_norm": 1.7734375,
      "learning_rate": 0.0002506901463318058,
      "loss": 4.3774,
      "mean_token_accuracy": 0.25838697999715804,
      "num_tokens": 15437765.0,
      "step": 19365
    },
    {
      "entropy": 4.68724570274353,
      "epoch": 5.476392422957309,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00025058789036373785,
      "loss": 4.3052,
      "mean_token_accuracy": 0.2661417931318283,
      "num_tokens": 15441394.0,
      "step": 19370
    },
    {
      "entropy": 4.610318660736084,
      "epoch": 5.4778060503251345,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002504856394976499,
      "loss": 4.1549,
      "mean_token_accuracy": 0.2893331930041313,
      "num_tokens": 15445194.0,
      "step": 19375
    },
    {
      "entropy": 4.755679941177368,
      "epoch": 5.47921967769296,
      "grad_norm": 1.734375,
      "learning_rate": 0.00025038339375491156,
      "loss": 4.3511,
      "mean_token_accuracy": 0.25704107284545896,
      "num_tokens": 15449196.0,
      "step": 19380
    },
    {
      "entropy": 4.6915136814117435,
      "epoch": 5.480633305060786,
      "grad_norm": 1.640625,
      "learning_rate": 0.00025028115315689183,
      "loss": 4.2972,
      "mean_token_accuracy": 0.2631926089525223,
      "num_tokens": 15453071.0,
      "step": 19385
    },
    {
      "entropy": 4.645698738098145,
      "epoch": 5.482046932428612,
      "grad_norm": 1.671875,
      "learning_rate": 0.00025017891772495814,
      "loss": 4.2619,
      "mean_token_accuracy": 0.27408865839242935,
      "num_tokens": 15457045.0,
      "step": 19390
    },
    {
      "entropy": 4.688718175888061,
      "epoch": 5.483460559796438,
      "grad_norm": 1.8125,
      "learning_rate": 0.00025007668748047723,
      "loss": 4.3192,
      "mean_token_accuracy": 0.26422131061553955,
      "num_tokens": 15461049.0,
      "step": 19395
    },
    {
      "entropy": 4.623563528060913,
      "epoch": 5.484874187164263,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002499744624448145,
      "loss": 4.2861,
      "mean_token_accuracy": 0.26320843398571014,
      "num_tokens": 15465164.0,
      "step": 19400
    },
    {
      "entropy": 4.675965118408203,
      "epoch": 5.48628781453209,
      "grad_norm": 1.9609375,
      "learning_rate": 0.00024987224263933435,
      "loss": 4.243,
      "mean_token_accuracy": 0.2751036286354065,
      "num_tokens": 15469373.0,
      "step": 19405
    },
    {
      "entropy": 4.668718528747559,
      "epoch": 5.487701441899915,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002497700280854001,
      "loss": 4.2269,
      "mean_token_accuracy": 0.2715396896004677,
      "num_tokens": 15473313.0,
      "step": 19410
    },
    {
      "entropy": 4.709665250778198,
      "epoch": 5.489115069267741,
      "grad_norm": 1.625,
      "learning_rate": 0.000249667818804374,
      "loss": 4.2174,
      "mean_token_accuracy": 0.2925337582826614,
      "num_tokens": 15477213.0,
      "step": 19415
    },
    {
      "entropy": 4.705202388763428,
      "epoch": 5.490528696635566,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00024956561481761716,
      "loss": 4.2246,
      "mean_token_accuracy": 0.27598472833633425,
      "num_tokens": 15481262.0,
      "step": 19420
    },
    {
      "entropy": 4.559599256515503,
      "epoch": 5.491942324003393,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002494634161464894,
      "loss": 4.2032,
      "mean_token_accuracy": 0.28377084285020826,
      "num_tokens": 15484856.0,
      "step": 19425
    },
    {
      "entropy": 4.721272563934326,
      "epoch": 5.4933559513712185,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002493612228123498,
      "loss": 4.3736,
      "mean_token_accuracy": 0.26757115721702573,
      "num_tokens": 15489161.0,
      "step": 19430
    },
    {
      "entropy": 4.674237012863159,
      "epoch": 5.494769578739044,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002492590348365561,
      "loss": 4.2662,
      "mean_token_accuracy": 0.26578062027692795,
      "num_tokens": 15493070.0,
      "step": 19435
    },
    {
      "entropy": 4.65609712600708,
      "epoch": 5.4961832061068705,
      "grad_norm": 1.75,
      "learning_rate": 0.000249156852240465,
      "loss": 4.2925,
      "mean_token_accuracy": 0.27246343344449997,
      "num_tokens": 15497083.0,
      "step": 19440
    },
    {
      "entropy": 4.606566286087036,
      "epoch": 5.497596833474696,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002490546750454319,
