{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22525421547174668,
  "eval_steps": 500,
  "global_step": 3500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 10.69194049835205,
      "epoch": 0.00032179173638820956,
      "grad_norm": 12.75,
      "learning_rate": 2e-06,
      "loss": 10.854,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11532.0,
      "step": 5
    },
    {
      "entropy": 10.691917896270752,
      "epoch": 0.0006435834727764191,
      "grad_norm": 16.25,
      "learning_rate": 4.5e-06,
      "loss": 10.7683,
      "mean_token_accuracy": 0.0002785657299682498,
      "num_tokens": 21733.0,
      "step": 10
    },
    {
      "entropy": 10.69142770767212,
      "epoch": 0.0009653752091646286,
      "grad_norm": 9.8125,
      "learning_rate": 7e-06,
      "loss": 10.5426,
      "mean_token_accuracy": 0.009345025499351322,
      "num_tokens": 32040.0,
      "step": 15
    },
    {
      "entropy": 10.687415981292725,
      "epoch": 0.0012871669455528382,
      "grad_norm": 6.09375,
      "learning_rate": 9.5e-06,
      "loss": 10.2464,
      "mean_token_accuracy": 0.035014028288424016,
      "num_tokens": 42645.0,
      "step": 20
    },
    {
      "entropy": 10.665467548370362,
      "epoch": 0.0016089586819410479,
      "grad_norm": 4.125,
      "learning_rate": 1.2e-05,
      "loss": 9.9923,
      "mean_token_accuracy": 0.03743175081908703,
      "num_tokens": 54374.0,
      "step": 25
    },
    {
      "entropy": 10.618990993499756,
      "epoch": 0.0019307504183292573,
      "grad_norm": 3.09375,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 9.8534,
      "mean_token_accuracy": 0.03550823628902435,
      "num_tokens": 66590.0,
      "step": 30
    },
    {
      "entropy": 10.615287399291992,
      "epoch": 0.002252542154717467,
      "grad_norm": 3.109375,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 9.7196,
      "mean_token_accuracy": 0.03219117037951946,
      "num_tokens": 78832.0,
      "step": 35
    },
    {
      "entropy": 10.627141761779786,
      "epoch": 0.0025743338911056765,
      "grad_norm": 2.703125,
      "learning_rate": 1.95e-05,
      "loss": 9.6667,
      "mean_token_accuracy": 0.046536854654550555,
      "num_tokens": 89158.0,
      "step": 40
    },
    {
      "entropy": 10.622460842132568,
      "epoch": 0.002896125627493886,
      "grad_norm": 2.671875,
      "learning_rate": 2.2e-05,
      "loss": 9.6509,
      "mean_token_accuracy": 0.04088520556688309,
      "num_tokens": 99546.0,
      "step": 45
    },
    {
      "entropy": 10.604833221435547,
      "epoch": 0.0032179173638820957,
      "grad_norm": 2.46875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.6314,
      "mean_token_accuracy": 0.043538968078792095,
      "num_tokens": 110770.0,
      "step": 50
    },
    {
      "entropy": 10.580564498901367,
      "epoch": 0.003539709100270305,
      "grad_norm": 2.40625,
      "learning_rate": 2.7e-05,
      "loss": 9.5314,
      "mean_token_accuracy": 0.04419821500778198,
      "num_tokens": 122202.0,
      "step": 55
    },
    {
      "entropy": 10.5360502243042,
      "epoch": 0.0038615008366585145,
      "grad_norm": 3.015625,
      "learning_rate": 2.95e-05,
      "loss": 9.4058,
      "mean_token_accuracy": 0.04586403630673885,
      "num_tokens": 134645.0,
      "step": 60
    },
    {
      "entropy": 10.41978416442871,
      "epoch": 0.004183292573046725,
      "grad_norm": 2.375,
      "learning_rate": 3.2e-05,
      "loss": 9.3388,
      "mean_token_accuracy": 0.05442853756248951,
      "num_tokens": 146962.0,
      "step": 65
    },
    {
      "entropy": 10.424466228485107,
      "epoch": 0.004505084309434934,
      "grad_norm": 2.78125,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.2536,
      "mean_token_accuracy": 0.05842369645833969,
      "num_tokens": 158200.0,
      "step": 70
    },
    {
      "entropy": 10.4311918258667,
      "epoch": 0.004826876045823143,
      "grad_norm": 2.34375,
      "learning_rate": 3.7e-05,
      "loss": 9.0956,
      "mean_token_accuracy": 0.0712333232164383,
      "num_tokens": 168437.0,
      "step": 75
    },
    {
      "entropy": 10.301731491088868,
      "epoch": 0.005148667782211353,
      "grad_norm": 2.40625,
      "learning_rate": 3.95e-05,
      "loss": 9.0239,
      "mean_token_accuracy": 0.06886589303612708,
      "num_tokens": 178953.0,
      "step": 80
    },
    {
      "entropy": 10.311075687408447,
      "epoch": 0.005470459518599562,
      "grad_norm": 2.4375,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 8.941,
      "mean_token_accuracy": 0.07178018912672997,
      "num_tokens": 189910.0,
      "step": 85
    },
    {
      "entropy": 10.289142322540282,
      "epoch": 0.005792251254987772,
      "grad_norm": 2.265625,
      "learning_rate": 4.45e-05,
      "loss": 8.8619,
      "mean_token_accuracy": 0.06988794207572938,
      "num_tokens": 201355.0,
      "step": 90
    },
    {
      "entropy": 10.256867218017579,
      "epoch": 0.006114042991375981,
      "grad_norm": 2.578125,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.7513,
      "mean_token_accuracy": 0.07205671109259129,
      "num_tokens": 212839.0,
      "step": 95
    },
    {
      "entropy": 10.217705726623535,
      "epoch": 0.0064358347277641914,
      "grad_norm": 2.1875,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.6829,
      "mean_token_accuracy": 0.07173232845962048,
      "num_tokens": 224453.0,
      "step": 100
    },
    {
      "entropy": 10.12640552520752,
      "epoch": 0.006757626464152401,
      "grad_norm": 2.125,
      "learning_rate": 5.2e-05,
      "loss": 8.5053,
      "mean_token_accuracy": 0.07860757485032081,
      "num_tokens": 236010.0,
      "step": 105
    },
    {
      "entropy": 10.044542694091797,
      "epoch": 0.00707941820054061,
      "grad_norm": 2.796875,
      "learning_rate": 5.45e-05,
      "loss": 8.4314,
      "mean_token_accuracy": 0.07552603259682655,
      "num_tokens": 247287.0,
      "step": 110
    },
    {
      "entropy": 10.028972911834718,
      "epoch": 0.00740120993692882,
      "grad_norm": 2.328125,
      "learning_rate": 5.7e-05,
      "loss": 8.3188,
      "mean_token_accuracy": 0.0754437144845724,
      "num_tokens": 258956.0,
      "step": 115
    },
    {
      "entropy": 9.909952545166016,
      "epoch": 0.007723001673317029,
      "grad_norm": 1.9765625,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.2245,
      "mean_token_accuracy": 0.08083591908216477,
      "num_tokens": 270356.0,
      "step": 120
    },
    {
      "entropy": 9.780113220214844,
      "epoch": 0.008044793409705238,
      "grad_norm": 2.21875,
      "learning_rate": 6.2e-05,
      "loss": 8.0979,
      "mean_token_accuracy": 0.08179116398096084,
      "num_tokens": 281711.0,
      "step": 125
    },
    {
      "entropy": 9.635583400726318,
      "epoch": 0.00836658514609345,
      "grad_norm": 2.140625,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.9786,
      "mean_token_accuracy": 0.08390471749007702,
      "num_tokens": 291995.0,
      "step": 130
    },
    {
      "entropy": 9.590358066558839,
      "epoch": 0.008688376882481658,
      "grad_norm": 2.203125,
      "learning_rate": 6.7e-05,
      "loss": 7.9706,
      "mean_token_accuracy": 0.07811799943447113,
      "num_tokens": 304222.0,
      "step": 135
    },
    {
      "entropy": 9.40874490737915,
      "epoch": 0.009010168618869868,
      "grad_norm": 1.7109375,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.8908,
      "mean_token_accuracy": 0.0770251952111721,
      "num_tokens": 316104.0,
      "step": 140
    },
    {
      "entropy": 9.228489303588868,
      "epoch": 0.009331960355258077,
      "grad_norm": 1.828125,
      "learning_rate": 7.2e-05,
      "loss": 7.7264,
      "mean_token_accuracy": 0.0914445884525776,
      "num_tokens": 327511.0,
      "step": 145
    },
    {
      "entropy": 9.002948188781739,
      "epoch": 0.009653752091646286,
      "grad_norm": 1.96875,
      "learning_rate": 7.45e-05,
      "loss": 7.5834,
      "mean_token_accuracy": 0.0950602874159813,
      "num_tokens": 338523.0,
      "step": 150
    },
    {
      "entropy": 8.787215518951417,
      "epoch": 0.009975543828034497,
      "grad_norm": 1.8828125,
      "learning_rate": 7.7e-05,
      "loss": 7.5805,
      "mean_token_accuracy": 0.09019249975681305,
      "num_tokens": 350247.0,
      "step": 155
    },
    {
      "entropy": 8.719713592529297,
      "epoch": 0.010297335564422706,
      "grad_norm": 1.328125,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.4723,
      "mean_token_accuracy": 0.09186024889349938,
      "num_tokens": 362539.0,
      "step": 160
    },
    {
      "entropy": 8.630542182922364,
      "epoch": 0.010619127300810915,
      "grad_norm": 1.3359375,
      "learning_rate": 8.2e-05,
      "loss": 7.5412,
      "mean_token_accuracy": 0.08679113686084747,
      "num_tokens": 374773.0,
      "step": 165
    },
    {
      "entropy": 8.505796527862548,
      "epoch": 0.010940919037199124,
      "grad_norm": 1.703125,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.3633,
      "mean_token_accuracy": 0.08832342699170112,
      "num_tokens": 385292.0,
      "step": 170
    },
    {
      "entropy": 8.319984436035156,
      "epoch": 0.011262710773587334,
      "grad_norm": 1.3984375,
      "learning_rate": 8.7e-05,
      "loss": 7.3207,
      "mean_token_accuracy": 0.09274004399776459,
      "num_tokens": 396662.0,
      "step": 175
    },
    {
      "entropy": 8.170382118225097,
      "epoch": 0.011584502509975544,
      "grad_norm": 1.453125,
      "learning_rate": 8.95e-05,
      "loss": 7.3063,
      "mean_token_accuracy": 0.09627912491559983,
      "num_tokens": 407836.0,
      "step": 180
    },
    {
      "entropy": 8.038551759719848,
      "epoch": 0.011906294246363754,
      "grad_norm": 1.59375,
      "learning_rate": 9.2e-05,
      "loss": 7.1831,
      "mean_token_accuracy": 0.10445229858160018,
      "num_tokens": 418626.0,
      "step": 185
    },
    {
      "entropy": 7.990694236755371,
      "epoch": 0.012228085982751963,
      "grad_norm": 1.40625,
      "learning_rate": 9.45e-05,
      "loss": 7.2592,
      "mean_token_accuracy": 0.09324222803115845,
      "num_tokens": 430714.0,
      "step": 190
    },
    {
      "entropy": 8.09174370765686,
      "epoch": 0.012549877719140172,
      "grad_norm": 1.375,
      "learning_rate": 9.7e-05,
      "loss": 7.2672,
      "mean_token_accuracy": 0.08939176499843597,
      "num_tokens": 442144.0,
      "step": 195
    },
    {
      "entropy": 7.912285184860229,
      "epoch": 0.012871669455528383,
      "grad_norm": 1.578125,
      "learning_rate": 9.95e-05,
      "loss": 7.3035,
      "mean_token_accuracy": 0.09566279277205467,
      "num_tokens": 454338.0,
      "step": 200
    },
    {
      "entropy": 7.9603489398956295,
      "epoch": 0.013193461191916592,
      "grad_norm": 1.53125,
      "learning_rate": 0.000102,
      "loss": 7.1568,
      "mean_token_accuracy": 0.10046104565262795,
      "num_tokens": 465493.0,
      "step": 205
    },
    {
      "entropy": 7.831574201583862,
      "epoch": 0.013515252928304801,
      "grad_norm": 1.6796875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.2202,
      "mean_token_accuracy": 0.09616154506802559,
      "num_tokens": 476753.0,
      "step": 210
    },
    {
      "entropy": 7.802942085266113,
      "epoch": 0.01383704466469301,
      "grad_norm": 1.5703125,
      "learning_rate": 0.000107,
      "loss": 7.1711,
      "mean_token_accuracy": 0.10192405954003333,
      "num_tokens": 488229.0,
      "step": 215
    },
    {
      "entropy": 7.800758314132691,
      "epoch": 0.01415883640108122,
      "grad_norm": 2.265625,
      "learning_rate": 0.0001095,
      "loss": 7.1771,
      "mean_token_accuracy": 0.09801913350820542,
      "num_tokens": 500060.0,
      "step": 220
    },
    {
      "entropy": 7.8091026782989506,
      "epoch": 0.01448062813746943,
      "grad_norm": 1.578125,
      "learning_rate": 0.000112,
      "loss": 7.1403,
      "mean_token_accuracy": 0.09416564255952835,
      "num_tokens": 512368.0,
      "step": 225
    },
    {
      "entropy": 7.6497923851013185,
      "epoch": 0.01480241987385764,
      "grad_norm": 1.5703125,
      "learning_rate": 0.0001145,
      "loss": 7.0236,
      "mean_token_accuracy": 0.10454239919781685,
      "num_tokens": 522546.0,
      "step": 230
    },
    {
      "entropy": 7.6424744606018065,
      "epoch": 0.015124211610245849,
      "grad_norm": 1.421875,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.106,
      "mean_token_accuracy": 0.10524164140224457,
      "num_tokens": 535441.0,
      "step": 235
    },
    {
      "entropy": 7.6858056545257565,
      "epoch": 0.015446003346634058,
      "grad_norm": 1.515625,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.0976,
      "mean_token_accuracy": 0.10354602560400963,
      "num_tokens": 547684.0,
      "step": 240
    },
    {
      "entropy": 7.588476228713989,
      "epoch": 0.015767795083022267,
      "grad_norm": 1.390625,
      "learning_rate": 0.000122,
      "loss": 7.1511,
      "mean_token_accuracy": 0.1012592412531376,
      "num_tokens": 558822.0,
      "step": 245
    },
    {
      "entropy": 7.707156753540039,
      "epoch": 0.016089586819410476,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0001245,
      "loss": 7.0529,
      "mean_token_accuracy": 0.10268613547086716,
      "num_tokens": 570343.0,
      "step": 250
    },
    {
      "entropy": 7.550539875030518,
      "epoch": 0.016411378555798686,
      "grad_norm": 1.4921875,
      "learning_rate": 0.000127,
      "loss": 7.0108,
      "mean_token_accuracy": 0.10793713703751565,
      "num_tokens": 581035.0,
      "step": 255
    },
    {
      "entropy": 7.545463514328003,
      "epoch": 0.0167331702921869,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0001295,
      "loss": 6.9715,
      "mean_token_accuracy": 0.103653634339571,
      "num_tokens": 594185.0,
      "step": 260
    },
    {
      "entropy": 7.447913885116577,
      "epoch": 0.017054962028575107,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000132,
      "loss": 6.9392,
      "mean_token_accuracy": 0.10639498457312584,
      "num_tokens": 605353.0,
      "step": 265
    },
    {
      "entropy": 7.561796474456787,
      "epoch": 0.017376753764963317,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.0345,
      "mean_token_accuracy": 0.10664849355816841,
      "num_tokens": 617477.0,
      "step": 270
    },
    {
      "entropy": 7.435778045654297,
      "epoch": 0.017698545501351526,
      "grad_norm": 1.71875,
      "learning_rate": 0.00013700000000000002,
      "loss": 6.9534,
      "mean_token_accuracy": 0.10435824021697045,
      "num_tokens": 627578.0,
      "step": 275
    },
    {
      "entropy": 7.4433777809143065,
      "epoch": 0.018020337237739735,
      "grad_norm": 1.5859375,
      "learning_rate": 0.0001395,
      "loss": 6.955,
      "mean_token_accuracy": 0.10273718237876892,
      "num_tokens": 639559.0,
      "step": 280
    },
    {
      "entropy": 7.458190488815307,
      "epoch": 0.018342128974127944,
      "grad_norm": 1.7578125,
      "learning_rate": 0.00014199999999999998,
      "loss": 6.962,
      "mean_token_accuracy": 0.1059377796947956,
      "num_tokens": 650407.0,
      "step": 285
    },
    {
      "entropy": 7.382494831085205,
      "epoch": 0.018663920710516153,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0001445,
      "loss": 6.8653,
      "mean_token_accuracy": 0.11828745529055595,
      "num_tokens": 662931.0,
      "step": 290
    },
    {
      "entropy": 7.4552978515625,
      "epoch": 0.018985712446904363,
      "grad_norm": 1.7421875,
      "learning_rate": 0.000147,
      "loss": 6.9431,
      "mean_token_accuracy": 0.11408839300274849,
      "num_tokens": 674686.0,
      "step": 295
    },
    {
      "entropy": 7.3142343044281,
      "epoch": 0.01930750418329257,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0001495,
      "loss": 6.8417,
      "mean_token_accuracy": 0.11158959865570069,
      "num_tokens": 685463.0,
      "step": 300
    },
    {
      "entropy": 7.456843852996826,
      "epoch": 0.01962929591968078,
      "grad_norm": 1.4453125,
      "learning_rate": 0.000152,
      "loss": 7.0395,
      "mean_token_accuracy": 0.11052512302994728,
      "num_tokens": 697397.0,
      "step": 305
    },
    {
      "entropy": 7.265558385848999,
      "epoch": 0.019951087656068994,
      "grad_norm": 1.515625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.8946,
      "mean_token_accuracy": 0.11685808449983597,
      "num_tokens": 708268.0,
      "step": 310
    },
    {
      "entropy": 7.297895240783691,
      "epoch": 0.020272879392457203,
      "grad_norm": 1.296875,
      "learning_rate": 0.000157,
      "loss": 6.8953,
      "mean_token_accuracy": 0.11156268939375877,
      "num_tokens": 720056.0,
      "step": 315
    },
    {
      "entropy": 7.366024732589722,
      "epoch": 0.020594671128845412,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0001595,
      "loss": 6.8679,
      "mean_token_accuracy": 0.1150714986026287,
      "num_tokens": 730835.0,
      "step": 320
    },
    {
      "entropy": 7.369055366516113,
      "epoch": 0.02091646286523362,
      "grad_norm": 1.328125,
      "learning_rate": 0.000162,
      "loss": 7.0326,
      "mean_token_accuracy": 0.108493110537529,
      "num_tokens": 742428.0,
      "step": 325
    },
    {
      "entropy": 7.274074840545654,
      "epoch": 0.02123825460162183,
      "grad_norm": 1.296875,
      "learning_rate": 0.00016450000000000001,
      "loss": 6.8866,
      "mean_token_accuracy": 0.1098675474524498,
      "num_tokens": 752147.0,
      "step": 330
    },
    {
      "entropy": 7.4121346950531,
      "epoch": 0.02156004633801004,
      "grad_norm": 1.328125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.9166,
      "mean_token_accuracy": 0.10750498473644257,
      "num_tokens": 763415.0,
      "step": 335
    },
    {
      "entropy": 7.304456186294556,
      "epoch": 0.02188183807439825,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.8702,
      "mean_token_accuracy": 0.10661920979619026,
      "num_tokens": 775175.0,
      "step": 340
    },
    {
      "entropy": 7.224200105667114,
      "epoch": 0.022203629810786458,
      "grad_norm": 1.359375,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8099,
      "mean_token_accuracy": 0.11928229033946991,
      "num_tokens": 785211.0,
      "step": 345
    },
    {
      "entropy": 7.240478086471557,
      "epoch": 0.022525421547174667,
      "grad_norm": 1.359375,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.869,
      "mean_token_accuracy": 0.11734312251210213,
      "num_tokens": 795946.0,
      "step": 350
    },
    {
      "entropy": 7.1357990264892575,
      "epoch": 0.02284721328356288,
      "grad_norm": 1.375,
      "learning_rate": 0.000177,
      "loss": 6.7101,
      "mean_token_accuracy": 0.12463226988911628,
      "num_tokens": 806529.0,
      "step": 355
    },
    {
      "entropy": 7.220741224288941,
      "epoch": 0.02316900501995109,
      "grad_norm": 1.140625,
      "learning_rate": 0.0001795,
      "loss": 6.8095,
      "mean_token_accuracy": 0.11709680408239365,
      "num_tokens": 817591.0,
      "step": 360
    },
    {
      "entropy": 7.153127908706665,
      "epoch": 0.023490796756339298,
      "grad_norm": 1.265625,
      "learning_rate": 0.000182,
      "loss": 6.8188,
      "mean_token_accuracy": 0.11239580363035202,
      "num_tokens": 828765.0,
      "step": 365
    },
    {
      "entropy": 7.202557897567749,
      "epoch": 0.023812588492727507,
      "grad_norm": 1.34375,
      "learning_rate": 0.0001845,
      "loss": 6.8802,
      "mean_token_accuracy": 0.1095006950199604,
      "num_tokens": 839442.0,
      "step": 370
    },
    {
      "entropy": 7.075786590576172,
      "epoch": 0.024134380229115716,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000187,
      "loss": 6.6783,
      "mean_token_accuracy": 0.11973611861467362,
      "num_tokens": 850471.0,
      "step": 375
    },
    {
      "entropy": 7.15129132270813,
      "epoch": 0.024456171965503926,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0001895,
      "loss": 6.7819,
      "mean_token_accuracy": 0.11633364856243134,
      "num_tokens": 861333.0,
      "step": 380
    },
    {
      "entropy": 7.195761871337891,
      "epoch": 0.024777963701892135,
      "grad_norm": 1.3359375,
      "learning_rate": 0.000192,
      "loss": 6.8712,
      "mean_token_accuracy": 0.11771900430321694,
      "num_tokens": 871915.0,
      "step": 385
    },
    {
      "entropy": 7.193553447723389,
      "epoch": 0.025099755438280344,
      "grad_norm": 1.34375,
      "learning_rate": 0.0001945,
      "loss": 6.8448,
      "mean_token_accuracy": 0.12030173614621162,
      "num_tokens": 884006.0,
      "step": 390
    },
    {
      "entropy": 7.1253314971923825,
      "epoch": 0.025421547174668553,
      "grad_norm": 1.609375,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.7347,
      "mean_token_accuracy": 0.1115813210606575,
      "num_tokens": 896393.0,
      "step": 395
    },
    {
      "entropy": 7.097787714004516,
      "epoch": 0.025743338911056766,
      "grad_norm": 1.3125,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.833,
      "mean_token_accuracy": 0.11548972800374031,
      "num_tokens": 907202.0,
      "step": 400
    },
    {
      "entropy": 6.983932209014893,
      "epoch": 0.026065130647444975,
      "grad_norm": 1.4375,
      "learning_rate": 0.000202,
      "loss": 6.513,
      "mean_token_accuracy": 0.12307731807231903,
      "num_tokens": 917063.0,
      "step": 405
    },
    {
      "entropy": 7.058181858062744,
      "epoch": 0.026386922383833184,
      "grad_norm": 1.390625,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.7534,
      "mean_token_accuracy": 0.11989552527666092,
      "num_tokens": 927593.0,
      "step": 410
    },
    {
      "entropy": 7.128173398971557,
      "epoch": 0.026708714120221393,
      "grad_norm": 1.390625,
      "learning_rate": 0.000207,
      "loss": 6.7926,
      "mean_token_accuracy": 0.11310968622565269,
      "num_tokens": 938942.0,
      "step": 415
    },
    {
      "entropy": 7.121332836151123,
      "epoch": 0.027030505856609603,
      "grad_norm": 1.7421875,
      "learning_rate": 0.0002095,
      "loss": 6.8324,
      "mean_token_accuracy": 0.11872464641928673,
      "num_tokens": 949912.0,
      "step": 420
    },
    {
      "entropy": 6.970332765579224,
      "epoch": 0.02735229759299781,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000212,
      "loss": 6.7019,
      "mean_token_accuracy": 0.12167602479457855,
      "num_tokens": 960959.0,
      "step": 425
    },
    {
      "entropy": 6.912196826934815,
      "epoch": 0.02767408932938602,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0002145,
      "loss": 6.607,
      "mean_token_accuracy": 0.12273273840546609,
      "num_tokens": 972908.0,
      "step": 430
    },
    {
      "entropy": 7.092562818527222,
      "epoch": 0.02799588106577423,
      "grad_norm": 1.34375,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.8127,
      "mean_token_accuracy": 0.10646726936101913,
      "num_tokens": 984115.0,
      "step": 435
    },
    {
      "entropy": 7.0991950035095215,
      "epoch": 0.02831767280216244,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002195,
      "loss": 6.862,
      "mean_token_accuracy": 0.11290395855903626,
      "num_tokens": 995697.0,
      "step": 440
    },
    {
      "entropy": 7.099041700363159,
      "epoch": 0.02863946453855065,
      "grad_norm": 1.390625,
      "learning_rate": 0.000222,
      "loss": 6.7961,
      "mean_token_accuracy": 0.11020021885633469,
      "num_tokens": 1007918.0,
      "step": 445
    },
    {
      "entropy": 7.085301446914673,
      "epoch": 0.02896125627493886,
      "grad_norm": 1.34375,
      "learning_rate": 0.0002245,
      "loss": 6.6977,
      "mean_token_accuracy": 0.12074138224124908,
      "num_tokens": 1017699.0,
      "step": 450
    },
    {
      "entropy": 7.032927083969116,
      "epoch": 0.02928304801132707,
      "grad_norm": 1.5078125,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.7618,
      "mean_token_accuracy": 0.12060360610485077,
      "num_tokens": 1029464.0,
      "step": 455
    },
    {
      "entropy": 7.0487706661224365,
      "epoch": 0.02960483974771528,
      "grad_norm": 1.296875,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7003,
      "mean_token_accuracy": 0.12441192641854286,
      "num_tokens": 1040241.0,
      "step": 460
    },
    {
      "entropy": 7.015468263626099,
      "epoch": 0.02992663148410349,
      "grad_norm": 1.34375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7482,
      "mean_token_accuracy": 0.11376301795244217,
      "num_tokens": 1052712.0,
      "step": 465
    },
    {
      "entropy": 6.893531227111817,
      "epoch": 0.030248423220491698,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6977,
      "mean_token_accuracy": 0.11767899766564369,
      "num_tokens": 1063569.0,
      "step": 470
    },
    {
      "entropy": 6.897192907333374,
      "epoch": 0.030570214956879907,
      "grad_norm": 1.46875,
      "learning_rate": 0.000237,
      "loss": 6.6361,
      "mean_token_accuracy": 0.12202028259634971,
      "num_tokens": 1074314.0,
      "step": 475
    },
    {
      "entropy": 7.062028932571411,
      "epoch": 0.030892006693268116,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0002395,
      "loss": 6.8464,
      "mean_token_accuracy": 0.1108027994632721,
      "num_tokens": 1086084.0,
      "step": 480
    },
    {
      "entropy": 7.020650625228882,
      "epoch": 0.031213798429656325,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000242,
      "loss": 6.7314,
      "mean_token_accuracy": 0.11836976334452629,
      "num_tokens": 1097116.0,
      "step": 485
    },
    {
      "entropy": 6.910376405715942,
      "epoch": 0.031535590166044535,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002445,
      "loss": 6.6115,
      "mean_token_accuracy": 0.12779454588890077,
      "num_tokens": 1108561.0,
      "step": 490
    },
    {
      "entropy": 6.949523115158081,
      "epoch": 0.031857381902432744,
      "grad_norm": 1.25,
      "learning_rate": 0.000247,
      "loss": 6.6493,
      "mean_token_accuracy": 0.12263330891728401,
      "num_tokens": 1120847.0,
      "step": 495
    },
    {
      "entropy": 6.9015583992004395,
      "epoch": 0.03217917363882095,
      "grad_norm": 1.375,
      "learning_rate": 0.0002495,
      "loss": 6.6208,
      "mean_token_accuracy": 0.1280594363808632,
      "num_tokens": 1132470.0,
      "step": 500
    },
    {
      "entropy": 6.869023752212525,
      "epoch": 0.03250096537520916,
      "grad_norm": 1.3671875,
      "learning_rate": 0.000252,
      "loss": 6.6691,
      "mean_token_accuracy": 0.1259506419301033,
      "num_tokens": 1144966.0,
      "step": 505
    },
    {
      "entropy": 6.998987627029419,
      "epoch": 0.03282275711159737,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002545,
      "loss": 6.6196,
      "mean_token_accuracy": 0.12219893932342529,
      "num_tokens": 1156413.0,
      "step": 510
    },
    {
      "entropy": 6.722204494476318,
      "epoch": 0.03314454884798558,
      "grad_norm": 1.171875,
      "learning_rate": 0.000257,
      "loss": 6.5395,
      "mean_token_accuracy": 0.1305614896118641,
      "num_tokens": 1167458.0,
      "step": 515
    },
    {
      "entropy": 6.878732395172119,
      "epoch": 0.0334663405843738,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002595,
      "loss": 6.6024,
      "mean_token_accuracy": 0.11557823643088341,
      "num_tokens": 1179015.0,
      "step": 520
    },
    {
      "entropy": 6.836273288726806,
      "epoch": 0.033788132320762006,
      "grad_norm": 1.296875,
      "learning_rate": 0.000262,
      "loss": 6.5532,
      "mean_token_accuracy": 0.12147759348154068,
      "num_tokens": 1189616.0,
      "step": 525
    },
    {
      "entropy": 6.74482798576355,
      "epoch": 0.034109924057150215,
      "grad_norm": 1.5390625,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.4766,
      "mean_token_accuracy": 0.12241575941443443,
      "num_tokens": 1200283.0,
      "step": 530
    },
    {
      "entropy": 6.877546215057373,
      "epoch": 0.034431715793538424,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.6075,
      "mean_token_accuracy": 0.12261987179517746,
      "num_tokens": 1211379.0,
      "step": 535
    },
    {
      "entropy": 6.855060482025147,
      "epoch": 0.03475350752992663,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.5896,
      "mean_token_accuracy": 0.1253580369055271,
      "num_tokens": 1222790.0,
      "step": 540
    },
    {
      "entropy": 6.859920072555542,
      "epoch": 0.03507529926631484,
      "grad_norm": 1.28125,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.5984,
      "mean_token_accuracy": 0.12788782343268396,
      "num_tokens": 1233801.0,
      "step": 545
    },
    {
      "entropy": 6.815193700790405,
      "epoch": 0.03539709100270305,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002745,
      "loss": 6.5805,
      "mean_token_accuracy": 0.12837038785219193,
      "num_tokens": 1245805.0,
      "step": 550
    },
    {
      "entropy": 6.745665168762207,
      "epoch": 0.03571888273909126,
      "grad_norm": 1.34375,
      "learning_rate": 0.000277,
      "loss": 6.5159,
      "mean_token_accuracy": 0.1349610432982445,
      "num_tokens": 1257107.0,
      "step": 555
    },
    {
      "entropy": 6.797294044494629,
      "epoch": 0.03604067447547947,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002795,
      "loss": 6.6277,
      "mean_token_accuracy": 0.12336362823843956,
      "num_tokens": 1268179.0,
      "step": 560
    },
    {
      "entropy": 6.868199729919434,
      "epoch": 0.03636246621186768,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.555,
      "mean_token_accuracy": 0.13341980427503586,
      "num_tokens": 1278442.0,
      "step": 565
    },
    {
      "entropy": 6.89066162109375,
      "epoch": 0.03668425794825589,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0002845,
      "loss": 6.6539,
      "mean_token_accuracy": 0.12870992943644524,
      "num_tokens": 1288623.0,
      "step": 570
    },
    {
      "entropy": 6.751723289489746,
      "epoch": 0.0370060496846441,
      "grad_norm": 1.3515625,
      "learning_rate": 0.000287,
      "loss": 6.5708,
      "mean_token_accuracy": 0.12620823234319686,
      "num_tokens": 1300981.0,
      "step": 575
    },
    {
      "entropy": 6.691731691360474,
      "epoch": 0.03732784142103231,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0002895,
      "loss": 6.4173,
      "mean_token_accuracy": 0.1356960915029049,
      "num_tokens": 1312496.0,
      "step": 580
    },
    {
      "entropy": 6.689237260818482,
      "epoch": 0.037649633157420516,
      "grad_norm": 1.078125,
      "learning_rate": 0.000292,
      "loss": 6.4696,
      "mean_token_accuracy": 0.13737872168421744,
      "num_tokens": 1323716.0,
      "step": 585
    },
    {
      "entropy": 6.680966424942016,
      "epoch": 0.037971424893808725,
      "grad_norm": 1.40625,
      "learning_rate": 0.0002945,
      "loss": 6.4844,
      "mean_token_accuracy": 0.1304626889526844,
      "num_tokens": 1335571.0,
      "step": 590
    },
    {
      "entropy": 6.6498072147369385,
      "epoch": 0.038293216630196934,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000297,
      "loss": 6.4942,
      "mean_token_accuracy": 0.13134991824626924,
      "num_tokens": 1347105.0,
      "step": 595
    },
    {
      "entropy": 6.765724849700928,
      "epoch": 0.03861500836658514,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002995,
      "loss": 6.5692,
      "mean_token_accuracy": 0.12424814403057098,
      "num_tokens": 1357640.0,
      "step": 600
    },
    {
      "entropy": 6.669166326522827,
      "epoch": 0.03893680010297335,
      "grad_norm": 1.28125,
      "learning_rate": 0.000302,
      "loss": 6.4196,
      "mean_token_accuracy": 0.13372000828385353,
      "num_tokens": 1368608.0,
      "step": 605
    },
    {
      "entropy": 6.75927505493164,
      "epoch": 0.03925859183936156,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003045,
      "loss": 6.5127,
      "mean_token_accuracy": 0.1207577109336853,
      "num_tokens": 1379853.0,
      "step": 610
    },
    {
      "entropy": 6.664323329925537,
      "epoch": 0.03958038357574978,
      "grad_norm": 1.25,
      "learning_rate": 0.000307,
      "loss": 6.4645,
      "mean_token_accuracy": 0.12720839604735373,
      "num_tokens": 1390630.0,
      "step": 615
    },
    {
      "entropy": 6.689991140365601,
      "epoch": 0.03990217531213799,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003095,
      "loss": 6.4366,
      "mean_token_accuracy": 0.13198668509721756,
      "num_tokens": 1401997.0,
      "step": 620
    },
    {
      "entropy": 6.634037685394287,
      "epoch": 0.040223967048526196,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000312,
      "loss": 6.536,
      "mean_token_accuracy": 0.1256448470056057,
      "num_tokens": 1413611.0,
      "step": 625
    },
    {
      "entropy": 6.804963159561157,
      "epoch": 0.040545758784914405,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003145,
      "loss": 6.6033,
      "mean_token_accuracy": 0.11850515380501747,
      "num_tokens": 1424639.0,
      "step": 630
    },
    {
      "entropy": 6.6990772724151615,
      "epoch": 0.040867550521302615,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000317,
      "loss": 6.5229,
      "mean_token_accuracy": 0.13084252625703813,
      "num_tokens": 1434802.0,
      "step": 635
    },
    {
      "entropy": 6.687566947937012,
      "epoch": 0.041189342257690824,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003195,
      "loss": 6.466,
      "mean_token_accuracy": 0.12602303102612494,
      "num_tokens": 1446942.0,
      "step": 640
    },
    {
      "entropy": 6.730192232131958,
      "epoch": 0.04151113399407903,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000322,
      "loss": 6.4753,
      "mean_token_accuracy": 0.12809659466147422,
      "num_tokens": 1458527.0,
      "step": 645
    },
    {
      "entropy": 6.495309972763062,
      "epoch": 0.04183292573046724,
      "grad_norm": 1.25,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.3417,
      "mean_token_accuracy": 0.13759842664003372,
      "num_tokens": 1469065.0,
      "step": 650
    },
    {
      "entropy": 6.648279190063477,
      "epoch": 0.04215471746685545,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.4067,
      "mean_token_accuracy": 0.13652142956852914,
      "num_tokens": 1479425.0,
      "step": 655
    },
    {
      "entropy": 6.641835165023804,
      "epoch": 0.04247650920324366,
      "grad_norm": 1.21875,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.4447,
      "mean_token_accuracy": 0.12913303449749947,
      "num_tokens": 1490661.0,
      "step": 660
    },
    {
      "entropy": 6.750196409225464,
      "epoch": 0.04279830093963187,
      "grad_norm": 1.4375,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.5532,
      "mean_token_accuracy": 0.12641846686601638,
      "num_tokens": 1502638.0,
      "step": 665
    },
    {
      "entropy": 6.594930934906006,
      "epoch": 0.04312009267602008,
      "grad_norm": 1.3125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.4584,
      "mean_token_accuracy": 0.1351857751607895,
      "num_tokens": 1512899.0,
      "step": 670
    },
    {
      "entropy": 6.697818374633789,
      "epoch": 0.04344188441240829,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000337,
      "loss": 6.4577,
      "mean_token_accuracy": 0.12779420837759972,
      "num_tokens": 1523914.0,
      "step": 675
    },
    {
      "entropy": 6.640419149398804,
      "epoch": 0.0437636761487965,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0003395,
      "loss": 6.4947,
      "mean_token_accuracy": 0.13016460612416267,
      "num_tokens": 1535498.0,
      "step": 680
    },
    {
      "entropy": 6.607001733779907,
      "epoch": 0.044085467885184706,
      "grad_norm": 1.296875,
      "learning_rate": 0.000342,
      "loss": 6.4335,
      "mean_token_accuracy": 0.13075666204094888,
      "num_tokens": 1545516.0,
      "step": 685
    },
    {
      "entropy": 6.660410928726196,
      "epoch": 0.044407259621572916,
      "grad_norm": 1.15625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.5728,
      "mean_token_accuracy": 0.12807587459683417,
      "num_tokens": 1556241.0,
      "step": 690
    },
    {
      "entropy": 6.73777813911438,
      "epoch": 0.044729051357961125,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000347,
      "loss": 6.4331,
      "mean_token_accuracy": 0.13530867099761962,
      "num_tokens": 1567229.0,
      "step": 695
    },
    {
      "entropy": 6.50676908493042,
      "epoch": 0.045050843094349334,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003495,
      "loss": 6.4477,
      "mean_token_accuracy": 0.12948062345385553,
      "num_tokens": 1578703.0,
      "step": 700
    },
    {
      "entropy": 6.497309255599975,
      "epoch": 0.04537263483073754,
      "grad_norm": 1.28125,
      "learning_rate": 0.000352,
      "loss": 6.3204,
      "mean_token_accuracy": 0.1343964785337448,
      "num_tokens": 1589839.0,
      "step": 705
    },
    {
      "entropy": 6.575838232040406,
      "epoch": 0.04569442656712576,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0003545,
      "loss": 6.3711,
      "mean_token_accuracy": 0.13568960577249528,
      "num_tokens": 1600277.0,
      "step": 710
    },
    {
      "entropy": 6.6485175609588625,
      "epoch": 0.04601621830351397,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000357,
      "loss": 6.5288,
      "mean_token_accuracy": 0.12966799288988112,
      "num_tokens": 1611894.0,
      "step": 715
    },
    {
      "entropy": 6.500547075271607,
      "epoch": 0.04633801003990218,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595,
      "loss": 6.3377,
      "mean_token_accuracy": 0.13781237751245498,
      "num_tokens": 1625165.0,
      "step": 720
    },
    {
      "entropy": 6.593970775604248,
      "epoch": 0.04665980177629039,
      "grad_norm": 1.15625,
      "learning_rate": 0.000362,
      "loss": 6.4257,
      "mean_token_accuracy": 0.13297282308340072,
      "num_tokens": 1636606.0,
      "step": 725
    },
    {
      "entropy": 6.567788457870483,
      "epoch": 0.046981593512678596,
      "grad_norm": 1.125,
      "learning_rate": 0.0003645,
      "loss": 6.3967,
      "mean_token_accuracy": 0.13927650898694993,
      "num_tokens": 1648528.0,
      "step": 730
    },
    {
      "entropy": 6.566985034942627,
      "epoch": 0.047303385249066805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000367,
      "loss": 6.4528,
      "mean_token_accuracy": 0.12781663686037065,
      "num_tokens": 1661540.0,
      "step": 735
    },
    {
      "entropy": 6.468529415130615,
      "epoch": 0.047625176985455014,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003695,
      "loss": 6.4452,
      "mean_token_accuracy": 0.12587922364473342,
      "num_tokens": 1673505.0,
      "step": 740
    },
    {
      "entropy": 6.576410531997681,
      "epoch": 0.047946968721843224,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000372,
      "loss": 6.3876,
      "mean_token_accuracy": 0.1358407884836197,
      "num_tokens": 1685297.0,
      "step": 745
    },
    {
      "entropy": 6.451522922515869,
      "epoch": 0.04826876045823143,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0003745,
      "loss": 6.3365,
      "mean_token_accuracy": 0.1313149116933346,
      "num_tokens": 1694649.0,
      "step": 750
    },
    {
      "entropy": 6.586151742935181,
      "epoch": 0.04859055219461964,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000377,
      "loss": 6.4005,
      "mean_token_accuracy": 0.1305807001888752,
      "num_tokens": 1705720.0,
      "step": 755
    },
    {
      "entropy": 6.542773485183716,
      "epoch": 0.04891234393100785,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003795,
      "loss": 6.485,
      "mean_token_accuracy": 0.12867986187338828,
      "num_tokens": 1717634.0,
      "step": 760
    },
    {
      "entropy": 6.590516901016235,
      "epoch": 0.04923413566739606,
      "grad_norm": 1.359375,
      "learning_rate": 0.000382,
      "loss": 6.3945,
      "mean_token_accuracy": 0.12419472783803939,
      "num_tokens": 1729954.0,
      "step": 765
    },
    {
      "entropy": 6.474939203262329,
      "epoch": 0.04955592740378427,
      "grad_norm": 1.234375,
      "learning_rate": 0.0003845,
      "loss": 6.3725,
      "mean_token_accuracy": 0.1370876781642437,
      "num_tokens": 1742173.0,
      "step": 770
    },
    {
      "entropy": 6.553148174285889,
      "epoch": 0.04987771914017248,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.3779,
      "mean_token_accuracy": 0.13408709540963173,
      "num_tokens": 1753682.0,
      "step": 775
    },
    {
      "entropy": 6.405833148956299,
      "epoch": 0.05019951087656069,
      "grad_norm": 1.171875,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.3635,
      "mean_token_accuracy": 0.13386630415916442,
      "num_tokens": 1764358.0,
      "step": 780
    },
    {
      "entropy": 6.4643104553222654,
      "epoch": 0.0505213026129489,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3553,
      "mean_token_accuracy": 0.13426898568868637,
      "num_tokens": 1775997.0,
      "step": 785
    },
    {
      "entropy": 6.612279558181763,
      "epoch": 0.050843094349337106,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.3925,
      "mean_token_accuracy": 0.12899470925331116,
      "num_tokens": 1787684.0,
      "step": 790
    },
    {
      "entropy": 6.357057094573975,
      "epoch": 0.051164886085725315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.3918,