      "loss": 4.1886,
      "mean_token_accuracy": 0.2820375844836235,
      "num_tokens": 15501162.0,
      "step": 19445
    },
    {
      "entropy": 4.61837248802185,
      "epoch": 5.499010460842522,
      "grad_norm": 1.703125,
      "learning_rate": 0.00024895250327281127,
      "loss": 4.1299,
      "mean_token_accuracy": 0.28364060670137403,
      "num_tokens": 15505241.0,
      "step": 19450
    },
    {
      "entropy": 4.649972677230835,
      "epoch": 5.500424088210348,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002488503369439564,
      "loss": 4.2462,
      "mean_token_accuracy": 0.28003452718257904,
      "num_tokens": 15508843.0,
      "step": 19455
    },
    {
      "entropy": 4.687499475479126,
      "epoch": 5.501837715578174,
      "grad_norm": 1.875,
      "learning_rate": 0.0002487481760802195,
      "loss": 4.2744,
      "mean_token_accuracy": 0.27426232397556305,
      "num_tokens": 15512656.0,
      "step": 19460
    },
    {
      "entropy": 4.623730134963989,
      "epoch": 5.503251342945999,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00024864602070295133,
      "loss": 4.1834,
      "mean_token_accuracy": 0.27945693135261535,
      "num_tokens": 15516444.0,
      "step": 19465
    },
    {
      "entropy": 4.6304404735565186,
      "epoch": 5.504664970313826,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002485438708335019,
      "loss": 4.1722,
      "mean_token_accuracy": 0.2812226340174675,
      "num_tokens": 15520445.0,
      "step": 19470
    },
    {
      "entropy": 4.546290349960327,
      "epoch": 5.506078597681651,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002484417264932199,
      "loss": 4.1773,
      "mean_token_accuracy": 0.2748391792178154,
      "num_tokens": 15524369.0,
      "step": 19475
    },
    {
      "entropy": 4.615872812271118,
      "epoch": 5.507492225049477,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00024833958770345283,
      "loss": 4.1767,
      "mean_token_accuracy": 0.2837546706199646,
      "num_tokens": 15528589.0,
      "step": 19480
    },
    {
      "entropy": 4.700962352752685,
      "epoch": 5.5089058524173025,
      "grad_norm": 1.703125,
      "learning_rate": 0.00024823745448554696,
      "loss": 4.3597,
      "mean_token_accuracy": 0.263947120308876,
      "num_tokens": 15532807.0,
      "step": 19485
    },
    {
      "entropy": 4.64639048576355,
      "epoch": 5.510319479785129,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002481353268608477,
      "loss": 4.2156,
      "mean_token_accuracy": 0.2831979274749756,
      "num_tokens": 15536876.0,
      "step": 19490
    },
    {
      "entropy": 4.6093708038330075,
      "epoch": 5.5117331071529545,
      "grad_norm": 1.5859375,
      "learning_rate": 0.000248033204850699,
      "loss": 4.1969,
      "mean_token_accuracy": 0.28140056133270264,
      "num_tokens": 15540897.0,
      "step": 19495
    },
    {
      "entropy": 4.695872116088867,
      "epoch": 5.51314673452078,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002479310884764438,
      "loss": 4.3624,
      "mean_token_accuracy": 0.2683891147375107,
      "num_tokens": 15544751.0,
      "step": 19500
    },
    {
      "entropy": 4.687041807174682,
      "epoch": 5.5145603618886065,
      "grad_norm": 1.765625,
      "learning_rate": 0.00024782897775942357,
      "loss": 4.2164,
      "mean_token_accuracy": 0.2705345243215561,
      "num_tokens": 15548735.0,
      "step": 19505
    },
    {
      "entropy": 4.670182561874389,
      "epoch": 5.515973989256432,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002477268727209791,
      "loss": 4.1988,
      "mean_token_accuracy": 0.2807599321007729,
      "num_tokens": 15552677.0,
      "step": 19510
    },
    {
      "entropy": 4.754320812225342,
      "epoch": 5.517387616624258,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002476247733824495,
      "loss": 4.3175,
      "mean_token_accuracy": 0.2691409230232239,
      "num_tokens": 15556805.0,
      "step": 19515
    },
    {
      "entropy": 4.661415910720825,
      "epoch": 5.518801243992083,
      "grad_norm": 1.734375,
      "learning_rate": 0.000247522679765173,
      "loss": 4.2294,
      "mean_token_accuracy": 0.286704882979393,
      "num_tokens": 15560589.0,
      "step": 19520
    },
    {
      "entropy": 4.632322788238525,
      "epoch": 5.52021487135991,
      "grad_norm": 1.7890625,
      "learning_rate": 0.0002474205918904865,
      "loss": 4.1967,
      "mean_token_accuracy": 0.2817597627639771,