      "mean_token_accuracy": 0.12932194545865058,
      "num_tokens": 1799291.0,
      "step": 795
    },
    {
      "entropy": 6.46887321472168,
      "epoch": 0.05148667782211353,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0003995,
      "loss": 6.3496,
      "mean_token_accuracy": 0.1412258230149746,
      "num_tokens": 1811976.0,
      "step": 800
    },
    {
      "entropy": 6.483779716491699,
      "epoch": 0.05180846955850174,
      "grad_norm": 1.234375,
      "learning_rate": 0.000402,
      "loss": 6.3974,
      "mean_token_accuracy": 0.12525486797094346,
      "num_tokens": 1824751.0,
      "step": 805
    },
    {
      "entropy": 6.414471673965454,
      "epoch": 0.05213026129488995,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004045,
      "loss": 6.3602,
      "mean_token_accuracy": 0.12907202765345574,
      "num_tokens": 1836676.0,
      "step": 810
    },
    {
      "entropy": 6.464478349685669,
      "epoch": 0.05245205303127816,
      "grad_norm": 1.203125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.2536,
      "mean_token_accuracy": 0.13829391375184058,
      "num_tokens": 1847803.0,
      "step": 815
    },
    {
      "entropy": 6.367223501205444,
      "epoch": 0.05277384476766637,
      "grad_norm": 1.125,
      "learning_rate": 0.0004095,
      "loss": 6.2625,
      "mean_token_accuracy": 0.12957098931074143,
      "num_tokens": 1859965.0,
      "step": 820
    },
    {
      "entropy": 6.430704641342163,
      "epoch": 0.05309563650405458,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000412,
      "loss": 6.2909,
      "mean_token_accuracy": 0.13130465224385263,
      "num_tokens": 1870950.0,
      "step": 825
    },
    {
      "entropy": 6.414825677871704,
      "epoch": 0.05341742824044279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004145,
      "loss": 6.279,
      "mean_token_accuracy": 0.14454965069890022,
      "num_tokens": 1882501.0,
      "step": 830
    },
    {
      "entropy": 6.502061080932617,
      "epoch": 0.053739219976830996,
      "grad_norm": 1.3828125,
      "learning_rate": 0.000417,
      "loss": 6.4967,
      "mean_token_accuracy": 0.13128656819462775,
      "num_tokens": 1895554.0,
      "step": 835
    },
    {
      "entropy": 6.353726434707641,
      "epoch": 0.054061011713219205,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004195,
      "loss": 6.3737,
      "mean_token_accuracy": 0.13272848650813102,
      "num_tokens": 1905982.0,
      "step": 840
    },
    {
      "entropy": 6.482131671905518,
      "epoch": 0.054382803449607414,
      "grad_norm": 1.203125,
      "learning_rate": 0.000422,
      "loss": 6.3661,
      "mean_token_accuracy": 0.1328873485326767,
      "num_tokens": 1917468.0,
      "step": 845
    },
    {
      "entropy": 6.4495148181915285,
      "epoch": 0.05470459518599562,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004245,
      "loss": 6.3424,
      "mean_token_accuracy": 0.13536591157317163,
      "num_tokens": 1927743.0,
      "step": 850
    },
    {
      "entropy": 6.410049724578857,
      "epoch": 0.05502638692238383,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000427,
      "loss": 6.2786,
      "mean_token_accuracy": 0.13671474531292915,
      "num_tokens": 1937721.0,
      "step": 855
    },
    {
      "entropy": 6.431154823303222,
      "epoch": 0.05534817865877204,
      "grad_norm": 1.125,
      "learning_rate": 0.0004295,
      "loss": 6.2665,
      "mean_token_accuracy": 0.13668931648135185,
      "num_tokens": 1949130.0,
      "step": 860
    },
    {
      "entropy": 6.5185517311096195,
      "epoch": 0.05566997039516025,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000432,
      "loss": 6.4461,
      "mean_token_accuracy": 0.12062366977334023,
      "num_tokens": 1961132.0,
      "step": 865
    },
    {
      "entropy": 6.3538782596588135,
      "epoch": 0.05599176213154846,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004345,
      "loss": 6.2272,
      "mean_token_accuracy": 0.14166060090065002,
      "num_tokens": 1971904.0,
      "step": 870
    },
    {
      "entropy": 6.424667167663574,
      "epoch": 0.05631355386793667,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000437,
      "loss": 6.3388,
      "mean_token_accuracy": 0.13849016577005385,
      "num_tokens": 1982246.0,
      "step": 875
    },
    {
      "entropy": 6.534474086761475,
      "epoch": 0.05663534560432488,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004395,
      "loss": 6.3182,
      "mean_token_accuracy": 0.1408639907836914,
      "num_tokens": 1993687.0,
      "step": 880
    },
    {
      "entropy": 6.196937131881714,
      "epoch": 0.05695713734071309,
      "grad_norm": 1.125,
      "learning_rate": 0.000442,
      "loss": 6.2508,
      "mean_token_accuracy": 0.14266577661037444,
      "num_tokens": 2004335.0,
      "step": 885
    },
    {
      "entropy": 6.529156827926636,
      "epoch": 0.0572789290771013,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004445,
      "loss": 6.4175,
      "mean_token_accuracy": 0.1389109380543232,
      "num_tokens": 2014875.0,
      "step": 890
    },
    {
      "entropy": 6.415221643447876,
      "epoch": 0.05760072081348951,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000447,
      "loss": 6.3593,
      "mean_token_accuracy": 0.14172261878848075,
      "num_tokens": 2025354.0,
      "step": 895
    },
    {
      "entropy": 6.325968408584595,
      "epoch": 0.05792251254987772,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2146,
      "mean_token_accuracy": 0.14011863470077515,
      "num_tokens": 2036306.0,
      "step": 900
    },
    {
      "entropy": 6.427873182296753,
      "epoch": 0.05824430428626593,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.3058,
      "mean_token_accuracy": 0.13970114216208457,
      "num_tokens": 2046232.0,
      "step": 905
    },
    {
      "entropy": 6.283337879180908,
      "epoch": 0.05856609602265414,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2769,
      "mean_token_accuracy": 0.13600695356726647,
      "num_tokens": 2056876.0,
      "step": 910
    },
    {
      "entropy": 6.4581481456756595,
      "epoch": 0.05888788775904235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.5089,
      "mean_token_accuracy": 0.12785363271832467,
      "num_tokens": 2069419.0,
      "step": 915
    },
    {
      "entropy": 6.396502590179443,
      "epoch": 0.05920967949543056,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.2433,
      "mean_token_accuracy": 0.13423478305339814,
      "num_tokens": 2079542.0,
      "step": 920
    },
    {
      "entropy": 6.267205047607422,
      "epoch": 0.05953147123181877,
      "grad_norm": 1.09375,
      "learning_rate": 0.000462,
      "loss": 6.2204,
      "mean_token_accuracy": 0.14118367657065392,
      "num_tokens": 2090022.0,
      "step": 925
    },
    {
      "entropy": 6.398083829879761,
      "epoch": 0.05985326296820698,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004645,
      "loss": 6.3334,
      "mean_token_accuracy": 0.13829256296157838,
      "num_tokens": 2100572.0,
      "step": 930
    },
    {
      "entropy": 6.3885338306427,
      "epoch": 0.060175054704595186,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000467,
      "loss": 6.2758,
      "mean_token_accuracy": 0.13688874542713164,
      "num_tokens": 2111868.0,
      "step": 935
    },
    {
      "entropy": 6.308207416534424,
      "epoch": 0.060496846440983396,
      "grad_norm": 1.125,
      "learning_rate": 0.0004695,
      "loss": 6.3022,
      "mean_token_accuracy": 0.13644594624638556,
      "num_tokens": 2122908.0,
      "step": 940
    },
    {
      "entropy": 6.246605396270752,
      "epoch": 0.060818638177371605,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000472,
      "loss": 6.1958,
      "mean_token_accuracy": 0.1409742645919323,
      "num_tokens": 2134667.0,
      "step": 945
    },
    {
      "entropy": 6.289280796051026,
      "epoch": 0.061140429913759814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004745,
      "loss": 6.2301,
      "mean_token_accuracy": 0.1406254932284355,
      "num_tokens": 2146183.0,
      "step": 950
    },
    {
      "entropy": 6.305873441696167,
      "epoch": 0.06146222165014802,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000477,
      "loss": 6.2783,
      "mean_token_accuracy": 0.13354712650179862,
      "num_tokens": 2157846.0,
      "step": 955
    },
    {
      "entropy": 6.296875238418579,
      "epoch": 0.06178401338653623,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795,
      "loss": 6.2284,
      "mean_token_accuracy": 0.14504019021987916,
      "num_tokens": 2168454.0,
      "step": 960
    },
    {
      "entropy": 6.360051965713501,
      "epoch": 0.06210580512292444,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000482,
      "loss": 6.3116,
      "mean_token_accuracy": 0.1352120630443096,
      "num_tokens": 2179457.0,
      "step": 965
    },
    {
      "entropy": 6.2297309875488285,
      "epoch": 0.06242759685931265,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004845,
      "loss": 6.2111,
      "mean_token_accuracy": 0.1407732456922531,
      "num_tokens": 2190780.0,
      "step": 970
    },
    {
      "entropy": 6.341225433349609,
      "epoch": 0.06274938859570087,
      "grad_norm": 1.1875,
      "learning_rate": 0.000487,
      "loss": 6.3645,
      "mean_token_accuracy": 0.12729336619377135,
      "num_tokens": 2201000.0,
      "step": 975
    },
    {
      "entropy": 6.179282903671265,
      "epoch": 0.06307118033208907,
      "grad_norm": 1.640625,
      "learning_rate": 0.0004895,
      "loss": 6.2269,
      "mean_token_accuracy": 0.1418264776468277,
      "num_tokens": 2213208.0,
      "step": 980
    },
    {
      "entropy": 6.357757472991944,
      "epoch": 0.06339297206847729,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000492,
      "loss": 6.2115,
      "mean_token_accuracy": 0.13693007826805115,
      "num_tokens": 2224550.0,
      "step": 985
    },
    {
      "entropy": 6.2303062915802006,
      "epoch": 0.06371476380486549,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004945,
      "loss": 6.2138,
      "mean_token_accuracy": 0.13869421929121017,
      "num_tokens": 2236341.0,
      "step": 990
    },
    {
      "entropy": 6.285347557067871,
      "epoch": 0.0640365555412537,
      "grad_norm": 1.15625,
      "learning_rate": 0.000497,
      "loss": 6.296,
      "mean_token_accuracy": 0.1306707262992859,
      "num_tokens": 2247462.0,
      "step": 995
    },
    {
      "entropy": 6.282100057601928,
      "epoch": 0.0643583472776419,
      "grad_norm": 1.578125,
      "learning_rate": 0.0004995,
      "loss": 6.1985,
      "mean_token_accuracy": 0.13407785445451736,
      "num_tokens": 2259136.0,
      "step": 1000
    },
    {
      "entropy": 6.265222263336182,
      "epoch": 0.06468013901403012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000499998026082006,
      "loss": 6.2387,
      "mean_token_accuracy": 0.13681457713246345,
      "num_tokens": 2270064.0,
      "step": 1005
    },
    {
      "entropy": 6.234803342819214,
      "epoch": 0.06500193075041832,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999900070995136,
      "loss": 6.1861,
      "mean_token_accuracy": 0.1417514629662037,
      "num_tokens": 2281104.0,
      "step": 1010
    },
    {
      "entropy": 6.302186393737793,
      "epoch": 0.06532372248680654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999758199023239,
      "loss": 6.3205,
      "mean_token_accuracy": 0.1325412757694721,
      "num_tokens": 2292906.0,
      "step": 1015
    },
    {
      "entropy": 6.214407348632813,
      "epoch": 0.06564551422319474,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999554648793858,
      "loss": 6.2279,
      "mean_token_accuracy": 0.13934953659772872,
      "num_tokens": 2304224.0,
      "step": 1020
    },
    {
      "entropy": 6.445757865905762,
      "epoch": 0.06596730595958296,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999289425887425,
      "loss": 6.2842,
      "mean_token_accuracy": 0.12618643864989282,
      "num_tokens": 2314904.0,
      "step": 1025
    },
    {
      "entropy": 6.222538137435913,
      "epoch": 0.06628909769597116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998962537575161,
      "loss": 6.3476,
      "mean_token_accuracy": 0.1342654272913933,
      "num_tokens": 2326352.0,
      "step": 1030
    },
    {
      "entropy": 6.3435663223266605,
      "epoch": 0.06661088943235938,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004998573992818874,
      "loss": 6.2911,
      "mean_token_accuracy": 0.14127109572291374,
      "num_tokens": 2337950.0,
      "step": 1035
    },
    {
      "entropy": 6.155977773666382,
      "epoch": 0.0669326811687476,
      "grad_norm": 1.125,
      "learning_rate": 0.0004998123802270715,
      "loss": 6.083,
      "mean_token_accuracy": 0.14717964977025985,
      "num_tokens": 2347847.0,
      "step": 1040
    },
    {
      "entropy": 6.259258651733399,
      "epoch": 0.0672544729051358,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004997611978272886,
      "loss": 6.1665,
      "mean_token_accuracy": 0.1488075628876686,
      "num_tokens": 2357835.0,
      "step": 1045
    },
    {
      "entropy": 6.239679956436158,
      "epoch": 0.06757626464152401,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004997038534857298,
      "loss": 6.2632,
      "mean_token_accuracy": 0.13494880571961404,
      "num_tokens": 2369081.0,
      "step": 1050
    },
    {
      "entropy": 6.193429708480835,
      "epoch": 0.06789805637791221,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004996403487745194,
      "loss": 6.2406,
      "mean_token_accuracy": 0.13763956353068352,
      "num_tokens": 2380243.0,
      "step": 1055
    },
    {
      "entropy": 6.336066341400146,
      "epoch": 0.06821984811430043,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499570685434671,
      "loss": 6.2294,
      "mean_token_accuracy": 0.14424258098006248,
      "num_tokens": 2392514.0,
      "step": 1060
    },
    {
      "entropy": 6.202878427505493,
      "epoch": 0.06854163985068863,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004994948653760405,
      "loss": 6.2643,
      "mean_token_accuracy": 0.1274372950196266,
      "num_tokens": 2404104.0,
      "step": 1065
    },
    {
      "entropy": 6.213591384887695,
      "epoch": 0.06886343158707685,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994128906772729,
      "loss": 6.1634,
      "mean_token_accuracy": 0.14054710865020753,
      "num_tokens": 2415740.0,
      "step": 1070
    },
    {
      "entropy": 6.262695741653443,
      "epoch": 0.06918522332346505,
      "grad_norm": 1.078125,
      "learning_rate": 0.000499324763585746,
      "loss": 6.2432,
      "mean_token_accuracy": 0.13545622304081917,
      "num_tokens": 2426952.0,
      "step": 1075
    },
    {
      "entropy": 6.151621580123901,
      "epoch": 0.06950701505985327,
      "grad_norm": 1.125,
      "learning_rate": 0.0004992304865175085,
      "loss": 6.1999,
      "mean_token_accuracy": 0.1342847764492035,
      "num_tokens": 2438085.0,
      "step": 1080
    },
    {
      "entropy": 6.251348161697388,
      "epoch": 0.06982880679624147,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004991300620572138,
      "loss": 6.1787,
      "mean_token_accuracy": 0.1454763390123844,
      "num_tokens": 2448426.0,
      "step": 1085
    },
    {
      "entropy": 6.1301616668701175,
      "epoch": 0.07015059853262968,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004990234929580494,
      "loss": 6.1395,
      "mean_token_accuracy": 0.13527224361896514,
      "num_tokens": 2459127.0,
      "step": 1090
    },
    {
      "entropy": 6.3425384044647215,
      "epoch": 0.07047239026901789,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989107821416609,
      "loss": 6.2899,
      "mean_token_accuracy": 0.133147981017828,
      "num_tokens": 2470913.0,
      "step": 1095
    },
    {
      "entropy": 6.121605920791626,
      "epoch": 0.0707941820054061,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004987919326980723,
      "loss": 6.2032,
      "mean_token_accuracy": 0.14263079017400743,
      "num_tokens": 2481214.0,
      "step": 1100
    },
    {
      "entropy": 6.2000030994415285,
      "epoch": 0.0711159737417943,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986669478856011,
      "loss": 6.1731,
      "mean_token_accuracy": 0.141845715790987,
      "num_tokens": 2493050.0,
      "step": 1105
    },
    {
      "entropy": 6.254453802108765,
      "epoch": 0.07143776547818252,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985358311307688,
      "loss": 6.1826,
      "mean_token_accuracy": 0.14366479367017745,
      "num_tokens": 2504406.0,
      "step": 1110
    },
    {
      "entropy": 6.1378173828125,
      "epoch": 0.07175955721457072,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004983985860282081,
      "loss": 6.258,
      "mean_token_accuracy": 0.1386260747909546,
      "num_tokens": 2514643.0,
      "step": 1115
    },
    {
      "entropy": 6.324863004684448,
      "epoch": 0.07208134895095894,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004982552163405623,
      "loss": 6.1865,
      "mean_token_accuracy": 0.14131532087922097,
      "num_tokens": 2524924.0,
      "step": 1120
    },
    {
      "entropy": 6.1816126823425295,
      "epoch": 0.07240314068734714,
      "grad_norm": 1.125,
      "learning_rate": 0.0004981057259983839,
      "loss": 6.2598,
      "mean_token_accuracy": 0.14671514555811882,
      "num_tokens": 2535447.0,
      "step": 1125
    },
    {
      "entropy": 6.25083646774292,
      "epoch": 0.07272493242373536,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004979501191000262,
      "loss": 6.1821,
      "mean_token_accuracy": 0.14264000579714775,
      "num_tokens": 2547308.0,
      "step": 1130
    },
    {
      "entropy": 6.189724493026733,
      "epoch": 0.07304672416012357,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004977883999115311,
      "loss": 6.1076,
      "mean_token_accuracy": 0.1410246193408966,
      "num_tokens": 2557319.0,
      "step": 1135
    },
    {
      "entropy": 6.084145212173462,
      "epoch": 0.07336851589651178,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004976205728665113,
      "loss": 6.096,
      "mean_token_accuracy": 0.14267066717147828,
      "num_tokens": 2570099.0,
      "step": 1140
    },
    {
      "entropy": 6.136931037902832,
      "epoch": 0.07369030763289999,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004974466425660307,
      "loss": 6.089,
      "mean_token_accuracy": 0.14493471682071685,
      "num_tokens": 2580492.0,
      "step": 1145
    },
    {
      "entropy": 6.092673969268799,
      "epoch": 0.0740120993692882,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004972666137784759,
      "loss": 6.1957,
      "mean_token_accuracy": 0.14505093023180962,
      "num_tokens": 2591564.0,
      "step": 1150
    },
    {
      "entropy": 6.166257286071778,
      "epoch": 0.07433389110567641,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004970804914394271,
      "loss": 6.0627,
      "mean_token_accuracy": 0.1405116766691208,
      "num_tokens": 2601611.0,
      "step": 1155
    },
    {
      "entropy": 6.123493099212647,
      "epoch": 0.07465568284206461,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004968882806515225,
      "loss": 6.0904,
      "mean_token_accuracy": 0.14291229397058486,
      "num_tokens": 2611945.0,
      "step": 1160
    },
    {
      "entropy": 6.176377582550049,
      "epoch": 0.07497747457845283,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004966899866843177,
      "loss": 6.1528,
      "mean_token_accuracy": 0.14379807338118553,
      "num_tokens": 2622520.0,
      "step": 1165
    },
    {
      "entropy": 6.097160577774048,
      "epoch": 0.07529926631484103,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496485614974142,
      "loss": 6.1361,
      "mean_token_accuracy": 0.14614415317773818,
      "num_tokens": 2634054.0,
      "step": 1170
    },
    {
      "entropy": 6.183728504180908,
      "epoch": 0.07562105805122925,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004962751711239492,
      "loss": 6.2013,
      "mean_token_accuracy": 0.14067911207675934,
      "num_tokens": 2645760.0,
      "step": 1175
    },
    {
      "entropy": 6.19887409210205,
      "epoch": 0.07594284978761745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004960586609031636,
      "loss": 6.1602,
      "mean_token_accuracy": 0.13677565157413482,
      "num_tokens": 2657599.0,
      "step": 1180
    },
    {
      "entropy": 6.193633031845093,
      "epoch": 0.07626464152400567,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004958360902475224,
      "loss": 6.2417,
      "mean_token_accuracy": 0.13924393579363822,
      "num_tokens": 2669962.0,
      "step": 1185
    },
    {
      "entropy": 6.172778654098511,
      "epoch": 0.07658643326039387,
      "grad_norm": 1.0,
      "learning_rate": 0.0004956074652589125,
      "loss": 6.1701,
      "mean_token_accuracy": 0.14136232435703278,
      "num_tokens": 2682416.0,
      "step": 1190
    },
    {
      "entropy": 6.083469009399414,
      "epoch": 0.07690822499678208,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004953727922052035,
      "loss": 6.0961,
      "mean_token_accuracy": 0.14915101528167723,
      "num_tokens": 2692095.0,
      "step": 1195
    },
    {
      "entropy": 6.182906866073608,
      "epoch": 0.07723001673317029,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004951320775200756,
      "loss": 6.1357,
      "mean_token_accuracy": 0.1437535248696804,
      "num_tokens": 2702705.0,
      "step": 1200
    },
    {
      "entropy": 6.094074058532715,
      "epoch": 0.0775518084695585,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004948853278028436,
      "loss": 6.1216,
      "mean_token_accuracy": 0.14604368135333062,
      "num_tokens": 2713280.0,
      "step": 1205
    },
    {
      "entropy": 6.195901679992676,
      "epoch": 0.0778736002059467,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004946325498182755,
      "loss": 6.159,
      "mean_token_accuracy": 0.14376554638147354,
      "num_tokens": 2725242.0,
      "step": 1210
    },
    {
      "entropy": 6.108611440658569,
      "epoch": 0.07819539194233492,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943737504964076,
      "loss": 6.1032,
      "mean_token_accuracy": 0.14644696861505507,
      "num_tokens": 2736895.0,
      "step": 1215
    },
    {
      "entropy": 6.193878316879273,
      "epoch": 0.07851718367872312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494108936932354,
      "loss": 6.0713,
      "mean_token_accuracy": 0.15232568681240083,
      "num_tokens": 2748919.0,
      "step": 1220
    },
    {
      "entropy": 6.20891547203064,
      "epoch": 0.07883897541511134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004938381163861124,
      "loss": 6.0809,
      "mean_token_accuracy": 0.14824541360139848,
      "num_tokens": 2760158.0,
      "step": 1225
    },
    {
      "entropy": 6.085719776153565,
      "epoch": 0.07916076715149956,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004935612962823645,
      "loss": 6.1627,
      "mean_token_accuracy": 0.1409171588718891,
      "num_tokens": 2771968.0,
      "step": 1230
    },
    {
      "entropy": 6.098658227920533,
      "epoch": 0.07948255888788776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004932784842102739,
      "loss": 6.0537,
      "mean_token_accuracy": 0.14979517236351966,
      "num_tokens": 2781611.0,
      "step": 1235
    },
    {
      "entropy": 5.994823932647705,
      "epoch": 0.07980435062427597,
      "grad_norm": 1.0,
      "learning_rate": 0.0004929896879232758,
      "loss": 6.0589,
      "mean_token_accuracy": 0.15257512778043747,
      "num_tokens": 2792800.0,
      "step": 1240
    },
    {
      "entropy": 6.129728031158447,
      "epoch": 0.08012614236066418,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004926949153388668,
      "loss": 6.059,
      "mean_token_accuracy": 0.14132918566465377,
      "num_tokens": 2803625.0,
      "step": 1245
    },
    {
      "entropy": 6.114728021621704,
      "epoch": 0.08044793409705239,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004923941745383859,
      "loss": 6.119,
      "mean_token_accuracy": 0.1411253795027733,
      "num_tokens": 2815836.0,
      "step": 1250
    },
    {
      "entropy": 6.101113939285279,
      "epoch": 0.0807697258334406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000492087473766794,
      "loss": 6.0744,
      "mean_token_accuracy": 0.1473376877605915,
      "num_tokens": 2827590.0,
      "step": 1255
    },
    {
      "entropy": 6.179189443588257,
      "epoch": 0.08109151756982881,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000491774821432448,
      "loss": 6.2092,
      "mean_token_accuracy": 0.14346556589007378,
      "num_tokens": 2839469.0,
      "step": 1260
    },
    {
      "entropy": 6.161050510406494,
      "epoch": 0.08141330930621701,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004914562261068693,
      "loss": 6.1127,
      "mean_token_accuracy": 0.14351990967988967,
      "num_tokens": 2851424.0,
      "step": 1265
    },
    {
      "entropy": 5.995301198959351,
      "epoch": 0.08173510104260523,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004911316965245098,
      "loss": 5.9972,
      "mean_token_accuracy": 0.14706434458494186,
      "num_tokens": 2862818.0,
      "step": 1270
    },
    {
      "entropy": 6.226345491409302,
      "epoch": 0.08205689277899343,
      "grad_norm": 1.0,
      "learning_rate": 0.000490801241582512,
      "loss": 6.1326,
      "mean_token_accuracy": 0.1427297107875347,
      "num_tokens": 2874795.0,
      "step": 1275
    },
    {
      "entropy": 6.0090076446533205,
      "epoch": 0.08237868451538165,
      "grad_norm": 0.984375,
      "learning_rate": 0.000490464870340465,
      "loss": 6.1186,
      "mean_token_accuracy": 0.14504720717668534,
      "num_tokens": 2884309.0,
      "step": 1280
    },
    {
      "entropy": 6.246135234832764,
      "epoch": 0.08270047625176985,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004901225920201563,
      "loss": 6.1841,
      "mean_token_accuracy": 0.13920361548662186,
      "num_tokens": 2895149.0,
      "step": 1285
    },
    {
      "entropy": 5.9457801342010494,
      "epoch": 0.08302226798815807,
      "grad_norm": 1.046875,
      "learning_rate": 0.000489774416005319,
      "loss": 5.953,
      "mean_token_accuracy": 0.15055820271372794,
      "num_tokens": 2906311.0,
      "step": 1290
    },
    {
      "entropy": 6.034160470962524,
      "epoch": 0.08334405972454627,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004894203518413742,
      "loss": 6.0461,
      "mean_token_accuracy": 0.15890783071517944,
      "num_tokens": 2918706.0,
      "step": 1295
    },
    {
      "entropy": 6.070583534240723,
      "epoch": 0.08366585146093448,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004890604092351701,
      "loss": 6.0141,
      "mean_token_accuracy": 0.1481314405798912,
      "num_tokens": 2929569.0,
      "step": 1300
    },
    {
      "entropy": 5.892015266418457,
      "epoch": 0.08398764319732269,
      "grad_norm": 1.0,
      "learning_rate": 0.000488694598054715,
      "loss": 5.9925,
      "mean_token_accuracy": 0.15369523614645003,
      "num_tokens": 2941286.0,
      "step": 1305
    },
    {
      "entropy": 6.097052526473999,
      "epoch": 0.0843094349337109,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004883229283289071,
      "loss": 6.0232,
      "mean_token_accuracy": 0.1539250060915947,
      "num_tokens": 2952490.0,
      "step": 1310
    },
    {
      "entropy": 6.123910570144654,
      "epoch": 0.0846312266700991,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00048794541024725993,
      "loss": 6.0977,
      "mean_token_accuracy": 0.14586264193058013,
      "num_tokens": 2964381.0,
      "step": 1315
    },
    {
      "entropy": 6.094405794143677,
      "epoch": 0.08495301840648732,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004875620541596221,
      "loss": 6.0476,
      "mean_token_accuracy": 0.14707987383008003,
      "num_tokens": 2974769.0,
      "step": 1320
    },
    {
      "entropy": 6.1349348545074465,
      "epoch": 0.08527481014287554,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048717287057589454,
      "loss": 6.1471,
      "mean_token_accuracy": 0.13929664865136146,
      "num_tokens": 2986333.0,
      "step": 1325
    },
    {
      "entropy": 6.084645557403564,
      "epoch": 0.08559660187926374,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004867778701657417,
      "loss": 6.0341,
      "mean_token_accuracy": 0.14553362131118774,
      "num_tokens": 2997642.0,
      "step": 1330
    },
    {
      "entropy": 6.0231305122375485,
      "epoch": 0.08591839361565196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00048637706375829955,
      "loss": 6.0573,
      "mean_token_accuracy": 0.15109685510396959,
      "num_tokens": 3009847.0,
      "step": 1335
    },
    {
      "entropy": 6.18050627708435,
      "epoch": 0.08624018535204016,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000485970462341878,
      "loss": 6.0891,
      "mean_token_accuracy": 0.1480099990963936,
      "num_tokens": 3019657.0,
      "step": 1340
    },
    {
      "entropy": 6.064053153991699,
      "epoch": 0.08656197708842837,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00048555807706366044,
      "loss": 6.1594,
      "mean_token_accuracy": 0.14406584575772285,
      "num_tokens": 3030507.0,
      "step": 1345
    },
    {
      "entropy": 6.17887053489685,
      "epoch": 0.08688376882481658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00048513991922939756,
      "loss": 6.1747,
      "mean_token_accuracy": 0.14376262128353118,
      "num_tokens": 3041751.0,
      "step": 1350
    },
    {
      "entropy": 6.123253107070923,
      "epoch": 0.08720556056120479,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048471600030309744,
      "loss": 6.0863,
      "mean_token_accuracy": 0.1487148091197014,
      "num_tokens": 3052754.0,
      "step": 1355
    },
    {
      "entropy": 6.192664575576782,
      "epoch": 0.087527352297593,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048428633190671186,
      "loss": 6.1462,
      "mean_token_accuracy": 0.1416320726275444,
      "num_tokens": 3064731.0,
      "step": 1360
    },
    {
      "entropy": 6.012926530838013,
      "epoch": 0.08784914403398121,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004838509258198167,
      "loss": 5.9544,
      "mean_token_accuracy": 0.1551859274506569,
      "num_tokens": 3075093.0,
      "step": 1365
    },
    {
      "entropy": 6.086758947372436,
      "epoch": 0.08817093577036941,
      "grad_norm": 1.0,
      "learning_rate": 0.00048340979397929,
      "loss": 6.112,
      "mean_token_accuracy": 0.1472473606467247,
      "num_tokens": 3086303.0,
      "step": 1370
    },
    {
      "entropy": 6.017410135269165,
      "epoch": 0.08849272750675763,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00048296294847898386,
      "loss": 6.0718,
      "mean_token_accuracy": 0.14377797544002532,
      "num_tokens": 3097916.0,
      "step": 1375
    },
    {
      "entropy": 6.118084335327149,
      "epoch": 0.08881451924314583,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004825104015693934,
      "loss": 5.997,
      "mean_token_accuracy": 0.15102456882596016,
      "num_tokens": 3109168.0,
      "step": 1380
    },
    {
      "entropy": 5.975411653518677,
      "epoch": 0.08913631097953405,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004820521656573208,
      "loss": 6.0083,
      "mean_token_accuracy": 0.15250723212957382,
      "num_tokens": 3120739.0,
      "step": 1385
    },
    {
      "entropy": 6.164347076416016,
      "epoch": 0.08945810271592225,
      "grad_norm": 1.0,
      "learning_rate": 0.00048158825330553505,
      "loss": 6.0871,
      "mean_token_accuracy": 0.1393938921391964,
      "num_tokens": 3131812.0,
      "step": 1390
    },
    {
      "entropy": 6.07136378288269,
      "epoch": 0.08977989445231047,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048111867723242763,
      "loss": 6.0846,
      "mean_token_accuracy": 0.14521935135126113,
      "num_tokens": 3142430.0,
      "step": 1395
    },
    {
      "entropy": 5.9969603538513185,
      "epoch": 0.09010168618869867,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004806434503116637,
      "loss": 5.9878,
      "mean_token_accuracy": 0.1448088675737381,
      "num_tokens": 3154552.0,
      "step": 1400
    },
    {
      "entropy": 6.107782697677612,
      "epoch": 0.09042347792508688,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004801625855718296,
      "loss": 6.0186,
      "mean_token_accuracy": 0.1484021157026291,
      "num_tokens": 3167117.0,
      "step": 1405
    },
    {
      "entropy": 6.196589136123658,
      "epoch": 0.09074526966147509,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047967609619607477,
      "loss": 6.176,
      "mean_token_accuracy": 0.14222314432263375,
      "num_tokens": 3177160.0,
      "step": 1410
    },
    {
      "entropy": 5.963523435592651,
      "epoch": 0.0910670613978633,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004791839955217513,
      "loss": 5.9977,
      "mean_token_accuracy": 0.15531311482191085,
      "num_tokens": 3188036.0,
      "step": 1415
    },
    {
      "entropy": 5.970466613769531,
      "epoch": 0.09138885313425152,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047868629704004786,
      "loss": 5.9436,
      "mean_token_accuracy": 0.15874468088150023,
      "num_tokens": 3199507.0,
      "step": 1420
    },
    {
      "entropy": 5.997041034698486,
      "epoch": 0.09171064487063972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047818301439561965,
      "loss": 5.8895,
      "mean_token_accuracy": 0.1560737043619156,
      "num_tokens": 3210316.0,
      "step": 1425
    },
    {
      "entropy": 6.003460884094238,
      "epoch": 0.09203243660702794,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047767416138621454,
      "loss": 6.058,
      "mean_token_accuracy": 0.14701004475355148,
      "num_tokens": 3221794.0,
      "step": 1430
    },
    {
      "entropy": 6.125612449645996,
      "epoch": 0.09235422834341614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000477159751962295,
      "loss": 6.0359,
      "mean_token_accuracy": 0.1510167494416237,
      "num_tokens": 3232950.0,
      "step": 1435
    },
    {
      "entropy": 6.005368185043335,
      "epoch": 0.09267602007980436,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047663980022665507,
      "loss": 5.9723,
      "mean_token_accuracy": 0.15661734193563462,
      "num_tokens": 3244012.0,
      "step": 1440
    },
    {
      "entropy": 5.998808765411377,
      "epoch": 0.09299781181619256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047611432043403437,
      "loss": 6.0681,
      "mean_token_accuracy": 0.1516854725778103,
      "num_tokens": 3255599.0,
      "step": 1445
    },
    {
      "entropy": 6.070061779022216,
      "epoch": 0.09331960355258077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004755833269907267,
      "loss": 5.9832,
      "mean_token_accuracy": 0.1570393592119217,
      "num_tokens": 3267060.0,
      "step": 1450
    },
    {
      "entropy": 6.018462991714477,
      "epoch": 0.09364139528896898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004750468344541857,
      "loss": 6.0012,
      "mean_token_accuracy": 0.15066305920481682,
      "num_tokens": 3278252.0,
      "step": 1455
    },
    {
      "entropy": 5.992702293395996,
      "epoch": 0.09396318702535719,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00047450485753262525,
      "loss": 5.7633,
      "mean_token_accuracy": 0.15733575001358985,
      "num_tokens": 3288551.0,
      "step": 1460
    },
    {
      "entropy": 5.891961193084716,
      "epoch": 0.0942849787617454,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047395741108461633,
      "loss": 6.0299,
      "mean_token_accuracy": 0.14577169418334962,
      "num_tokens": 3300424.0,
      "step": 1465
    },
    {
      "entropy": 6.028947257995606,
      "epoch": 0.09460677049813361,
      "grad_norm": 0.953125,
      "learning_rate": 0.00047340451011867985,
      "loss": 5.9416,
      "mean_token_accuracy": 0.1503148078918457,
      "num_tokens": 3312224.0,
      "step": 1470
    },
    {
      "entropy": 5.910190534591675,
      "epoch": 0.09492856223452181,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00047284616979287515,
      "loss": 5.9129,
      "mean_token_accuracy": 0.1531186506152153,
      "num_tokens": 3323262.0,
      "step": 1475
    },
    {
      "entropy": 5.984435129165649,
      "epoch": 0.09525035397091003,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00047228240541438433,
      "loss": 5.9634,
      "mean_token_accuracy": 0.15215728729963302,
      "num_tokens": 3334016.0,
      "step": 1480
    },
    {
      "entropy": 6.027677917480469,
      "epoch": 0.09557214570729823,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047171323243909257,
      "loss": 6.0225,
      "mean_token_accuracy": 0.15151964053511618,
      "num_tokens": 3344026.0,
      "step": 1485
    },
    {
      "entropy": 6.137910842895508,
      "epoch": 0.09589393744368645,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00047113866647116457,
      "loss": 6.1397,
      "mean_token_accuracy": 0.14101183116436006,
      "num_tokens": 3355453.0,
      "step": 1490
    },
    {
      "entropy": 5.969770812988282,
      "epoch": 0.09621572918007465,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004705587232626164,
      "loss": 5.9421,
      "mean_token_accuracy": 0.15476928949356078,
      "num_tokens": 3366719.0,
      "step": 1495
    },
    {
      "entropy": 5.972850036621094,
      "epoch": 0.09653752091646287,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046997341871288424,
      "loss": 5.9765,
      "mean_token_accuracy": 0.1517869286239147,
      "num_tokens": 3377689.0,
      "step": 1500
    },
    {
      "entropy": 6.007156610488892,
      "epoch": 0.09685931265285108,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004693827688683879,
      "loss": 5.9625,
      "mean_token_accuracy": 0.15451329201459885,
      "num_tokens": 3387506.0,
      "step": 1505
    },
    {
      "entropy": 5.956726312637329,
      "epoch": 0.09718110438923928,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004687867899220914,
      "loss": 5.9864,
      "mean_token_accuracy": 0.1476556710898876,
      "num_tokens": 3400673.0,
      "step": 1510
    },
    {
      "entropy": 6.018198251724243,
      "epoch": 0.0975028961256275,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046818549821305846,
      "loss": 5.8886,
      "mean_token_accuracy": 0.1627383753657341,
      "num_tokens": 3410402.0,
      "step": 1515
    },
    {
      "entropy": 5.961309385299683,
      "epoch": 0.0978246878620157,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046757891022600494,
      "loss": 5.9154,
      "mean_token_accuracy": 0.157576847076416,
      "num_tokens": 3421752.0,
      "step": 1520
    },
    {
      "entropy": 6.011680936813354,
      "epoch": 0.09814647959840392,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004669670425908471,
      "loss": 6.0598,
      "mean_token_accuracy": 0.14632627367973328,
      "num_tokens": 3433293.0,
      "step": 1525
    },
    {
      "entropy": 5.9564330101013185,
      "epoch": 0.09846827133479212,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004663499120822451,
      "loss": 5.8276,
      "mean_token_accuracy": 0.1553707927465439,
      "num_tokens": 3444768.0,
      "step": 1530
    },
    {
      "entropy": 5.846206903457642,
      "epoch": 0.09879006307118034,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004657275356191437,
      "loss": 5.9117,
      "mean_token_accuracy": 0.15863873660564423,
      "num_tokens": 3456174.0,
      "step": 1535
    },
    {
      "entropy": 5.99799599647522,
      "epoch": 0.09911185480756854,
      "grad_norm": 1.125,
      "learning_rate": 0.00046509993026430804,
      "loss": 5.8807,
      "mean_token_accuracy": 0.15284504294395446,
      "num_tokens": 3466512.0,
      "step": 1540
    },
    {
      "entropy": 5.948553133010864,
      "epoch": 0.09943364654395676,
      "grad_norm": 1.0,
      "learning_rate": 0.0004644671132238558,
      "loss": 5.9194,
      "mean_token_accuracy": 0.15121473073959352,
      "num_tokens": 3476458.0,
      "step": 1545
    },
    {
      "entropy": 5.954727649688721,
      "epoch": 0.09975543828034496,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046382910184678585,
      "loss": 5.9136,
      "mean_token_accuracy": 0.15346899926662444,
      "num_tokens": 3487844.0,
      "step": 1550
    },
    {
      "entropy": 5.8821799755096436,
      "epoch": 0.10007723001673317,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004631859136245025,
      "loss": 5.9371,
      "mean_token_accuracy": 0.15183866620063782,
      "num_tokens": 3498516.0,
      "step": 1555
    },
    {
      "entropy": 5.948780822753906,
      "epoch": 0.10039902175312138,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004625375661903357,
      "loss": 5.8911,
      "mean_token_accuracy": 0.15053738802671432,
      "num_tokens": 3510341.0,
      "step": 1560
    },
    {
      "entropy": 6.01388030052185,
      "epoch": 0.10072081348950959,
      "grad_norm": 1.125,
      "learning_rate": 0.00046188407731905787,
      "loss": 6.0,
      "mean_token_accuracy": 0.15072501227259635,
      "num_tokens": 3521209.0,
      "step": 1565
    },