      "num_tokens": 15564683.0,
      "step": 19525
    },
    {
      "entropy": 4.6200432777404785,
      "epoch": 5.521628498727735,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002473185097797258,
      "loss": 4.1781,
      "mean_token_accuracy": 0.2892683491110802,
      "num_tokens": 15568513.0,
      "step": 19530
    },
    {
      "entropy": 4.61649341583252,
      "epoch": 5.523042126095561,
      "grad_norm": 1.703125,
      "learning_rate": 0.00024721643345422534,
      "loss": 4.2512,
      "mean_token_accuracy": 0.27922495603561404,
      "num_tokens": 15572614.0,
      "step": 19535
    },
    {
      "entropy": 4.556638383865357,
      "epoch": 5.524455753463387,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002471143629353185,
      "loss": 4.2859,
      "mean_token_accuracy": 0.2749201327562332,
      "num_tokens": 15576596.0,
      "step": 19540
    },
    {
      "entropy": 4.6875988960266115,
      "epoch": 5.525869380831213,
      "grad_norm": 1.8125,
      "learning_rate": 0.0002470122982443373,
      "loss": 4.3118,
      "mean_token_accuracy": 0.2689857453107834,
      "num_tokens": 15580524.0,
      "step": 19545
    },
    {
      "entropy": 4.661058950424194,
      "epoch": 5.5272830081990385,
      "grad_norm": 1.7109375,
      "learning_rate": 0.0002469102394026127,
      "loss": 4.2139,
      "mean_token_accuracy": 0.27813027799129486,
      "num_tokens": 15584561.0,
      "step": 19550
    },
    {
      "entropy": 4.697919034957886,
      "epoch": 5.528696635566865,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002468081864314745,
      "loss": 4.3038,
      "mean_token_accuracy": 0.2699168220162392,
      "num_tokens": 15588603.0,
      "step": 19555
    },
    {
      "entropy": 4.64435830116272,
      "epoch": 5.5301102629346905,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002467061393522509,
      "loss": 4.2204,
      "mean_token_accuracy": 0.27239142507314684,
      "num_tokens": 15592889.0,
      "step": 19560
    },
    {
      "entropy": 4.646487474441528,
      "epoch": 5.531523890302516,
      "grad_norm": 1.734375,
      "learning_rate": 0.00024660409818626916,
      "loss": 4.2509,
      "mean_token_accuracy": 0.271334508061409,
      "num_tokens": 15596602.0,
      "step": 19565
    },
    {
      "entropy": 4.593534517288208,
      "epoch": 5.532937517670343,
      "grad_norm": 1.7734375,
      "learning_rate": 0.00024650206295485536,
      "loss": 4.2932,
      "mean_token_accuracy": 0.28398210406303404,
      "num_tokens": 15600401.0,
      "step": 19570
    },
    {
      "entropy": 4.608480596542359,
      "epoch": 5.534351145038168,
      "grad_norm": 1.75,
      "learning_rate": 0.00024640003367933425,
      "loss": 4.1755,
      "mean_token_accuracy": 0.28554896265268326,
      "num_tokens": 15604159.0,
      "step": 19575
    },
    {
      "entropy": 4.6109209060668945,
      "epoch": 5.535764772405994,
      "grad_norm": 1.71875,
      "learning_rate": 0.00024629801038102914,
      "loss": 4.1565,
      "mean_token_accuracy": 0.2777742251753807,
      "num_tokens": 15608088.0,
      "step": 19580
    },
    {
      "entropy": 4.606988334655762,
      "epoch": 5.537178399773819,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00024619599308126246,
      "loss": 4.2407,
      "mean_token_accuracy": 0.27509729713201525,
      "num_tokens": 15612243.0,
      "step": 19585
    },
    {
      "entropy": 4.600841760635376,
      "epoch": 5.538592027141646,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002460939818013551,
      "loss": 4.229,
      "mean_token_accuracy": 0.2785800457000732,
      "num_tokens": 15616392.0,
      "step": 19590
    },
    {
      "entropy": 4.700632762908936,
      "epoch": 5.540005654509471,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024599197656262675,
      "loss": 4.248,
      "mean_token_accuracy": 0.2634484425187111,
      "num_tokens": 15620680.0,
      "step": 19595
    },
    {
      "entropy": 4.672055006027222,
      "epoch": 5.541419281877297,
      "grad_norm": 1.78125,
      "learning_rate": 0.0002458899773863959,
      "loss": 4.2413,
      "mean_token_accuracy": 0.27624692022800446,
      "num_tokens": 15624435.0,
      "step": 19600
    },
    {
      "entropy": 4.7090685844421385,
      "epoch": 5.5428329092451225,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002457879842939797,
      "loss": 4.3797,
      "mean_token_accuracy": 0.266084623336792,
      "num_tokens": 15628569.0,
      "step": 19605
    },
    {
      "entropy": 4.6327437400817875,
      "epoch": 5.544246536612949,
      "grad_norm": 1.6171875,
      "learning_rate": 0.0002456859973066942,
      "loss": 4.2562,
      "mean_token_accuracy": 0.2763918966054916,
      "num_tokens": 15632452.0,
      "step": 19610
    },
    {
      "entropy": 4.659896993637085,
      "epoch": 5.5456601639807745,
      "grad_norm": 1.578125,
      "learning_rate": 0.0002455840164458541,
      "loss": 4.2055,
      "mean_token_accuracy": 0.28411660343408585,
      "num_tokens": 15636982.0,
      "step": 19615
    },
    {
      "entropy": 4.639674472808838,
      "epoch": 5.5470737913486,
      "grad_norm": 1.75,
      "learning_rate": 0.00024548204173277256,
      "loss": 4.2525,
      "mean_token_accuracy": 0.26117789894342425,
      "num_tokens": 15641027.0,
      "step": 19620
    },
    {
      "entropy": 4.691533136367798,
      "epoch": 5.548487418716427,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00024538007318876176,
      "loss": 4.2256,
      "mean_token_accuracy": 0.2717921271920204,
      "num_tokens": 15645137.0,
      "step": 19625
    },
    {
      "entropy": 4.746392917633057,
      "epoch": 5.549901046084252,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0002452781108351325,
      "loss": 4.3999,
      "mean_token_accuracy": 0.2597831696271896,
      "num_tokens": 15649240.0,
      "step": 19630
    },
    {
      "entropy": 4.680949687957764,
      "epoch": 5.551314673452078,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024517615469319434,
      "loss": 4.2651,
      "mean_token_accuracy": 0.2748690083622932,
      "num_tokens": 15653069.0,
      "step": 19635
    },
    {
      "entropy": 4.713500165939331,
      "epoch": 5.552728300819904,
      "grad_norm": 1.640625,
      "learning_rate": 0.0002450742047842554,
      "loss": 4.3109,
      "mean_token_accuracy": 0.2801923111081123,
      "num_tokens": 15657011.0,
      "step": 19640
    },
    {
      "entropy": 4.618224000930786,
      "epoch": 5.55414192818773,
      "grad_norm": 1.8203125,
      "learning_rate": 0.0002449722611296226,
      "loss": 4.2068,
      "mean_token_accuracy": 0.2770373851060867,
      "num_tokens": 15660902.0,
      "step": 19645
    },
    {
      "entropy": 4.68924732208252,
      "epoch": 5.555555555555555,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002448703237506015,
      "loss": 4.3721,
      "mean_token_accuracy": 0.25445225536823274,
      "num_tokens": 15665026.0,
      "step": 19650
    },
    {
      "entropy": 4.607961130142212,
      "epoch": 5.556969182923382,
      "grad_norm": 1.75,
      "learning_rate": 0.0002447683926684965,
      "loss": 4.2142,
      "mean_token_accuracy": 0.27678134590387343,
      "num_tokens": 15668966.0,
      "step": 19655
    },
    {
      "entropy": 4.580642938613892,
      "epoch": 5.558382810291207,
      "grad_norm": 1.65625,
      "learning_rate": 0.0002446664679046104,
      "loss": 4.1003,
      "mean_token_accuracy": 0.28856033384799956,
      "num_tokens": 15672791.0,
      "step": 19660
    },
    {
      "entropy": 4.698865795135498,
      "epoch": 5.559796437659033,
      "grad_norm": 1.65625,
      "learning_rate": 0.00024456454948024485,
      "loss": 4.3149,
      "mean_token_accuracy": 0.26684743314981463,
      "num_tokens": 15676725.0,
      "step": 19665
    },
    {
      "entropy": 4.600046062469483,
      "epoch": 5.5612100650268586,
      "grad_norm": 1.65625,
      "learning_rate": 0.00024446263741670045,
      "loss": 4.2453,
      "mean_token_accuracy": 0.2705697566270828,
      "num_tokens": 15680923.0,
      "step": 19670
    },
    {
      "entropy": 4.652478647232056,
      "epoch": 5.562623692394685,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0002443607317352759,
      "loss": 4.213,
      "mean_token_accuracy": 0.2736234351992607,
      "num_tokens": 15684976.0,
      "step": 19675
    },
    {
      "entropy": 4.65725474357605,
      "epoch": 5.564037319762511,
      "grad_norm": 1.65625,
      "learning_rate": 0.000244258832457269,
      "loss": 4.1924,
      "mean_token_accuracy": 0.27809516340494156,
      "num_tokens": 15689085.0,
      "step": 19680
    },
    {
      "entropy": 4.586135816574097,
      "epoch": 5.565450947130336,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00024415693960397594,
      "loss": 4.2111,
      "mean_token_accuracy": 0.27843959629535675,
      "num_tokens": 15692958.0,
      "step": 19685
    },
    {
      "entropy": 4.6016734600067135,