    {
      "entropy": 5.85715069770813,
      "epoch": 0.1010426052258978,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00046122546492639643,
      "loss": 5.893,
      "mean_token_accuracy": 0.15645911991596223,
      "num_tokens": 3532217.0,
      "step": 1570
    },
    {
      "entropy": 5.904603004455566,
      "epoch": 0.10136439696228601,
      "grad_norm": 1.0,
      "learning_rate": 0.000460561747068543,
      "loss": 5.9774,
      "mean_token_accuracy": 0.15602791607379912,
      "num_tokens": 3543995.0,
      "step": 1575
    },
    {
      "entropy": 6.0624183177947994,
      "epoch": 0.10168618869867421,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004598929419416578,
      "loss": 5.9029,
      "mean_token_accuracy": 0.15403971821069717,
      "num_tokens": 3555652.0,
      "step": 1580
    },
    {
      "entropy": 5.792616319656372,
      "epoch": 0.10200798043506243,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045921906788137123,
      "loss": 5.8913,
      "mean_token_accuracy": 0.1603339657187462,
      "num_tokens": 3566492.0,
      "step": 1585
    },
    {
      "entropy": 6.011936378479004,
      "epoch": 0.10232977217145063,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045854014336228115,
      "loss": 5.8699,
      "mean_token_accuracy": 0.1546788766980171,
      "num_tokens": 3577935.0,
      "step": 1590
    },
    {
      "entropy": 5.8856230735778805,
      "epoch": 0.10265156390783885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045785618699744615,
      "loss": 5.9684,
      "mean_token_accuracy": 0.15323716551065444,
      "num_tokens": 3589301.0,
      "step": 1595
    },
    {
      "entropy": 5.999621391296387,
      "epoch": 0.10297335564422706,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045716721753787543,
      "loss": 5.9862,
      "mean_token_accuracy": 0.15104353725910186,
      "num_tokens": 3601626.0,
      "step": 1600
    },
    {
      "entropy": 5.85779504776001,
      "epoch": 0.10329514738061527,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004564732538720148,
      "loss": 5.8508,
      "mean_token_accuracy": 0.15395467430353166,
      "num_tokens": 3611436.0,
      "step": 1605
    },
    {
      "entropy": 5.937434530258178,
      "epoch": 0.10361693911700348,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045577431502522877,
      "loss": 5.9026,
      "mean_token_accuracy": 0.15675225034356116,
      "num_tokens": 3622973.0,
      "step": 1610
    },
    {
      "entropy": 5.909781885147095,
      "epoch": 0.10393873085339168,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004550704201592787,
      "loss": 5.9017,
      "mean_token_accuracy": 0.15919461250305175,
      "num_tokens": 3634623.0,
      "step": 1615
    },
    {
      "entropy": 6.012297630310059,
      "epoch": 0.1042605225897799,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004543615885717981,
      "loss": 6.0083,
      "mean_token_accuracy": 0.15407174974679946,
      "num_tokens": 3646664.0,
      "step": 1620
    },
    {
      "entropy": 5.945712995529175,
      "epoch": 0.1045823143261681,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045364783969576296,
      "loss": 5.9058,
      "mean_token_accuracy": 0.1542764276266098,
      "num_tokens": 3657890.0,
      "step": 1625
    },
    {
      "entropy": 5.919182443618775,
      "epoch": 0.10490410606255632,
      "grad_norm": 1.125,
      "learning_rate": 0.0004529291930989592,
      "loss": 5.9567,
      "mean_token_accuracy": 0.14882618859410285,
      "num_tokens": 3669696.0,
      "step": 1630
    },
    {
      "entropy": 5.951112079620361,
      "epoch": 0.10522589779894452,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004522056684834464,
      "loss": 5.9076,
      "mean_token_accuracy": 0.15821209847927092,
      "num_tokens": 3680100.0,
      "step": 1635
    },
    {
      "entropy": 5.9215904712677006,
      "epoch": 0.10554768953533274,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004514772856850173,
      "loss": 5.9342,
      "mean_token_accuracy": 0.1526174545288086,
      "num_tokens": 3691440.0,
      "step": 1640
    },
    {
      "entropy": 6.0002028465271,
      "epoch": 0.10586948127172094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004507440646726542,
      "loss": 5.9316,
      "mean_token_accuracy": 0.14958438128232956,
      "num_tokens": 3701883.0,
      "step": 1645
    },
    {
      "entropy": 5.941017007827758,
      "epoch": 0.10619127300810915,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004500060255479818,
      "loss": 5.9856,
      "mean_token_accuracy": 0.15235226228833199,
      "num_tokens": 3713617.0,
      "step": 1650
    },
    {
      "entropy": 6.003496313095093,
      "epoch": 0.10651306474449736,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004492631885447151,
      "loss": 5.8788,
      "mean_token_accuracy": 0.15904206037521362,
      "num_tokens": 3724337.0,
      "step": 1655
    },
    {
      "entropy": 6.044408845901489,
      "epoch": 0.10683485648088557,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00044851557402810616,
      "loss": 6.0336,
      "mean_token_accuracy": 0.1441546455025673,
      "num_tokens": 3736709.0,
      "step": 1660
    },
    {
      "entropy": 6.070526647567749,
      "epoch": 0.10715664821727378,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00044776320249438444,
      "loss": 5.9531,
      "mean_token_accuracy": 0.15391699373722076,
      "num_tokens": 3747474.0,
      "step": 1665
    },
    {
      "entropy": 5.804151821136474,
      "epoch": 0.10747843995366199,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044700609457019565,
      "loss": 5.8801,
      "mean_token_accuracy": 0.15678186863660812,
      "num_tokens": 3760883.0,
      "step": 1670
    },
    {
      "entropy": 5.909248018264771,
      "epoch": 0.1078002316900502,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004462442710120359,
      "loss": 5.9592,
      "mean_token_accuracy": 0.15394372045993804,
      "num_tokens": 3772465.0,
      "step": 1675
    },
    {
      "entropy": 5.912482404708863,
      "epoch": 0.10812202342643841,
      "grad_norm": 0.96875,
      "learning_rate": 0.000445477752705683,
      "loss": 5.7492,
      "mean_token_accuracy": 0.16123215705156327,
      "num_tokens": 3782682.0,
      "step": 1680
    },
    {
      "entropy": 5.917457246780396,
      "epoch": 0.10844381516282661,
      "grad_norm": 0.9375,
      "learning_rate": 0.00044470656066562336,
      "loss": 5.9236,
      "mean_token_accuracy": 0.15440376996994018,
      "num_tokens": 3794918.0,
      "step": 1685
    },
    {
      "entropy": 5.912576055526733,
      "epoch": 0.10876560689921483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004439307160344765,
      "loss": 5.8839,
      "mean_token_accuracy": 0.15873944461345674,
      "num_tokens": 3806041.0,
      "step": 1690
    },
    {
      "entropy": 5.97290005683899,
      "epoch": 0.10908739863560304,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044315024008241473,
      "loss": 5.9917,
      "mean_token_accuracy": 0.1478461280465126,
      "num_tokens": 3816569.0,
      "step": 1695
    },
    {
      "entropy": 5.902381753921508,
      "epoch": 0.10940919037199125,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004423651542065806,
      "loss": 5.9301,
      "mean_token_accuracy": 0.1564306400716305,
      "num_tokens": 3828266.0,
      "step": 1700
    },
    {
      "entropy": 6.009257650375366,
      "epoch": 0.10973098210837946,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044157547993050006,
      "loss": 5.913,
      "mean_token_accuracy": 0.1588406190276146,
      "num_tokens": 3840099.0,
      "step": 1705
    },
    {
      "entropy": 5.864715099334717,
      "epoch": 0.11005277384476767,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00044078123890349227,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16254316568374633,
      "num_tokens": 3851774.0,
      "step": 1710
    },
    {
      "entropy": 5.899453449249267,
      "epoch": 0.11037456558115588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043998245290007606,
      "loss": 5.8625,
      "mean_token_accuracy": 0.1496282361447811,
      "num_tokens": 3863398.0,
      "step": 1715
    },
    {
      "entropy": 5.895662784576416,
      "epoch": 0.11069635731754408,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043917914381937323,
      "loss": 5.912,
      "mean_token_accuracy": 0.16176212728023528,
      "num_tokens": 3875168.0,
      "step": 1720
    },
    {
      "entropy": 5.9350645542144775,
      "epoch": 0.1110181490539323,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043837133368450815,
      "loss": 5.7814,
      "mean_token_accuracy": 0.16604321599006652,
      "num_tokens": 3886296.0,
      "step": 1725
    },
    {
      "entropy": 5.854771709442138,
      "epoch": 0.1113399407903205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004375590446420037,
      "loss": 5.9224,
      "mean_token_accuracy": 0.15965267568826674,
      "num_tokens": 3897896.0,
      "step": 1730
    },
    {
      "entropy": 6.008559322357177,
      "epoch": 0.11166173252670872,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004367422989611743,
      "loss": 5.9959,
      "mean_token_accuracy": 0.14704702496528627,
      "num_tokens": 3909335.0,
      "step": 1735
    },
    {
      "entropy": 5.960231637954712,
      "epoch": 0.11198352426309692,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004359211190335153,
      "loss": 5.9273,
      "mean_token_accuracy": 0.15591442286968232,
      "num_tokens": 3921230.0,
      "step": 1740
    },
    {
      "entropy": 5.870255851745606,
      "epoch": 0.11230531599948514,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00043509552737208923,
      "loss": 5.9079,
      "mean_token_accuracy": 0.15321207642555237,
      "num_tokens": 3932497.0,
      "step": 1745
    },
    {
      "entropy": 5.926720762252808,
      "epoch": 0.11262710773587334,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043426554661090853,
      "loss": 5.8566,
      "mean_token_accuracy": 0.1600069060921669,
      "num_tokens": 3945275.0,
      "step": 1750
    },
    {
      "entropy": 5.876068925857544,
      "epoch": 0.11294889947226155,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043343119950431516,
      "loss": 5.9136,
      "mean_token_accuracy": 0.15576501786708832,
      "num_tokens": 3956901.0,
      "step": 1755
    },
    {
      "entropy": 5.908150672912598,
      "epoch": 0.11327069120864976,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043259250892635644,
      "loss": 5.8642,
      "mean_token_accuracy": 0.16266916692256927,
      "num_tokens": 3967989.0,
      "step": 1760
    },
    {
      "entropy": 6.0435584545135494,
      "epoch": 0.11359248294503797,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004317494978701582,
      "loss": 5.9015,
      "mean_token_accuracy": 0.15826912820339203,
      "num_tokens": 3980079.0,
      "step": 1765
    },
    {
      "entropy": 5.878367137908936,
      "epoch": 0.11391427468142618,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004309021894472943,
      "loss": 5.8487,
      "mean_token_accuracy": 0.15181449800729752,
      "num_tokens": 3991485.0,
      "step": 1770
    },
    {
      "entropy": 5.942022466659546,
      "epoch": 0.11423606641781439,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004300506068871534,
      "loss": 5.9173,
      "mean_token_accuracy": 0.1530771479010582,
      "num_tokens": 4002632.0,
      "step": 1775
    },
    {
      "entropy": 5.915815925598144,
      "epoch": 0.1145578581542026,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042919477353630135,
      "loss": 5.889,
      "mean_token_accuracy": 0.15727661401033402,
      "num_tokens": 4013676.0,
      "step": 1780
    },
    {
      "entropy": 5.946446180343628,
      "epoch": 0.11487964989059081,
      "grad_norm": 1.3125,
      "learning_rate": 0.000428334712857842,
      "loss": 5.9189,
      "mean_token_accuracy": 0.1588667020201683,
      "num_tokens": 4025738.0,
      "step": 1785
    },
    {
      "entropy": 5.924416637420654,
      "epoch": 0.11520144162697903,
      "grad_norm": 1.125,
      "learning_rate": 0.00042747044843077304,
      "loss": 5.9106,
      "mean_token_accuracy": 0.16056165993213653,
      "num_tokens": 4037930.0,
      "step": 1790
    },
    {
      "entropy": 6.00054235458374,
      "epoch": 0.11552323336336723,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00042660200394934047,
      "loss": 6.0129,
      "mean_token_accuracy": 0.14755067974328995,
      "num_tokens": 4049732.0,
      "step": 1795
    },
    {
      "entropy": 5.885663604736328,
      "epoch": 0.11584502509975544,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042572940322238844,
      "loss": 5.8486,
      "mean_token_accuracy": 0.15762704685330392,
      "num_tokens": 4063370.0,
      "step": 1800
    },
    {
      "entropy": 5.926276350021363,
      "epoch": 0.11616681683614365,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00042485267017270664,
      "loss": 5.9645,
      "mean_token_accuracy": 0.1550281137228012,
      "num_tokens": 4075646.0,
      "step": 1805
    },
    {
      "entropy": 5.849887084960938,
      "epoch": 0.11648860857253186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004239718288363745,
      "loss": 5.7667,
      "mean_token_accuracy": 0.16589755862951278,
      "num_tokens": 4086088.0,
      "step": 1810
    },
    {
      "entropy": 5.852992105484009,
      "epoch": 0.11681040030892006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230869033621023,
      "loss": 5.8435,
      "mean_token_accuracy": 0.1609729766845703,
      "num_tokens": 4097735.0,
      "step": 1815
    },
    {
      "entropy": 5.979506397247315,
      "epoch": 0.11713219204530828,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004221979180105688,
      "loss": 5.8795,
      "mean_token_accuracy": 0.1602598875761032,
      "num_tokens": 4109687.0,
      "step": 1820
    },
    {
      "entropy": 5.909818267822265,
      "epoch": 0.11745398378169648,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042130489715375645,
      "loss": 5.8386,
      "mean_token_accuracy": 0.16082719191908837,
      "num_tokens": 4120275.0,
      "step": 1825
    },
    {
      "entropy": 5.771799325942993,
      "epoch": 0.1177757755180847,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042040786527428335,
      "loss": 5.7417,
      "mean_token_accuracy": 0.16540927290916443,
      "num_tokens": 4130680.0,
      "step": 1830
    },
    {
      "entropy": 5.872281885147094,
      "epoch": 0.1180975672544729,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004195068469647315,
      "loss": 5.91,
      "mean_token_accuracy": 0.15430775731801988,
      "num_tokens": 4142272.0,
      "step": 1835
    },
    {
      "entropy": 5.8957343101501465,
      "epoch": 0.11841935899086112,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041860186692697297,
      "loss": 5.8161,
      "mean_token_accuracy": 0.15842420160770415,
      "num_tokens": 4153619.0,
      "step": 1840
    },
    {
      "entropy": 5.818892955780029,
      "epoch": 0.11874115072724932,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041769294997149264,
      "loss": 5.7716,
      "mean_token_accuracy": 0.16130125522613525,
      "num_tokens": 4164026.0,
      "step": 1845
    },
    {
      "entropy": 5.903949880599976,
      "epoch": 0.11906294246363754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004167801210167081,
      "loss": 5.8816,
      "mean_token_accuracy": 0.15510665476322175,
      "num_tokens": 4176772.0,
      "step": 1850
    },
    {
      "entropy": 5.934047603607178,
      "epoch": 0.11938473420002574,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004158634050882861,
      "loss": 5.7464,
      "mean_token_accuracy": 0.17128876298666001,
      "num_tokens": 4188268.0,
      "step": 1855
    },
    {
      "entropy": 5.727901649475098,
      "epoch": 0.11970652593641395,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004149428273184569,
      "loss": 5.8305,
      "mean_token_accuracy": 0.16112859100103377,
      "num_tokens": 4199558.0,
      "step": 1860
    },
    {
      "entropy": 5.964790534973145,
      "epoch": 0.12002831767280216,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004140184129453253,
      "loss": 5.9023,
      "mean_token_accuracy": 0.15668266266584396,
      "num_tokens": 4212412.0,
      "step": 1865
    },
    {
      "entropy": 5.853475856781006,
      "epoch": 0.12035010940919037,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000413090187312178,
      "loss": 5.6953,
      "mean_token_accuracy": 0.16073490902781487,
      "num_tokens": 4222361.0,
      "step": 1870
    },
    {
      "entropy": 5.685002326965332,
      "epoch": 0.12067190114557858,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004121581758667898,
      "loss": 5.6881,
      "mean_token_accuracy": 0.16696619540452956,
      "num_tokens": 4233396.0,
      "step": 1875
    },
    {
      "entropy": 5.866187334060669,
      "epoch": 0.12099369288196679,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041122240416072533,
      "loss": 5.9006,
      "mean_token_accuracy": 0.15792281329631805,
      "num_tokens": 4244902.0,
      "step": 1880
    },
    {
      "entropy": 5.915771865844727,
      "epoch": 0.12131548461835501,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004102828978486385,
      "loss": 5.7624,
      "mean_token_accuracy": 0.17442939132452012,
      "num_tokens": 4256055.0,
      "step": 1885
    },
    {
      "entropy": 5.868275308609009,
      "epoch": 0.12163727635474321,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004093396826875695,
      "loss": 5.9111,
      "mean_token_accuracy": 0.15428673028945922,
      "num_tokens": 4267075.0,
      "step": 1890
    },
    {
      "entropy": 5.802133178710937,
      "epoch": 0.12195906809113143,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040839278453623837,
      "loss": 5.7418,
      "mean_token_accuracy": 0.1663813665509224,
      "num_tokens": 4277888.0,
      "step": 1895
    },
    {
      "entropy": 5.923374128341675,
      "epoch": 0.12228085982751963,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004074422293543363,
      "loss": 5.8357,
      "mean_token_accuracy": 0.15506165772676467,
      "num_tokens": 4288680.0,
      "step": 1900
    },
    {
      "entropy": 5.859901380538941,
      "epoch": 0.12260265156390784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004064880432018137,
      "loss": 5.8526,
      "mean_token_accuracy": 0.1631172552704811,
      "num_tokens": 4298820.0,
      "step": 1905
    },
    {
      "entropy": 5.857502079010009,
      "epoch": 0.12292444330029605,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040553025223816615,
      "loss": 5.7659,
      "mean_token_accuracy": 0.1576807200908661,
      "num_tokens": 4309365.0,
      "step": 1910
    },
    {
      "entropy": 5.877448987960816,
      "epoch": 0.12324623503668426,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040456888272171653,
      "loss": 5.8271,
      "mean_token_accuracy": 0.1554803043603897,
      "num_tokens": 4321799.0,
      "step": 1915
    },
    {
      "entropy": 5.8920072555542,
      "epoch": 0.12356802677307246,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040360396100889577,
      "loss": 5.865,
      "mean_token_accuracy": 0.158425410091877,
      "num_tokens": 4332765.0,
      "step": 1920
    },
    {
      "entropy": 5.952794742584229,
      "epoch": 0.12388981850946068,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004026355135535202,
      "loss": 5.8025,
      "mean_token_accuracy": 0.15583222210407258,
      "num_tokens": 4344280.0,
      "step": 1925
    },
    {
      "entropy": 5.843797206878662,
      "epoch": 0.12421161024584888,
      "grad_norm": 1.0,
      "learning_rate": 0.000401663566906066,
      "loss": 5.7491,
      "mean_token_accuracy": 0.16600394546985625,
      "num_tokens": 4355177.0,
      "step": 1930
    },
    {
      "entropy": 5.863084220886231,
      "epoch": 0.1245334019822371,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040068814771294134,
      "loss": 5.7915,
      "mean_token_accuracy": 0.16675261706113814,
      "num_tokens": 4366898.0,
      "step": 1935
    },
    {
      "entropy": 5.891567659378052,
      "epoch": 0.1248551937186253,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003997092827157562,
      "loss": 5.8661,
      "mean_token_accuracy": 0.1553776204586029,
      "num_tokens": 4377869.0,
      "step": 1940
    },
    {
      "entropy": 5.845063591003418,
      "epoch": 0.12517698545501352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000398726998750589,
      "loss": 5.7076,
      "mean_token_accuracy": 0.16778974533081054,
      "num_tokens": 4388499.0,
      "step": 1945
    },
    {