      "epoch": 5.566864574498163,
      "grad_norm": 1.75,
      "learning_rate": 0.0002440550531966918,
      "loss": 4.1551,
      "mean_token_accuracy": 0.2832563668489456,
      "num_tokens": 15696842.0,
      "step": 19690
    },
    {
      "entropy": 4.6128603458404545,
      "epoch": 5.568278201865988,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00024395317325671023,
      "loss": 4.3162,
      "mean_token_accuracy": 0.27016914784908297,
      "num_tokens": 15700776.0,
      "step": 19695
    },
    {
      "entropy": 4.589887809753418,
      "epoch": 5.569691829233814,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024385129980532345,
      "loss": 4.1611,
      "mean_token_accuracy": 0.2986323028802872,
      "num_tokens": 15704707.0,
      "step": 19700
    },
    {
      "entropy": 4.687882757186889,
      "epoch": 5.571105456601639,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002437494328638224,
      "loss": 4.2836,
      "mean_token_accuracy": 0.2715510904788971,
      "num_tokens": 15708855.0,
      "step": 19705
    },
    {
      "entropy": 4.6500293731689455,
      "epoch": 5.572519083969466,
      "grad_norm": 1.640625,
      "learning_rate": 0.00024364757245349655,
      "loss": 4.254,
      "mean_token_accuracy": 0.27813707292079926,
      "num_tokens": 15712890.0,
      "step": 19710
    },
    {
      "entropy": 4.555387926101685,
      "epoch": 5.573932711337291,
      "grad_norm": 1.640625,
      "learning_rate": 0.00024354571859563424,
      "loss": 4.1977,
      "mean_token_accuracy": 0.2800738841295242,
      "num_tokens": 15716755.0,
      "step": 19715
    },
    {
      "entropy": 4.631353569030762,
      "epoch": 5.575346338705117,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00024344387131152223,
      "loss": 4.2338,
      "mean_token_accuracy": 0.2641952529549599,
      "num_tokens": 15720694.0,
      "step": 19720
    },
    {
      "entropy": 4.669527006149292,
      "epoch": 5.576759966072943,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024334203062244597,
      "loss": 4.2358,
      "mean_token_accuracy": 0.2689785614609718,
      "num_tokens": 15724659.0,
      "step": 19725
    },
    {
      "entropy": 4.668330717086792,
      "epoch": 5.578173593440769,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00024324019654968948,
      "loss": 4.262,
      "mean_token_accuracy": 0.2766865938901901,
      "num_tokens": 15728214.0,
      "step": 19730
    },
    {
      "entropy": 4.747522783279419,
      "epoch": 5.579587220808595,
      "grad_norm": 1.6640625,
      "learning_rate": 0.00024313836911453562,
      "loss": 4.3472,
      "mean_token_accuracy": 0.265335151553154,
      "num_tokens": 15732149.0,
      "step": 19735
    },
    {
      "entropy": 4.546465492248535,
      "epoch": 5.581000848176421,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0002430365483382656,
      "loss": 4.0791,
      "mean_token_accuracy": 0.30231479555368423,
      "num_tokens": 15736128.0,
      "step": 19740
    },
    {
      "entropy": 4.630575037002563,
      "epoch": 5.582414475544247,
      "grad_norm": 1.65625,
      "learning_rate": 0.00024293473424215928,
      "loss": 4.3328,
      "mean_token_accuracy": 0.2662311315536499,
      "num_tokens": 15739956.0,
      "step": 19745
    },
    {
      "entropy": 4.5313104629516605,
      "epoch": 5.583828102912072,
      "grad_norm": 1.8125,
      "learning_rate": 0.00024283292684749514,
      "loss": 4.0769,
      "mean_token_accuracy": 0.29772662073373796,
      "num_tokens": 15743873.0,
      "step": 19750
    },
    {
      "entropy": 4.549833965301514,
      "epoch": 5.585241730279899,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002427311261755505,
      "loss": 4.2297,
      "mean_token_accuracy": 0.2822910413146019,
      "num_tokens": 15747687.0,
      "step": 19755
    },
    {
      "entropy": 4.555965518951416,
      "epoch": 5.586655357647724,
      "grad_norm": 1.6171875,
      "learning_rate": 0.00024262933224760092,
      "loss": 4.2199,
      "mean_token_accuracy": 0.2837903842329979,
      "num_tokens": 15751641.0,
      "step": 19760
    },
    {
      "entropy": 4.612254571914673,
      "epoch": 5.58806898501555,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002425275450849207,
      "loss": 4.2767,
      "mean_token_accuracy": 0.27603536993265154,
      "num_tokens": 15755747.0,
      "step": 19765
    },
    {
      "entropy": 4.734816074371338,