      "entropy": 5.754357147216797,
      "epoch": 0.12549877719140173,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039774132274725076,
      "loss": 5.7962,
      "mean_token_accuracy": 0.15787893384695054,
      "num_tokens": 4400576.0,
      "step": 1950
    },
    {
      "entropy": 5.8626384258270265,
      "epoch": 0.12582056892778992,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039675228172854707,
      "loss": 5.783,
      "mean_token_accuracy": 0.15784820839762687,
      "num_tokens": 4411972.0,
      "step": 1955
    },
    {
      "entropy": 5.849202680587768,
      "epoch": 0.12614236066417814,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003957599028095371,
      "loss": 5.8243,
      "mean_token_accuracy": 0.1586673527956009,
      "num_tokens": 4422196.0,
      "step": 1960
    },
    {
      "entropy": 5.818286466598511,
      "epoch": 0.12646415240056635,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039476421319679017,
      "loss": 5.8093,
      "mean_token_accuracy": 0.16714178770780563,
      "num_tokens": 4433423.0,
      "step": 1965
    },
    {
      "entropy": 5.92006549835205,
      "epoch": 0.12678594413695457,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039376524018764,
      "loss": 5.816,
      "mean_token_accuracy": 0.16039222329854966,
      "num_tokens": 4444683.0,
      "step": 1970
    },
    {
      "entropy": 5.897944211959839,
      "epoch": 0.12710773587334276,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039276301116943616,
      "loss": 5.8231,
      "mean_token_accuracy": 0.16646796762943267,
      "num_tokens": 4455532.0,
      "step": 1975
    },
    {
      "entropy": 5.73743314743042,
      "epoch": 0.12742952760973097,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003917575536187936,
      "loss": 5.7323,
      "mean_token_accuracy": 0.17379816323518754,
      "num_tokens": 4466177.0,
      "step": 1980
    },
    {
      "entropy": 5.889675617218018,
      "epoch": 0.1277513193461192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039074889510083894,
      "loss": 5.9075,
      "mean_token_accuracy": 0.1522969052195549,
      "num_tokens": 4478623.0,
      "step": 1985
    },
    {
      "entropy": 5.958529806137085,
      "epoch": 0.1280731110825074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038973706326845495,
      "loss": 5.83,
      "mean_token_accuracy": 0.16366453915834428,
      "num_tokens": 4489080.0,
      "step": 1990
    },
    {
      "entropy": 5.827965545654297,
      "epoch": 0.12839490281889562,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003887220858615225,
      "loss": 5.8017,
      "mean_token_accuracy": 0.16294108033180238,
      "num_tokens": 4499648.0,
      "step": 1995
    },
    {
      "entropy": 5.852404594421387,
      "epoch": 0.1287166945552838,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003877039907061597,
      "loss": 5.816,
      "mean_token_accuracy": 0.1611374318599701,
      "num_tokens": 4510937.0,
      "step": 2000
    },
    {
      "entropy": 5.825767660140992,
      "epoch": 0.12903848629167203,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003866828057139598,
      "loss": 5.7396,
      "mean_token_accuracy": 0.16263994723558425,
      "num_tokens": 4522479.0,
      "step": 2005
    },
    {
      "entropy": 5.769980478286743,
      "epoch": 0.12936027802806024,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038565855888122503,
      "loss": 5.7414,
      "mean_token_accuracy": 0.16560350060462953,
      "num_tokens": 4533167.0,
      "step": 2010
    },
    {
      "entropy": 5.8879228115081785,
      "epoch": 0.12968206976444846,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00038463127828819975,
      "loss": 5.8311,
      "mean_token_accuracy": 0.15922270715236664,
      "num_tokens": 4544064.0,
      "step": 2015
    },
    {
      "entropy": 5.792943096160888,
      "epoch": 0.13000386150083665,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038360099209830043,
      "loss": 5.7325,
      "mean_token_accuracy": 0.1662492945790291,
      "num_tokens": 4555540.0,
      "step": 2020
    },
    {
      "entropy": 5.878731203079224,
      "epoch": 0.13032565323722486,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003825677285573433,
      "loss": 5.8076,
      "mean_token_accuracy": 0.15965683460235597,
      "num_tokens": 4568111.0,
      "step": 2025
    },
    {
      "entropy": 5.941011619567871,
      "epoch": 0.13064744497361308,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038153151599277027,
      "loss": 5.8995,
      "mean_token_accuracy": 0.15820185244083404,
      "num_tokens": 4580315.0,
      "step": 2030
    },
    {
      "entropy": 5.7433160781860355,
      "epoch": 0.1309692367100013,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003804923828128723,
      "loss": 5.7622,
      "mean_token_accuracy": 0.16896363347768784,
      "num_tokens": 4591623.0,
      "step": 2035
    },
    {
      "entropy": 5.844108963012696,
      "epoch": 0.13129102844638948,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003794503575060104,
      "loss": 5.8124,
      "mean_token_accuracy": 0.15706413984298706,
      "num_tokens": 4602929.0,
      "step": 2040
    },
    {
      "entropy": 5.938379287719727,
      "epoch": 0.1316128201827777,
      "grad_norm": 1.4921875,
      "learning_rate": 0.00037840546863983484,
      "loss": 5.8476,
      "mean_token_accuracy": 0.15860359370708466,
      "num_tokens": 4615073.0,
      "step": 2045
    },
    {
      "entropy": 5.824937200546264,
      "epoch": 0.13193461191916592,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003773577448605015,
      "loss": 5.7878,
      "mean_token_accuracy": 0.1596840962767601,
      "num_tokens": 4625249.0,
      "step": 2050
    },
    {
      "entropy": 5.87605094909668,
      "epoch": 0.13225640365555413,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003763072148918872,
      "loss": 5.7944,
      "mean_token_accuracy": 0.15860157012939452,
      "num_tokens": 4638544.0,
      "step": 2055
    },
    {
      "entropy": 5.7525852680206295,
      "epoch": 0.13257819539194232,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003752539075348017,
      "loss": 5.7285,
      "mean_token_accuracy": 0.1704973801970482,
      "num_tokens": 4649167.0,
      "step": 2060
    },
    {
      "entropy": 5.853913736343384,
      "epoch": 0.13289998712833054,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037419785166619817,
      "loss": 5.8019,
      "mean_token_accuracy": 0.158880615234375,
      "num_tokens": 4661637.0,
      "step": 2065
    },
    {
      "entropy": 5.9116246700286865,
      "epoch": 0.13322177886471875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003731390762383818,
      "loss": 5.9144,
      "mean_token_accuracy": 0.15258290395140647,
      "num_tokens": 4673723.0,
      "step": 2070
    },
    {
      "entropy": 5.849474906921387,
      "epoch": 0.13354357060110697,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003720776102782158,
      "loss": 5.7971,
      "mean_token_accuracy": 0.1537196971476078,
      "num_tokens": 4685735.0,
      "step": 2075
    },
    {
      "entropy": 5.750076198577881,
      "epoch": 0.1338653623374952,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037101348288632555,
      "loss": 5.6408,
      "mean_token_accuracy": 0.1739220142364502,
      "num_tokens": 4695664.0,
      "step": 2080
    },
    {
      "entropy": 5.897701787948608,
      "epoch": 0.13418715407388337,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003699467232363012,
      "loss": 5.8256,
      "mean_token_accuracy": 0.15790853947401046,
      "num_tokens": 4708242.0,
      "step": 2085
    },
    {
      "entropy": 5.7580366134643555,
      "epoch": 0.1345089458102716,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003688773605738973,
      "loss": 5.7587,
      "mean_token_accuracy": 0.1662400022149086,
      "num_tokens": 4719361.0,
      "step": 2090
    },
    {
      "entropy": 5.912790679931641,
      "epoch": 0.1348307375466598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036780542421623134,
      "loss": 5.8442,
      "mean_token_accuracy": 0.16789428889751434,
      "num_tokens": 4730672.0,
      "step": 2095
    },
    {
      "entropy": 5.820644998550415,
      "epoch": 0.13515252928304802,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003667309435509802,
      "loss": 5.7108,
      "mean_token_accuracy": 0.16895635426044464,
      "num_tokens": 4742066.0,
      "step": 2100
    },
    {
      "entropy": 5.826292037963867,
      "epoch": 0.1354743210194362,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003656539480355741,
      "loss": 5.8127,
      "mean_token_accuracy": 0.15991458892822266,
      "num_tokens": 4752693.0,
      "step": 2105
    },
    {
      "entropy": 5.785766649246216,
      "epoch": 0.13579611275582443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003645744671963891,
      "loss": 5.6275,
      "mean_token_accuracy": 0.16639572978019715,
      "num_tokens": 4763301.0,
      "step": 2110
    },
    {
      "entropy": 5.769074964523315,
      "epoch": 0.13611790449221264,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003634925306279376,
      "loss": 5.7386,
      "mean_token_accuracy": 0.16180252134799958,
      "num_tokens": 4774845.0,
      "step": 2115
    },
    {
      "entropy": 5.814271116256714,
      "epoch": 0.13643969622860086,
      "grad_norm": 1.0,
      "learning_rate": 0.0003624081679920574,
      "loss": 5.7176,
      "mean_token_accuracy": 0.16672129184007645,
      "num_tokens": 4786435.0,
      "step": 2120
    },
    {
      "entropy": 5.83293342590332,
      "epoch": 0.13676148796498905,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003613214090170977,
      "loss": 5.7464,
      "mean_token_accuracy": 0.1640954092144966,
      "num_tokens": 4797773.0,
      "step": 2125
    },
    {
      "entropy": 5.851455116271973,
      "epoch": 0.13708327970137726,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003602322834971048,
      "loss": 5.7814,
      "mean_token_accuracy": 0.15917347148060798,
      "num_tokens": 4808426.0,
      "step": 2130
    },
    {
      "entropy": 5.823632717132568,
      "epoch": 0.13740507143776548,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003591408212910051,
      "loss": 5.6825,
      "mean_token_accuracy": 0.17304720878601074,
      "num_tokens": 4818849.0,
      "step": 2135
    },
    {
      "entropy": 5.740329837799072,
      "epoch": 0.1377268631741537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003580470523217863,
      "loss": 5.708,
      "mean_token_accuracy": 0.16744430959224701,
      "num_tokens": 4829951.0,
      "step": 2140
    },
    {
      "entropy": 5.8014449119567875,
      "epoch": 0.13804865491054188,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003569510065756771,
      "loss": 5.7068,
      "mean_token_accuracy": 0.16946078836917877,
      "num_tokens": 4840661.0,
      "step": 2145
    },
    {
      "entropy": 5.749332523345947,
      "epoch": 0.1383704466469301,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003558527141013254,
      "loss": 5.7067,
      "mean_token_accuracy": 0.1629394695162773,
      "num_tokens": 4850863.0,
      "step": 2150
    },
    {
      "entropy": 5.787845277786255,
      "epoch": 0.13869223838331832,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003547522050089742,
      "loss": 5.7352,
      "mean_token_accuracy": 0.16636962741613387,
      "num_tokens": 4862527.0,
      "step": 2155
    },
    {
      "entropy": 5.874751996994019,
      "epoch": 0.13901403011970653,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00035364950946963606,
      "loss": 5.7838,
      "mean_token_accuracy": 0.1608475163578987,
      "num_tokens": 4873645.0,
      "step": 2160
    },
    {
      "entropy": 5.749224758148193,
      "epoch": 0.13933582185609475,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003525446577142663,
      "loss": 5.6549,
      "mean_token_accuracy": 0.16208767145872116,
      "num_tokens": 4886481.0,
      "step": 2165
    },
    {
      "entropy": 5.792453145980835,
      "epoch": 0.13965761359248294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035143768003293395,
      "loss": 5.7877,
      "mean_token_accuracy": 0.1617497146129608,
      "num_tokens": 4898122.0,
      "step": 2170
    },
    {
      "entropy": 5.804708766937256,
      "epoch": 0.13997940532887115,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003503286067739913,
      "loss": 5.6727,
      "mean_token_accuracy": 0.16071560382843017,
      "num_tokens": 4909313.0,
      "step": 2175
    },
    {
      "entropy": 5.7051926612854,
      "epoch": 0.14030119706525937,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034921746834324193,
      "loss": 5.6663,
      "mean_token_accuracy": 0.17719542980194092,
      "num_tokens": 4921582.0,
      "step": 2180
    },
    {
      "entropy": 5.740246820449829,
      "epoch": 0.14062298880164759,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003481042952031072,
      "loss": 5.5813,
      "mean_token_accuracy": 0.17565540969371796,
      "num_tokens": 4933138.0,
      "step": 2185
    },
    {
      "entropy": 5.707650947570801,
      "epoch": 0.14094478053803577,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003469891178717911,
      "loss": 5.6352,
      "mean_token_accuracy": 0.17529215514659882,
      "num_tokens": 4944348.0,
      "step": 2190
    },
    {
      "entropy": 5.759362363815308,
      "epoch": 0.141266572274424,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003458719669224436,
      "loss": 5.666,
      "mean_token_accuracy": 0.17190855294466018,
      "num_tokens": 4954958.0,
      "step": 2195
    },
    {
      "entropy": 5.798312330245972,
      "epoch": 0.1415883640108122,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003447528729823221,
      "loss": 5.7003,
      "mean_token_accuracy": 0.16672273576259614,
      "num_tokens": 4965579.0,
      "step": 2200
    },
    {
      "entropy": 5.63726487159729,
      "epoch": 0.14191015574720042,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003436318667319525,
      "loss": 5.6563,
      "mean_token_accuracy": 0.1711287721991539,
      "num_tokens": 4975561.0,
      "step": 2205
    },
    {
      "entropy": 5.819374322891235,
      "epoch": 0.1422319474835886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034250897890428716,
      "loss": 5.7074,
      "mean_token_accuracy": 0.17213211506605147,
      "num_tokens": 4986676.0,
      "step": 2210
    },
    {
      "entropy": 5.728911304473877,
      "epoch": 0.14255373921997683,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003413842402838633,
      "loss": 5.6928,
      "mean_token_accuracy": 0.17019706964492798,
      "num_tokens": 4996943.0,
      "step": 2215
    },
    {
      "entropy": 5.69297080039978,
      "epoch": 0.14287553095636504,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00034025768170595834,
      "loss": 5.5753,
      "mean_token_accuracy": 0.17739512175321578,
      "num_tokens": 5009171.0,
      "step": 2220
    },
    {
      "entropy": 5.77791543006897,
      "epoch": 0.14319732269275326,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003391293340557446,
      "loss": 5.7586,
      "mean_token_accuracy": 0.1646817520260811,
      "num_tokens": 5020130.0,
      "step": 2225
    },
    {
      "entropy": 5.796299028396606,
      "epoch": 0.14351911442914145,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003379992282674431,
      "loss": 5.6274,
      "mean_token_accuracy": 0.17734064161777496,
      "num_tokens": 5031603.0,
      "step": 2230
    },
    {
      "entropy": 5.6813554763793945,
      "epoch": 0.14384090616552966,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003368673953234749,
      "loss": 5.75,
      "mean_token_accuracy": 0.15818410217761994,
      "num_tokens": 5043694.0,
      "step": 2235
    },
    {
      "entropy": 5.779007339477539,
      "epoch": 0.14416269790191788,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033573386625361176,
      "loss": 5.669,
      "mean_token_accuracy": 0.17275520116090776,
      "num_tokens": 5054802.0,
      "step": 2240
    },
    {
      "entropy": 5.767786312103271,
      "epoch": 0.1444844896383061,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00033459867213412567,
      "loss": 5.6723,
      "mean_token_accuracy": 0.16912684291601182,
      "num_tokens": 5065705.0,
      "step": 2245
    },
    {
      "entropy": 5.7344136238098145,
      "epoch": 0.14480628137469428,
      "grad_norm": 0.984375,
      "learning_rate": 0.000333461844086937,
      "loss": 5.7389,
      "mean_token_accuracy": 0.1672192856669426,
      "num_tokens": 5077687.0,
      "step": 2250
    },
    {
      "entropy": 5.785981845855713,
      "epoch": 0.1451280731110825,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033232341327876097,
      "loss": 5.5786,
      "mean_token_accuracy": 0.18196143656969072,
      "num_tokens": 5088839.0,
      "step": 2255
    },
    {
      "entropy": 5.74149260520935,
      "epoch": 0.14544986484747072,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003311834109202531,
      "loss": 5.7074,
      "mean_token_accuracy": 0.1680561050772667,
      "num_tokens": 5100614.0,
      "step": 2260
    },
    {
      "entropy": 5.772172975540161,
      "epoch": 0.14577165658385893,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00033004186826515416,
      "loss": 5.7358,
      "mean_token_accuracy": 0.16374193727970124,
      "num_tokens": 5111015.0,
      "step": 2265
    },
    {
      "entropy": 5.871637439727783,
      "epoch": 0.14609344832024715,
      "grad_norm": 1.0,
      "learning_rate": 0.0003288988166094324,
      "loss": 5.844,
      "mean_token_accuracy": 0.1605599269270897,
      "num_tokens": 5121979.0,
      "step": 2270
    },
    {
      "entropy": 5.816112041473389,
      "epoch": 0.14641524005663534,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032775428729042656,
      "loss": 5.6607,
      "mean_token_accuracy": 0.1708015352487564,
      "num_tokens": 5133129.0,
      "step": 2275
    },
    {
      "entropy": 5.639659690856933,
      "epoch": 0.14673703179302355,
      "grad_norm": 0.9375,
      "learning_rate": 0.000326608311685986,
      "loss": 5.6274,
      "mean_token_accuracy": 0.17164263874292374,
      "num_tokens": 5144146.0,
      "step": 2280
    },
    {
      "entropy": 5.859129095077515,
      "epoch": 0.14705882352941177,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003254609212136108,
      "loss": 5.6658,
      "mean_token_accuracy": 0.16665443032979965,
      "num_tokens": 5155841.0,
      "step": 2285
    },
    {
      "entropy": 5.71760630607605,
      "epoch": 0.14738061526579999,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032431214732959036,
      "loss": 5.6952,
      "mean_token_accuracy": 0.16795089840888977,
      "num_tokens": 5166890.0,
      "step": 2290
    },
    {
      "entropy": 5.748434686660767,
      "epoch": 0.14770240700218817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000323162021528141,
      "loss": 5.7687,
      "mean_token_accuracy": 0.16394128352403642,
      "num_tokens": 5178174.0,
      "step": 2295
    },
    {
      "entropy": 5.836808395385742,
      "epoch": 0.1480241987385764,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032201057534054264,
      "loss": 5.6624,
      "mean_token_accuracy": 0.172088360786438,
      "num_tokens": 5189132.0,
      "step": 2300
    },
    {
      "entropy": 5.792338418960571,
      "epoch": 0.1483459904749646,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032085784033427414,
      "loss": 5.7244,
      "mean_token_accuracy": 0.16786212623119354,
      "num_tokens": 5199934.0,
      "step": 2305
    },
    {
      "entropy": 5.741599273681641,
      "epoch": 0.14866778221135282,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003197038481121478,
      "loss": 5.7482,
      "mean_token_accuracy": 0.16209147423505782,
      "num_tokens": 5210381.0,
      "step": 2310
    },
    {
      "entropy": 5.768185663223266,
      "epoch": 0.148989573947741,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003185486303114436,
      "loss": 5.6521,
      "mean_token_accuracy": 0.1676488548517227,
      "num_tokens": 5221169.0,
      "step": 2315
    },
    {
      "entropy": 5.8220250606536865,
      "epoch": 0.14931136568412923,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003173922186030409,
      "loss": 5.6266,
      "mean_token_accuracy": 0.1705146238207817,
      "num_tokens": 5233527.0,
      "step": 2320
    },
    {
      "entropy": 5.801087808609009,
      "epoch": 0.14963315742051744,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000316234644690551,
      "loss": 5.6812,
      "mean_token_accuracy": 0.16586119830608367,
      "num_tokens": 5243884.0,
      "step": 2325
    },
    {
      "entropy": 5.66407790184021,
      "epoch": 0.14995494915690566,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003150759403094473,
      "loss": 5.6779,
      "mean_token_accuracy": 0.17341048419475555,
      "num_tokens": 5254829.0,
      "step": 2330
    },
    {
      "entropy": 5.806618738174438,
      "epoch": 0.15027674089329385,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031391613722619587,
      "loss": 5.7024,
      "mean_token_accuracy": 0.16236732006072999,
      "num_tokens": 5266783.0,
      "step": 2335
    },
    {
      "entropy": 5.856207323074341,
      "epoch": 0.15059853262968206,
      "grad_norm": 1.046875,
      "learning_rate": 0.000312755267237384,
      "loss": 5.6879,
      "mean_token_accuracy": 0.16239377856254578,
      "num_tokens": 5278574.0,
      "step": 2340
    },
    {
      "entropy": 5.742423248291016,
      "epoch": 0.15092032436607028,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003115933621688488,
      "loss": 5.6268,
      "mean_token_accuracy": 0.17340437918901444,
      "num_tokens": 5289743.0,
      "step": 2345
    },
    {
      "entropy": 5.740512990951538,
      "epoch": 0.1512421161024585,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031043045387480487,
      "loss": 5.7034,
      "mean_token_accuracy": 0.17053953558206558,
      "num_tokens": 5299568.0,
      "step": 2350
    },
    {
      "entropy": 5.7228960514068605,
      "epoch": 0.1515639078388467,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003092665742369703,
      "loss": 5.5644,
      "mean_token_accuracy": 0.17459591031074523,
      "num_tokens": 5311177.0,
      "step": 2355
    },
    {
      "entropy": 5.751312971115112,
      "epoch": 0.1518856995752349,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00030810175516369343,
      "loss": 5.7135,
      "mean_token_accuracy": 0.1629110351204872,
      "num_tokens": 5322670.0,
      "step": 2360
    },
    {
      "entropy": 5.7899586200714115,
      "epoch": 0.15220749131162312,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003069360285890775,
      "loss": 5.6872,
      "mean_token_accuracy": 0.16645462512969972,
      "num_tokens": 5334385.0,
      "step": 2365
    },
    {
      "entropy": 5.834907865524292,
      "epoch": 0.15252928304801133,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030576942647210547,
      "loss": 5.7175,
      "mean_token_accuracy": 0.16685152053833008,
      "num_tokens": 5345467.0,
      "step": 2370
    },
    {
      "entropy": 5.775135898590088,
      "epoch": 0.15285107478439955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030460198079576355,
      "loss": 5.7355,
      "mean_token_accuracy": 0.166574227809906,
      "num_tokens": 5357793.0,
      "step": 2375
    },
    {
      "entropy": 5.7258546352386475,
      "epoch": 0.15317286652078774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003034337235661648,
      "loss": 5.6335,
      "mean_token_accuracy": 0.16347366571426392,
      "num_tokens": 5369111.0,
      "step": 2380
    },
    {
      "entropy": 5.739697504043579,
      "epoch": 0.15349465825717595,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003022646868116714,
      "loss": 5.6448,
      "mean_token_accuracy": 0.1721239745616913,
      "num_tokens": 5380592.0,
      "step": 2385
    },
    {
      "entropy": 5.7774739265441895,
      "epoch": 0.15381644999356417,
      "grad_norm": 0.875,
      "learning_rate": 0.0003010949025820163,
      "loss": 5.6682,
      "mean_token_accuracy": 0.17445676624774933,
      "num_tokens": 5392624.0,
      "step": 2390
    },
    {
      "entropy": 5.6770154476165775,
      "epoch": 0.15413824172995239,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002999244029474252,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18305357098579406,
      "num_tokens": 5403846.0,
      "step": 2395
    },
    {
      "entropy": 5.752689981460572,
      "epoch": 0.15446003346634057,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029875321999773684,
      "loss": 5.7557,
      "mean_token_accuracy": 0.16041475385427476,
      "num_tokens": 5416447.0,
      "step": 2400
    },
    {
      "entropy": 5.824324226379394,
      "epoch": 0.1547818252027288,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029758138584152333,
      "loss": 5.6466,
      "mean_token_accuracy": 0.17074086517095566,
      "num_tokens": 5428128.0,
      "step": 2405
    },
    {
      "entropy": 5.715564489364624,
      "epoch": 0.155103616939117,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002964089326052102,
      "loss": 5.6248,
      "mean_token_accuracy": 0.1692623719573021,
      "num_tokens": 5440291.0,
      "step": 2410
    },
    {
      "entropy": 5.796578359603882,
      "epoch": 0.15542540867550522,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002952358924321949,
      "loss": 5.6709,
      "mean_token_accuracy": 0.16312159299850465,
      "num_tokens": 5451103.0,
      "step": 2415
    },
    {
      "entropy": 5.743859052658081,
      "epoch": 0.1557472004118934,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029406229748196657,
      "loss": 5.6411,
      "mean_token_accuracy": 0.16831670552492142,
      "num_tokens": 5462657.0,
      "step": 2420
    },
    {
      "entropy": 5.77202033996582,
      "epoch": 0.15606899214828163,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002928881799292235,
      "loss": 5.6011,
      "mean_token_accuracy": 0.1700947880744934,
      "num_tokens": 5474090.0,
      "step": 2425
    },
    {
      "entropy": 5.761477375030518,
      "epoch": 0.15639078388466984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029171357196299154,
      "loss": 5.6999,
      "mean_token_accuracy": 0.16784273386001586,
      "num_tokens": 5485476.0,
      "step": 2430
    },
    {
      "entropy": 5.825471639633179,
      "epoch": 0.15671257562105806,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002905385057857414,
      "loss": 5.6991,
      "mean_token_accuracy": 0.1641829714179039,
      "num_tokens": 5495917.0,
      "step": 2435
    },
    {
      "entropy": 5.815099668502808,
      "epoch": 0.15703436735744625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002893630136125058,
      "loss": 5.6471,
      "mean_token_accuracy": 0.16917823553085326,
      "num_tokens": 5507374.0,
      "step": 2440
    },
    {
      "entropy": 5.694417810440063,
      "epoch": 0.15735615909383446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002881871276699967,
      "loss": 5.5991,
      "mean_token_accuracy": 0.17832678854465484,
      "num_tokens": 5518030.0,
      "step": 2445
    },
    {
      "entropy": 5.649110889434814,
      "epoch": 0.15767795083022268,
      "grad_norm": 1.125,
      "learning_rate": 0.00028701088019572114,
      "loss": 5.5829,
      "mean_token_accuracy": 0.1709439381957054,
      "num_tokens": 5528271.0,
      "step": 2450
    },
    {
      "entropy": 5.757630252838135,
      "epoch": 0.1579997425666109,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002858343034370977,
      "loss": 5.6871,
      "mean_token_accuracy": 0.16742459684610367,
      "num_tokens": 5538937.0,
      "step": 2455
    },
    {
      "entropy": 5.9393689155578615,
      "epoch": 0.1583215343029991,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028465742965057267,
      "loss": 5.9055,
      "mean_token_accuracy": 0.1508650816977024,
      "num_tokens": 5549335.0,
      "step": 2460
    },
    {
      "entropy": 5.752944040298462,
      "epoch": 0.1586433260393873,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028348029110073533,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17041779905557633,
      "num_tokens": 5560462.0,
      "step": 2465
    },
    {
      "entropy": 5.685638093948365,
      "epoch": 0.15896511777577552,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00028230292005943365,
      "loss": 5.612,
      "mean_token_accuracy": 0.17755805253982543,
      "num_tokens": 5570454.0,
      "step": 2470
    },
    {
      "entropy": 5.7484986782073975,
      "epoch": 0.15928690951216373,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028112534880488945,
      "loss": 5.5746,
      "mean_token_accuracy": 0.17453639209270477,
      "num_tokens": 5582446.0,
      "step": 2475
    },
    {
      "entropy": 5.715127992630005,
      "epoch": 0.15960870124855195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002799476096208137,
      "loss": 5.6722,
      "mean_token_accuracy": 0.16698170751333236,
      "num_tokens": 5594635.0,
      "step": 2480
    },
    {
      "entropy": 5.734315681457519,
      "epoch": 0.15993049298494014,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027876973479552087,
      "loss": 5.7102,
      "mean_token_accuracy": 0.17686639428138734,
      "num_tokens": 5606157.0,
      "step": 2485
    },
    {
      "entropy": 5.796161127090454,
      "epoch": 0.16025228472132835,
      "grad_norm": 1.0,
      "learning_rate": 0.00027759175662104424,
      "loss": 5.7053,
      "mean_token_accuracy": 0.1668169803917408,
      "num_tokens": 5617909.0,
      "step": 2490
    },
    {
      "entropy": 5.787411499023437,
      "epoch": 0.16057407645771657,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002764137073922508,
      "loss": 5.6795,
      "mean_token_accuracy": 0.16229669451713563,
      "num_tokens": 5630995.0,
      "step": 2495
    },
    {
      "entropy": 5.72302656173706,
      "epoch": 0.16089586819410479,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027523561940595505,
      "loss": 5.7186,
      "mean_token_accuracy": 0.16211511939764023,
      "num_tokens": 5642510.0,
      "step": 2500
    },
    {
      "entropy": 5.745364141464234,
      "epoch": 0.16121765993049297,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002740575249600342,
      "loss": 5.6271,
      "mean_token_accuracy": 0.16727463603019715,
      "num_tokens": 5654554.0,
      "step": 2505
    },
    {
      "entropy": 5.740041732788086,
      "epoch": 0.1615394516668812,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027287945635254263,
      "loss": 5.5445,
      "mean_token_accuracy": 0.17487518191337587,
      "num_tokens": 5666173.0,
      "step": 2510
    },
    {
      "entropy": 5.733009243011475,
      "epoch": 0.1618612434032694,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027170144588082635,
      "loss": 5.6234,
      "mean_token_accuracy": 0.17679100334644318,
      "num_tokens": 5677178.0,
      "step": 2515
    },
    {
      "entropy": 5.6678074359893795,
      "epoch": 0.16218303513965762,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027052352584063763,
      "loss": 5.6145,
      "mean_token_accuracy": 0.17861917316913606,
      "num_tokens": 5689195.0,
      "step": 2520
    },
    {
      "entropy": 5.956272315979004,
      "epoch": 0.1625048268760458,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026934572852524907,
      "loss": 5.7824,
      "mean_token_accuracy": 0.16314957439899444,
      "num_tokens": 5700771.0,
      "step": 2525
    },
    {
      "entropy": 5.7851142406463625,
      "epoch": 0.16282661861243403,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026816808622456937,
      "loss": 5.7097,
      "mean_token_accuracy": 0.16848561763763428,
      "num_tokens": 5711699.0,
      "step": 2530
    },
    {
      "entropy": 5.683857059478759,
      "epoch": 0.16314841034882224,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0002669906312242569,
      "loss": 5.6184,
      "mean_token_accuracy": 0.17141083776950836,
      "num_tokens": 5723286.0,
      "step": 2535
    },
    {
      "entropy": 5.7497316837310795,
      "epoch": 0.16347020208521046,
      "grad_norm": 0.921875,
      "learning_rate": 0.00026581339580483525,
      "loss": 5.5967,
      "mean_token_accuracy": 0.1723111167550087,
      "num_tokens": 5736061.0,
      "step": 2540
    },
    {
      "entropy": 5.745066452026367,
      "epoch": 0.16379199382159867,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002646364122408082,
      "loss": 5.6347,
      "mean_token_accuracy": 0.17187002152204514,
      "num_tokens": 5747620.0,
      "step": 2545
    },
    {
      "entropy": 5.676752519607544,
      "epoch": 0.16411378555798686,
      "grad_norm": 1.0,
      "learning_rate": 0.0002634597127997749,
      "loss": 5.5337,
      "mean_token_accuracy": 0.17709117084741594,
      "num_tokens": 5758695.0,
      "step": 2550
    },
    {
      "entropy": 5.665675210952759,
      "epoch": 0.16443557729437508,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002622833297415445,
      "loss": 5.5777,
      "mean_token_accuracy": 0.17751290649175644,
      "num_tokens": 5768572.0,
      "step": 2555
    },
    {
      "entropy": 5.689227628707886,
      "epoch": 0.1647573690307633,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002611072953172531,
      "loss": 5.5895,
      "mean_token_accuracy": 0.17748772203922272,
      "num_tokens": 5779798.0,
      "step": 2560
    },
    {
      "entropy": 5.776288938522339,
      "epoch": 0.1650791607671515,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025993164176847845,
      "loss": 5.6309,
      "mean_token_accuracy": 0.1727111294865608,
      "num_tokens": 5791005.0,
      "step": 2565
    },
    {
      "entropy": 5.751083278656006,
      "epoch": 0.1654009525035397,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002587564013263564,
      "loss": 5.6338,
      "mean_token_accuracy": 0.17171423882246017,
      "num_tokens": 5803648.0,
      "step": 2570
    },
    {
      "entropy": 5.747887659072876,
      "epoch": 0.16572274423992792,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002575816062106974,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17470741793513297,
      "num_tokens": 5814410.0,
      "step": 2575
    },
    {
      "entropy": 5.767975997924805,
      "epoch": 0.16604453597631613,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025640728862910293,
      "loss": 5.5942,
      "mean_token_accuracy": 0.18073396682739257,
      "num_tokens": 5826169.0,
      "step": 2580
    },
    {
      "entropy": 5.768696975708008,
      "epoch": 0.16636632771270435,
      "grad_norm": 0.921875,
      "learning_rate": 0.00025523348077608285,
      "loss": 5.6739,
      "mean_token_accuracy": 0.16068278551101683,
      "num_tokens": 5839379.0,
      "step": 2585
    },
    {
      "entropy": 5.659948492050171,
      "epoch": 0.16668811944909254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025406021483217225,
      "loss": 5.5936,
      "mean_token_accuracy": 0.174931663274765,
      "num_tokens": 5850734.0,
      "step": 2590
    },
    {
      "entropy": 5.683897399902344,
      "epoch": 0.16700991118548075,
      "grad_norm": 0.9375,
      "learning_rate": 0.00025288752296304963,
      "loss": 5.6192,
      "mean_token_accuracy": 0.17865339815616607,
      "num_tokens": 5862323.0,
      "step": 2595
    },
    {
      "entropy": 5.706236410140991,
      "epoch": 0.16733170292186897,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000251715437318655,
      "loss": 5.4733,
      "mean_token_accuracy": 0.18056690096855163,
      "num_tokens": 5873140.0,
      "step": 2600
    },
    {
      "entropy": 5.7186017513275145,
      "epoch": 0.16765349465825719,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002505439900323084,
      "loss": 5.5091,
      "mean_token_accuracy": 0.18042383342981339,
      "num_tokens": 5884282.0,
      "step": 2605
    },
    {
      "entropy": 5.701045083999634,
      "epoch": 0.16797528639464537,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00024937321321982894,
      "loss": 5.5835,
      "mean_token_accuracy": 0.1805478647351265,
      "num_tokens": 5896067.0,
      "step": 2610
    },
    {
      "entropy": 5.58365216255188,
      "epoch": 0.1682970781310336,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00024820313897865433,
      "loss": 5.5104,
      "mean_token_accuracy": 0.18009457737207413,
      "num_tokens": 5907051.0,
      "step": 2615
    },
    {
      "entropy": 5.648576402664185,
      "epoch": 0.1686188698674218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00024703379938696105,
      "loss": 5.574,
      "mean_token_accuracy": 0.17690564692020416,
      "num_tokens": 5917937.0,
      "step": 2620
    },
    {
      "entropy": 5.746658658981323,
      "epoch": 0.16894066160381002,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024586522650278447,
      "loss": 5.5995,
      "mean_token_accuracy": 0.17674737721681594,
      "num_tokens": 5929208.0,
      "step": 2625
    },
    {
      "entropy": 5.74280366897583,
      "epoch": 0.1692624533401982,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00024469745236314064,
      "loss": 5.6606,
      "mean_token_accuracy": 0.16710000336170197,
      "num_tokens": 5939824.0,
      "step": 2630
    },
    {
      "entropy": 5.809232091903686,
      "epoch": 0.16958424507658643,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00024353050898314767,
      "loss": 5.6659,
      "mean_token_accuracy": 0.16944222897291183,
      "num_tokens": 5952022.0,
      "step": 2635
    },
    {
      "entropy": 5.7748181343078615,
      "epoch": 0.16990603681297464,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00024236442835514743,
      "loss": 5.6841,
      "mean_token_accuracy": 0.1673639178276062,
      "num_tokens": 5964102.0,
      "step": 2640
    },
    {
      "entropy": 5.736437654495239,
      "epoch": 0.17022782854936286,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00024119924244782965,
      "loss": 5.6561,
      "mean_token_accuracy": 0.16534417420625686,
      "num_tokens": 5976174.0,
      "step": 2645
    },
    {
      "entropy": 5.759584856033325,
      "epoch": 0.17054962028575107,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00024003498320535462,
      "loss": 5.581,
      "mean_token_accuracy": 0.17213369607925416,
      "num_tokens": 5987340.0,
      "step": 2650
    },
    {
      "entropy": 5.683378028869629,
      "epoch": 0.17087141202213926,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00023887168254647727,
      "loss": 5.5712,
      "mean_token_accuracy": 0.17500948309898376,
      "num_tokens": 5998398.0,
      "step": 2655
    },
    {
      "entropy": 5.6339733600616455,
      "epoch": 0.17119320375852748,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023770937236367308,
      "loss": 5.4887,
      "mean_token_accuracy": 0.17725965976715088,
      "num_tokens": 6009310.0,
      "step": 2660
    },
    {
      "entropy": 5.717236757278442,
      "epoch": 0.1715149954949157,
      "grad_norm": 0.96875,
      "learning_rate": 0.00023654808452226278,
      "loss": 5.6107,
      "mean_token_accuracy": 0.17078125476837158,
      "num_tokens": 6021035.0,
      "step": 2665
    },
    {
      "entropy": 5.769252967834473,
      "epoch": 0.1718367872313039,
      "grad_norm": 1.03125,
      "learning_rate": 0.00023538785085953912,
      "loss": 5.5421,
      "mean_token_accuracy": 0.1711832508444786,
      "num_tokens": 6032849.0,
      "step": 2670
    },
    {
      "entropy": 5.745126247406006,
      "epoch": 0.1721585789676921,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00023422870318389404,
      "loss": 5.6687,
      "mean_token_accuracy": 0.17119213938713074,
      "num_tokens": 6043867.0,
      "step": 2675
    },
    {
      "entropy": 5.7412035942077635,
      "epoch": 0.17248037070408032,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002330706732739468,
      "loss": 5.6647,
      "mean_token_accuracy": 0.16564337611198426,
      "num_tokens": 6055633.0,
      "step": 2680
    },
    {
      "entropy": 5.678151798248291,
      "epoch": 0.17280216244046853,
      "grad_norm": 0.984375,
      "learning_rate": 0.00023191379287767211,
      "loss": 5.5887,
      "mean_token_accuracy": 0.17996419221162796,
      "num_tokens": 6066986.0,
      "step": 2685
    },
    {
      "entropy": 5.719368124008179,
      "epoch": 0.17312395417685675,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002307580937115305,
      "loss": 5.6628,
      "mean_token_accuracy": 0.17769541144371032,
      "num_tokens": 6078897.0,
      "step": 2690
    },
    {
      "entropy": 5.776931953430176,
      "epoch": 0.17344574591324494,
      "grad_norm": 0.984375,
      "learning_rate": 0.00022960360745959846,
      "loss": 5.566,
      "mean_token_accuracy": 0.17624957114458084,
      "num_tokens": 6090034.0,
      "step": 2695
    },
    {
      "entropy": 5.754274797439575,
      "epoch": 0.17376753764963315,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00022845036577269972,
      "loss": 5.6075,
      "mean_token_accuracy": 0.17573596388101578,
      "num_tokens": 6101296.0,
      "step": 2700
    },
    {
      "entropy": 5.703330135345459,
      "epoch": 0.17408932938602137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00022729840026753777,
      "loss": 5.6291,
      "mean_token_accuracy": 0.16433238834142685,
      "num_tokens": 6113823.0,