      "epoch": 5.589482612383375,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002424257647087827,
      "loss": 4.2653,
      "mean_token_accuracy": 0.2665874406695366,
      "num_tokens": 15759595.0,
      "step": 19770
    },
    {
      "entropy": 4.634348964691162,
      "epoch": 5.590896239751202,
      "grad_norm": 1.6875,
      "learning_rate": 0.0002423239911404585,
      "loss": 4.1632,
      "mean_token_accuracy": 0.2868500009179115,
      "num_tokens": 15763649.0,
      "step": 19775
    },
    {
      "entropy": 4.695278263092041,
      "epoch": 5.5923098671190274,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002422222244012181,
      "loss": 4.308,
      "mean_token_accuracy": 0.26301450878381727,
      "num_tokens": 15767619.0,
      "step": 19780
    },
    {
      "entropy": 4.666971445083618,
      "epoch": 5.593723494486853,
      "grad_norm": 1.734375,
      "learning_rate": 0.00024212046451233015,
      "loss": 4.28,
      "mean_token_accuracy": 0.27173508852720263,
      "num_tokens": 15771494.0,
      "step": 19785
    },
    {
      "entropy": 4.67068657875061,
      "epoch": 5.5951371218546795,
      "grad_norm": 1.609375,
      "learning_rate": 0.00024201871149506166,
      "loss": 4.2315,
      "mean_token_accuracy": 0.2823155790567398,
      "num_tokens": 15775421.0,
      "step": 19790
    },
    {
      "entropy": 4.645075035095215,
      "epoch": 5.596550749222505,
      "grad_norm": 1.7265625,
      "learning_rate": 0.00024191696537067858,
      "loss": 4.2558,
      "mean_token_accuracy": 0.2697598680853844,
      "num_tokens": 15779383.0,
      "step": 19795
    },
    {
      "entropy": 4.526032972335815,
      "epoch": 5.597964376590331,
      "grad_norm": 1.6328125,
      "learning_rate": 0.0002418152261604452,
      "loss": 4.1501,
      "mean_token_accuracy": 0.28109719306230546,
      "num_tokens": 15783452.0,
      "step": 19800
    },
    {
      "entropy": 4.645971727371216,
      "epoch": 5.599378003958156,
      "grad_norm": 1.6875,
      "learning_rate": 0.00024171349388562435,
      "loss": 4.3521,
      "mean_token_accuracy": 0.26674690693616865,
      "num_tokens": 15787488.0,
      "step": 19805
    },
    {
      "entropy": 4.607998323440552,
      "epoch": 5.600791631325983,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00024161176856747735,
      "loss": 4.1521,
      "mean_token_accuracy": 0.2825265884399414,
      "num_tokens": 15791676.0,
      "step": 19810
    },
    {
      "entropy": 4.494440746307373,
      "epoch": 5.602205258693808,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00024151005022726436,
      "loss": 4.098,
      "mean_token_accuracy": 0.29091261327266693,
      "num_tokens": 15795603.0,
      "step": 19815
    },
    {
      "entropy": 4.660259246826172,
      "epoch": 5.603618886061634,
      "grad_norm": 1.765625,
      "learning_rate": 0.0002414083388862437,
      "loss": 4.3258,
      "mean_token_accuracy": 0.2733373031020164,
      "num_tokens": 15799820.0,
      "step": 19820
    },
    {
      "entropy": 4.718490934371948,
      "epoch": 5.60503251342946,
      "grad_norm": 1.734375,
      "learning_rate": 0.00024130663456567255,
      "loss": 4.3982,
      "mean_token_accuracy": 0.25824896842241285,
      "num_tokens": 15803641.0,
      "step": 19825
    },
    {
      "entropy": 4.7006346702575685,
      "epoch": 5.606446140797286,
      "grad_norm": 1.703125,
      "learning_rate": 0.00024120493728680637,
      "loss": 4.1976,
      "mean_token_accuracy": 0.27732548415660857,
      "num_tokens": 15807273.0,
      "step": 19830
    },
    {
      "entropy": 4.659915447235107,
      "epoch": 5.6078597681651114,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0002411032470708993,
      "loss": 4.2119,
      "mean_token_accuracy": 0.27991697043180463,
      "num_tokens": 15811209.0,
      "step": 19835
    },
    {
      "entropy": 4.578808450698853,
      "epoch": 5.609273395532938,
      "grad_norm": 1.7421875,
      "learning_rate": 0.00024100156393920408,
      "loss": 4.1853,
      "mean_token_accuracy": 0.27597877383232117,
      "num_tokens": 15815274.0,
      "step": 19840
    },
    {
      "entropy": 4.514270544052124,
      "epoch": 5.6106870229007635,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00024089988791297168,
      "loss": 4.151,
      "mean_token_accuracy": 0.28049186915159224,
      "num_tokens": 15819325.0,
      "step": 19845
    },
    {
      "entropy": 4.725122499465942,
      "epoch": 5.612100650268589,
      "grad_norm": 1.75,
      "learning_rate": 0.0002407982190134518,
      "loss": 4.3411,
      "mean_token_accuracy": 0.2658290207386017,
      "num_tokens": 15823699.0,
      "step": 19850
    },
    {
      "entropy": 4.662112092971801,
      "epoch": 5.6135142776364155,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00024069655726189265,
      "loss": 4.2945,
      "mean_token_accuracy": 0.2629629343748093,
      "num_tokens": 15827605.0,
      "step": 19855
    },
    {
      "entropy": 4.623284482955933,
      "epoch": 5.614927905004241,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00024059490267954099,
      "loss": 4.2126,
      "mean_token_accuracy": 0.2871454402804375,
      "num_tokens": 15831679.0,
      "step": 19860
    },
    {
      "entropy": 4.694229555130005,
      "epoch": 5.616341532372067,
      "grad_norm": 1.5234375,
      "learning_rate": 0.00024049325528764193,
      "loss": 4.2404,
      "mean_token_accuracy": 0.276295530796051,
      "num_tokens": 15835684.0,
      "step": 19865
    },
    {
      "entropy": 4.713327789306641,
      "epoch": 5.617755159739892,
      "grad_norm": 1.609375,
      "learning_rate": 0.000240391615107439,
      "loss": 4.3265,
      "mean_token_accuracy": 0.269145530462265,
      "num_tokens": 15839590.0,
      "step": 19870
    },
    {
      "entropy": 4.626586103439331,
      "epoch": 5.619168787107719,
      "grad_norm": 1.8125,
      "learning_rate": 0.00024028998216017462,
      "loss": 4.2318,
      "mean_token_accuracy": 0.2843716964125633,
      "num_tokens": 15843495.0,
      "step": 19875
    },
    {
      "entropy": 4.656545829772949,
      "epoch": 5.620582414475544,
      "grad_norm": 1.671875,
      "learning_rate": 0.0002401883564670893,
      "loss": 4.2973,
      "mean_token_accuracy": 0.27081269174814226,
      "num_tokens": 15847506.0,
      "step": 19880
    },
    {
      "entropy": 4.648951196670533,
      "epoch": 5.62199604184337,
      "grad_norm": 1.75,
      "learning_rate": 0.00024008673804942223,
      "loss": 4.2528,
      "mean_token_accuracy": 0.27466840147972105,
      "num_tokens": 15851438.0,
      "step": 19885
    },
    {
      "entropy": 4.7030843734741214,
      "epoch": 5.6234096692111954,
      "grad_norm": 1.703125,
      "learning_rate": 0.0002399851269284109,
      "loss": 4.2132,
      "mean_token_accuracy": 0.2792676344513893,
      "num_tokens": 15855509.0,
      "step": 19890
    },
    {
      "entropy": 4.6753661155700685,
      "epoch": 5.624823296579022,
      "grad_norm": 1.59375,
      "learning_rate": 0.00023988352312529165,
      "loss": 4.19,
      "mean_token_accuracy": 0.27191189378499986,
      "num_tokens": 15859734.0,
      "step": 19895
    },
    {
      "entropy": 4.643299961090088,
      "epoch": 5.6262369239468475,
      "grad_norm": 1.65625,
      "learning_rate": 0.00023978192666129888,
      "loss": 4.2012,
      "mean_token_accuracy": 0.2783049136400223,
      "num_tokens": 15863682.0,
      "step": 19900
    },
    {
      "entropy": 4.677227783203125,
      "epoch": 5.627650551314673,
      "grad_norm": 1.578125,
      "learning_rate": 0.00023968033755766567,
      "loss": 4.2743,
      "mean_token_accuracy": 0.2798586830496788,
      "num_tokens": 15867876.0,
      "step": 19905
    },
    {
      "entropy": 4.726441097259522,
      "epoch": 5.6290641786824995,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0002395787558356234,
      "loss": 4.4376,
      "mean_token_accuracy": 0.26062889099121095,
      "num_tokens": 15871959.0,
      "step": 19910
    },
    {
      "entropy": 4.720166873931885,
      "epoch": 5.630477806050325,
      "grad_norm": 1.796875,
      "learning_rate": 0.00023947718151640218,
      "loss": 4.307,
      "mean_token_accuracy": 0.2770163580775261,
      "num_tokens": 15876127.0,
      "step": 19915
    },
    {
      "entropy": 4.705212926864624,
      "epoch": 5.631891433418151,
      "grad_norm": 1.8203125,
      "learning_rate": 0.00023937561462123035,
      "loss": 4.2548,
      "mean_token_accuracy": 0.28063480257987977,
      "num_tokens": 15879769.0,
      "step": 19920
    },
    {
      "entropy": 4.641681814193726,
      "epoch": 5.633305060785977,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00023927405517133483,
      "loss": 4.0994,
      "mean_token_accuracy": 0.29197325706481936,
      "num_tokens": 15883827.0,
      "step": 19925
    },
    {
      "entropy": 4.6417186737060545,
      "epoch": 5.634718688153803,
      "grad_norm": 1.7109375,
      "learning_rate": 0.00023917250318794076,
      "loss": 4.1458,
      "mean_token_accuracy": 0.293830069899559,
      "num_tokens": 15887824.0,
      "step": 19930
    },
    {
      "entropy": 4.580664920806885,
      "epoch": 5.636132315521628,
      "grad_norm": 1.71875,
      "learning_rate": 0.00023907095869227203,
      "loss": 4.2174,
      "mean_token_accuracy": 0.2807564675807953,
      "num_tokens": 15891634.0,
      "step": 19935
    },
    {
      "entropy": 4.6348131656646725,
      "epoch": 5.637545942889455,
      "grad_norm": 1.75,
      "learning_rate": 0.00023896942170555075,
      "loss": 4.2175,
      "mean_token_accuracy": 0.2917804718017578,
      "num_tokens": 15895624.0,
      "step": 19940
    },
    {
      "entropy": 4.589888858795166,
      "epoch": 5.63895957025728,
      "grad_norm": 1.6640625,
      "learning_rate": 0.0002388678922489975,
      "loss": 4.2076,
      "mean_token_accuracy": 0.27270770221948626,
      "num_tokens": 15899842.0,
      "step": 19945
    },
    {
      "entropy": 4.6108537197113035,
      "epoch": 5.640373197625106,
      "grad_norm": 1.8671875,
      "learning_rate": 0.00023876637034383135,
      "loss": 4.2181,
      "mean_token_accuracy": 0.2781824812293053,
      "num_tokens": 15903849.0,
      "step": 19950
    },
    {
      "entropy": 4.594720792770386,
      "epoch": 5.6417868249929315,
      "grad_norm": 1.796875,
      "learning_rate": 0.0002386648560112698,
      "loss": 4.17,
      "mean_token_accuracy": 0.2897926583886147,
      "num_tokens": 15907729.0,
      "step": 19955
    },
    {
      "entropy": 4.521228075027466,
      "epoch": 5.643200452360758,
      "grad_norm": 1.75,
      "learning_rate": 0.00023856334927252858,
      "loss": 4.1042,
      "mean_token_accuracy": 0.29376896321773527,
      "num_tokens": 15911515.0,
      "step": 19960
    },
    {
      "entropy": 4.6577842235565186,
      "epoch": 5.6446140797285835,
      "grad_norm": 1.609375,
      "learning_rate": 0.00023846185014882211,
      "loss": 4.2086,
      "mean_token_accuracy": 0.26874958276748656,
      "num_tokens": 15915072.0,
      "step": 19965
    },
    {
      "entropy": 4.702977895736694,
      "epoch": 5.646027707096409,
      "grad_norm": 1.671875,
      "learning_rate": 0.00023836035866136302,
      "loss": 4.2825,
      "mean_token_accuracy": 0.2709368035197258,
      "num_tokens": 15919215.0,
      "step": 19970
    },
    {
      "entropy": 4.675715017318725,
      "epoch": 5.647441334464236,
      "grad_norm": 1.71875,
      "learning_rate": 0.00023825887483136223,
      "loss": 4.2556,
      "mean_token_accuracy": 0.27941168248653414,
      "num_tokens": 15923022.0,
      "step": 19975
    },
    {
      "entropy": 4.654994916915894,
      "epoch": 5.648854961832061,
      "grad_norm": 1.6953125,
      "learning_rate": 0.00023815739868002955,
      "loss": 4.2835,
      "mean_token_accuracy": 0.28814811259508133,
      "num_tokens": 15926955.0,
      "step": 19980
    },
    {
      "entropy": 4.630128669738769,
      "epoch": 5.650268589199887,
      "grad_norm": 1.7265625,
      "learning_rate": 0.0002380559302285727,
      "loss": 4.2509,
      "mean_token_accuracy": 0.27425051778554915,
      "num_tokens": 15931434.0,
      "step": 19985
    },
    {
      "entropy": 4.624661302566528,
      "epoch": 5.651682216567712,
      "grad_norm": 1.609375,
      "learning_rate": 0.00023795446949819793,
      "loss": 4.2169,
      "mean_token_accuracy": 0.28021185994148257,
      "num_tokens": 15935694.0,
      "step": 19990
    },
    {
      "entropy": 4.675784778594971,
      "epoch": 5.653095843935539,
      "grad_norm": 1.6875,
      "learning_rate": 0.00023785301651010998,
      "loss": 4.2353,
      "mean_token_accuracy": 0.2924942135810852,
      "num_tokens": 15939646.0,
      "step": 19995
    },
    {
      "entropy": 4.674494314193725,
      "epoch": 5.654509471303364,
      "grad_norm": 1.75,
      "learning_rate": 0.0002377515712855119,
      "loss": 4.1844,
      "mean_token_accuracy": 0.2877335101366043,
      "num_tokens": 15943682.0,
      "step": 20000
    },
    {
      "epoch": 5.654509471303364,
      "eval_entropy": 4.478157658005842,
      "eval_loss": 4.945027828216553,
      "eval_mean_token_accuracy": 0.2531574366259119,
      "eval_num_tokens": 15943682.0,
      "eval_runtime": 6.6896,
      "eval_samples_per_second": 1686.806,
      "eval_steps_per_second": 210.926,
      "step": 20000
    }
  ],
  "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": 1.9352625990912e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}