      "step": 2705
    },
    {
      "entropy": 5.699364376068115,
      "epoch": 0.17441112112240958,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002261477425258287,
      "loss": 5.5487,
      "mean_token_accuracy": 0.17384159564971924,
      "num_tokens": 6125121.0,
      "step": 2710
    },
    {
      "entropy": 5.8121678829193115,
      "epoch": 0.17473291285879777,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002249984240934358,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16082816421985627,
      "num_tokens": 6136594.0,
      "step": 2715
    },
    {
      "entropy": 5.736780261993408,
      "epoch": 0.175054704595186,
      "grad_norm": 1.046875,
      "learning_rate": 0.00022385047647950464,
      "loss": 5.5377,
      "mean_token_accuracy": 0.1771623447537422,
      "num_tokens": 6146945.0,
      "step": 2720
    },
    {
      "entropy": 5.736144351959228,
      "epoch": 0.1753764963315742,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002227039311555986,
      "loss": 5.6493,
      "mean_token_accuracy": 0.16945586502552032,
      "num_tokens": 6158521.0,
      "step": 2725
    },
    {
      "entropy": 5.757169485092163,
      "epoch": 0.17569828806796242,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002215588195548372,
      "loss": 5.6459,
      "mean_token_accuracy": 0.17539671659469605,
      "num_tokens": 6169240.0,
      "step": 2730
    },
    {
      "entropy": 5.6294848918914795,
      "epoch": 0.17602007980435064,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00022041517307103337,
      "loss": 5.4641,
      "mean_token_accuracy": 0.18372700363397598,
      "num_tokens": 6180686.0,
      "step": 2735
    },
    {
      "entropy": 5.715571403503418,
      "epoch": 0.17634187154073883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002192730230578331,
      "loss": 5.5729,
      "mean_token_accuracy": 0.1786137655377388,
      "num_tokens": 6191065.0,
      "step": 2740
    },
    {
      "entropy": 5.79557056427002,
      "epoch": 0.17666366327712704,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002181324008278559,
      "loss": 5.6518,
      "mean_token_accuracy": 0.17339885234832764,
      "num_tokens": 6202878.0,
      "step": 2745
    },
    {
      "entropy": 5.693873023986816,
      "epoch": 0.17698545501351526,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00021699333765183655,
      "loss": 5.6046,
      "mean_token_accuracy": 0.17037773877382278,
      "num_tokens": 6213622.0,
      "step": 2750
    },
    {
      "entropy": 5.622701358795166,
      "epoch": 0.17730724674990347,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002158558647577673,
      "loss": 5.5581,
      "mean_token_accuracy": 0.17680486291646957,
      "num_tokens": 6224548.0,
      "step": 2755
    },
    {
      "entropy": 5.705058670043945,
      "epoch": 0.17762903848629166,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00021472001333004215,
      "loss": 5.517,
      "mean_token_accuracy": 0.1789616346359253,
      "num_tokens": 6235921.0,
      "step": 2760
    },
    {
      "entropy": 5.678872680664062,
      "epoch": 0.17795083022267988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00021358581450860186,
      "loss": 5.4965,
      "mean_token_accuracy": 0.18053624480962754,
      "num_tokens": 6246970.0,
      "step": 2765
    },
    {
      "entropy": 5.640407514572144,
      "epoch": 0.1782726219590681,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0002124532993880799,
      "loss": 5.5643,
      "mean_token_accuracy": 0.17441155016422272,
      "num_tokens": 6258524.0,
      "step": 2770
    },
    {
      "entropy": 5.686666011810303,
      "epoch": 0.1785944136954563,
      "grad_norm": 1.046875,
      "learning_rate": 0.00021132249901695044,
      "loss": 5.445,
      "mean_token_accuracy": 0.192951437830925,
      "num_tokens": 6269292.0,
      "step": 2775
    },
    {
      "entropy": 5.694063806533814,
      "epoch": 0.1789162054318445,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00021019344439667705,
      "loss": 5.6199,
      "mean_token_accuracy": 0.17343329042196273,
      "num_tokens": 6281055.0,
      "step": 2780
    },
    {
      "entropy": 5.727927398681641,
      "epoch": 0.17923799716823272,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00020906616648086213,
      "loss": 5.5443,
      "mean_token_accuracy": 0.17099367082118988,
      "num_tokens": 6293609.0,
      "step": 2785
    },
    {
      "entropy": 5.756379699707031,
      "epoch": 0.17955978890462093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00020794069617439942,
      "loss": 5.6042,
      "mean_token_accuracy": 0.1755429908633232,
      "num_tokens": 6304327.0,
      "step": 2790
    },
    {
      "entropy": 5.652706098556519,
      "epoch": 0.17988158064100915,
      "grad_norm": 1.0,
      "learning_rate": 0.00020681706433262593,
      "loss": 5.5383,
      "mean_token_accuracy": 0.17942026257514954,
      "num_tokens": 6315080.0,
      "step": 2795
    },
    {
      "entropy": 5.796750879287719,
      "epoch": 0.18020337237739734,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00020569530176047602,
      "loss": 5.7145,
      "mean_token_accuracy": 0.16154513657093048,
      "num_tokens": 6326772.0,
      "step": 2800
    },
    {
      "entropy": 5.774946546554565,
      "epoch": 0.18052516411378555,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002045754392116374,
      "loss": 5.5963,
      "mean_token_accuracy": 0.172379070520401,
      "num_tokens": 6337918.0,
      "step": 2805
    },
    {
      "entropy": 5.668222045898437,
      "epoch": 0.18084695585017377,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00020345750738770757,
      "loss": 5.5836,
      "mean_token_accuracy": 0.17521809041500092,
      "num_tokens": 6348954.0,
      "step": 2810
    },
    {
      "entropy": 5.701242160797119,
      "epoch": 0.18116874758656198,
      "grad_norm": 1.140625,
      "learning_rate": 0.00020234153693735214,
      "loss": 5.5009,
      "mean_token_accuracy": 0.17282357066869736,
      "num_tokens": 6359863.0,
      "step": 2815
    },
    {
      "entropy": 5.7502178192138675,
      "epoch": 0.18149053932295017,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002012275584554647,
      "loss": 5.6219,
      "mean_token_accuracy": 0.16867788434028624,
      "num_tokens": 6370528.0,
      "step": 2820
    },
    {
      "entropy": 5.678765726089478,
      "epoch": 0.1818123310593384,
      "grad_norm": 1.09375,
      "learning_rate": 0.00020011560248232803,
      "loss": 5.5851,
      "mean_token_accuracy": 0.1763036862015724,
      "num_tokens": 6381041.0,
      "step": 2825
    },
    {
      "entropy": 5.758502578735351,
      "epoch": 0.1821341227957266,
      "grad_norm": 0.9375,
      "learning_rate": 0.00019900569950277692,
      "loss": 5.5958,
      "mean_token_accuracy": 0.17222651094198227,
      "num_tokens": 6393876.0,
      "step": 2830
    },
    {
      "entropy": 5.71019926071167,
      "epoch": 0.18245591453211482,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00019789787994536228,
      "loss": 5.5821,
      "mean_token_accuracy": 0.17190856039524077,
      "num_tokens": 6405367.0,
      "step": 2835
    },
    {
      "entropy": 5.7003566265106205,
      "epoch": 0.18277770626850304,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00019679217418151667,
      "loss": 5.5759,
      "mean_token_accuracy": 0.17772747427225113,
      "num_tokens": 6416340.0,
      "step": 2840
    },
    {
      "entropy": 5.714562940597534,
      "epoch": 0.18309949800489123,
      "grad_norm": 1.0,
      "learning_rate": 0.00019568861252472236,
      "loss": 5.5691,
      "mean_token_accuracy": 0.1808065801858902,
      "num_tokens": 6427479.0,
      "step": 2845
    },
    {
      "entropy": 5.718938398361206,
      "epoch": 0.18342128974127944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00019458722522967952,
      "loss": 5.5369,
      "mean_token_accuracy": 0.18007902055978775,
      "num_tokens": 6437781.0,
      "step": 2850
    },
    {
      "entropy": 5.652391481399536,
      "epoch": 0.18374308147766766,
      "grad_norm": 1.03125,
      "learning_rate": 0.00019348804249147723,
      "loss": 5.4885,
      "mean_token_accuracy": 0.1808314561843872,
      "num_tokens": 6449835.0,
      "step": 2855
    },
    {
      "entropy": 5.704882049560547,
      "epoch": 0.18406487321405587,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0001923910944447655,
      "loss": 5.537,
      "mean_token_accuracy": 0.18006470203399658,
      "num_tokens": 6461589.0,
      "step": 2860
    },
    {
      "entropy": 5.8008747577667235,
      "epoch": 0.18438666495044406,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00019129641116292928,
      "loss": 5.725,
      "mean_token_accuracy": 0.16126319169998168,
      "num_tokens": 6472816.0,
      "step": 2865
    },
    {
      "entropy": 5.775189208984375,
      "epoch": 0.18470845668683228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00019020402265726343,
      "loss": 5.6478,
      "mean_token_accuracy": 0.17083903700113295,
      "num_tokens": 6485009.0,
      "step": 2870
    },
    {
      "entropy": 5.604465532302856,
      "epoch": 0.1850302484232205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0001891139588761509,
      "loss": 5.3975,
      "mean_token_accuracy": 0.18840058892965317,
      "num_tokens": 6495985.0,
      "step": 2875
    },
    {
      "entropy": 5.6450989723205565,
      "epoch": 0.1853520401596087,
      "grad_norm": 1.5,
      "learning_rate": 0.00018802624970424076,
      "loss": 5.5737,
      "mean_token_accuracy": 0.1801061749458313,
      "num_tokens": 6508603.0,
      "step": 2880
    },
    {
      "entropy": 5.747278213500977,
      "epoch": 0.1856738318959969,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00018694092496162945,
      "loss": 5.6144,
      "mean_token_accuracy": 0.17110922634601594,
      "num_tokens": 6520783.0,
      "step": 2885
    },
    {
      "entropy": 5.7772563934326175,
      "epoch": 0.18599562363238512,
      "grad_norm": 0.953125,
      "learning_rate": 0.00018585801440304306,
      "loss": 5.6304,
      "mean_token_accuracy": 0.17218098789453506,
      "num_tokens": 6532257.0,
      "step": 2890
    },
    {
      "entropy": 5.664489316940307,
      "epoch": 0.18631741536877333,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00018477754771702165,
      "loss": 5.5861,
      "mean_token_accuracy": 0.17788384556770326,
      "num_tokens": 6543295.0,
      "step": 2895
    },
    {
      "entropy": 5.667116689682007,
      "epoch": 0.18663920710516155,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00018369955452510506,
      "loss": 5.471,
      "mean_token_accuracy": 0.1864291697740555,
      "num_tokens": 6555037.0,
      "step": 2900
    },
    {
      "entropy": 5.672414684295655,
      "epoch": 0.18696099884154974,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0001826240643810212,
      "loss": 5.5482,
      "mean_token_accuracy": 0.18003589808940887,
      "num_tokens": 6566730.0,
      "step": 2905
    },
    {
      "entropy": 5.61477255821228,
      "epoch": 0.18728279057793795,
      "grad_norm": 1.015625,
      "learning_rate": 0.0001815511067698758,
      "loss": 5.4704,
      "mean_token_accuracy": 0.18574921488761903,
      "num_tokens": 6577080.0,
      "step": 2910
    },
    {
      "entropy": 5.692835092544556,
      "epoch": 0.18760458231432617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0001804807111073436,
      "loss": 5.462,
      "mean_token_accuracy": 0.17626544833183289,
      "num_tokens": 6587714.0,
      "step": 2915
    },
    {
      "entropy": 5.739736890792846,
      "epoch": 0.18792637405071438,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0001794129067388625,
      "loss": 5.6253,
      "mean_token_accuracy": 0.17472572922706603,
      "num_tokens": 6599442.0,
      "step": 2920
    },
    {
      "entropy": 5.8049914836883545,
      "epoch": 0.1882481657871026,
      "grad_norm": 1.078125,
      "learning_rate": 0.00017834772293882868,
      "loss": 5.65,
      "mean_token_accuracy": 0.16884101331233978,
      "num_tokens": 6610406.0,
      "step": 2925
    },
    {
      "entropy": 5.8121825695037845,
      "epoch": 0.1885699575234908,
      "grad_norm": 0.96875,
      "learning_rate": 0.000177285188909794,
      "loss": 5.649,
      "mean_token_accuracy": 0.16940721571445466,
      "num_tokens": 6621396.0,
      "step": 2930
    },
    {
      "entropy": 5.77225341796875,
      "epoch": 0.188891749259879,
      "grad_norm": 0.984375,
      "learning_rate": 0.0001762253337816656,
      "loss": 5.5953,
      "mean_token_accuracy": 0.17360182255506515,
      "num_tokens": 6633391.0,
      "step": 2935
    },
    {
      "entropy": 5.738308620452881,
      "epoch": 0.18921354099626722,
      "grad_norm": 1.015625,
      "learning_rate": 0.00017516818661090738,
      "loss": 5.5776,
      "mean_token_accuracy": 0.17296917736530304,
      "num_tokens": 6645574.0,
      "step": 2940
    },
    {
      "entropy": 5.69874119758606,
      "epoch": 0.18953533273265544,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0001741137763797428,
      "loss": 5.4534,
      "mean_token_accuracy": 0.18476555794477462,
      "num_tokens": 6656839.0,
      "step": 2945
    },
    {
      "entropy": 5.674434041976928,
      "epoch": 0.18985712446904363,
      "grad_norm": 1.03125,
      "learning_rate": 0.00017306213199536115,
      "loss": 5.6163,
      "mean_token_accuracy": 0.17427906692028045,
      "num_tokens": 6669177.0,
      "step": 2950
    },
    {
      "entropy": 5.748139667510986,
      "epoch": 0.19017891620543184,
      "grad_norm": 1.0,
      "learning_rate": 0.0001720132822891243,
      "loss": 5.6176,
      "mean_token_accuracy": 0.174359992146492,
      "num_tokens": 6681171.0,
      "step": 2955
    },
    {
      "entropy": 5.731746482849121,
      "epoch": 0.19050070794182006,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0001709672560157769,
      "loss": 5.4987,
      "mean_token_accuracy": 0.18219124376773835,
      "num_tokens": 6694005.0,
      "step": 2960
    },
    {
      "entropy": 5.724074411392212,
      "epoch": 0.19082249967820827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00016992408185265758,
      "loss": 5.5515,
      "mean_token_accuracy": 0.1818791389465332,
      "num_tokens": 6705131.0,
      "step": 2965
    },
    {
      "entropy": 5.668516206741333,
      "epoch": 0.19114429141459646,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00016888378839891298,
      "loss": 5.514,
      "mean_token_accuracy": 0.18113282769918443,
      "num_tokens": 6716993.0,
      "step": 2970
    },
    {
      "entropy": 5.645058488845825,
      "epoch": 0.19146608315098468,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0001678464041747137,
      "loss": 5.5056,
      "mean_token_accuracy": 0.1877603128552437,
      "num_tokens": 6728475.0,
      "step": 2975
    },
    {
      "entropy": 5.680294704437256,
      "epoch": 0.1917878748873729,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00016681195762047223,
      "loss": 5.6187,
      "mean_token_accuracy": 0.1748928114771843,
      "num_tokens": 6740103.0,
      "step": 2980
    },
    {
      "entropy": 5.758680152893066,
      "epoch": 0.1921096666237611,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00016578047709606337,
      "loss": 5.6349,
      "mean_token_accuracy": 0.16935867220163345,
      "num_tokens": 6752667.0,
      "step": 2985
    },
    {
      "entropy": 5.67816390991211,
      "epoch": 0.1924314583601493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00016475199088004678,
      "loss": 5.4015,
      "mean_token_accuracy": 0.19075028896331786,
      "num_tokens": 6764225.0,
      "step": 2990
    },
    {
      "entropy": 5.688899850845337,
      "epoch": 0.19275325009653752,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00016372652716889163,
      "loss": 5.5174,
      "mean_token_accuracy": 0.18526418507099152,
      "num_tokens": 6775582.0,
      "step": 2995
    },
    {
      "entropy": 5.558641862869263,
      "epoch": 0.19307504183292573,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0001627041140762035,
      "loss": 5.4904,
      "mean_token_accuracy": 0.18614665418863297,
      "num_tokens": 6787143.0,
      "step": 3000
    },
    {
      "entropy": 5.6982316970825195,
      "epoch": 0.19339683356931395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00016168477963195382,
      "loss": 5.5969,
      "mean_token_accuracy": 0.1700701579451561,
      "num_tokens": 6798331.0,
      "step": 3005
    },
    {
      "entropy": 5.812347269058227,
      "epoch": 0.19371862530570216,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0001606685517817114,
      "loss": 5.5896,
      "mean_token_accuracy": 0.17314366847276688,
      "num_tokens": 6809174.0,
      "step": 3010
    },
    {
      "entropy": 5.695400857925415,
      "epoch": 0.19404041704209035,
      "grad_norm": 1.03125,
      "learning_rate": 0.00015965545838587592,
      "loss": 5.4838,
      "mean_token_accuracy": 0.1871499627828598,
      "num_tokens": 6821430.0,
      "step": 3015
    },
    {
      "entropy": 5.699887084960937,
      "epoch": 0.19436220877847857,
      "grad_norm": 1.03125,
      "learning_rate": 0.00015864552721891467,
      "loss": 5.4952,
      "mean_token_accuracy": 0.18506361842155455,
      "num_tokens": 6833543.0,
      "step": 3020
    },
    {
      "entropy": 5.548637390136719,
      "epoch": 0.19468400051486678,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00015763878596860076,
      "loss": 5.3546,
      "mean_token_accuracy": 0.19219871312379838,
      "num_tokens": 6843535.0,
      "step": 3025
    },
    {
      "entropy": 5.6764974117279055,
      "epoch": 0.195005792251255,
      "grad_norm": 0.953125,
      "learning_rate": 0.00015663526223525412,
      "loss": 5.6167,
      "mean_token_accuracy": 0.17830319777131082,
      "num_tokens": 6855675.0,
      "step": 3030
    },
    {
      "entropy": 5.640204763412475,
      "epoch": 0.1953275839876432,
      "grad_norm": 1.09375,
      "learning_rate": 0.0001556349835309848,
      "loss": 5.4516,
      "mean_token_accuracy": 0.178354512155056,
      "num_tokens": 6867209.0,
      "step": 3035
    },
    {
      "entropy": 5.634023904800415,
      "epoch": 0.1956493757240314,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001546379772789389,
      "loss": 5.4462,
      "mean_token_accuracy": 0.18591717332601548,
      "num_tokens": 6878631.0,
      "step": 3040
    },
    {
      "entropy": 5.704431819915771,
      "epoch": 0.19597116746041962,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00015364427081254622,
      "loss": 5.519,
      "mean_token_accuracy": 0.18015825897455215,
      "num_tokens": 6889264.0,
      "step": 3045
    },
    {
      "entropy": 5.801104736328125,
      "epoch": 0.19629295919680784,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00015265389137477165,
      "loss": 5.6561,
      "mean_token_accuracy": 0.17016648277640342,
      "num_tokens": 6899769.0,
      "step": 3050
    },
    {
      "entropy": 5.756004047393799,
      "epoch": 0.19661475093319603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00015166686611736786,
      "loss": 5.533,
      "mean_token_accuracy": 0.17748568803071976,
      "num_tokens": 6910779.0,
      "step": 3055
    },
    {
      "entropy": 5.769090080261231,
      "epoch": 0.19693654266958424,
      "grad_norm": 1.21875,
      "learning_rate": 0.00015068322210013064,
      "loss": 5.5515,
      "mean_token_accuracy": 0.18222929090261458,
      "num_tokens": 6921653.0,
      "step": 3060
    },
    {
      "entropy": 5.624980401992798,
      "epoch": 0.19725833440597246,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0001497029862901578,
      "loss": 5.4713,
      "mean_token_accuracy": 0.1844359040260315,
      "num_tokens": 6933836.0,
      "step": 3065
    },
    {
      "entropy": 5.612014818191528,
      "epoch": 0.19758012614236067,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00014872618556110905,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18860455006361007,
      "num_tokens": 6944312.0,
      "step": 3070
    },
    {
      "entropy": 5.679592180252075,
      "epoch": 0.19790191787874886,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00014775284669246992,
      "loss": 5.5213,
      "mean_token_accuracy": 0.18389405608177184,
      "num_tokens": 6955779.0,
      "step": 3075
    },
    {
      "entropy": 5.597935438156128,
      "epoch": 0.19822370961513708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00014678299636881716,
      "loss": 5.4889,
      "mean_token_accuracy": 0.18244732469320296,
      "num_tokens": 6967598.0,
      "step": 3080
    },
    {
      "entropy": 5.680601692199707,
      "epoch": 0.1985455013515253,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001458166611790873,
      "loss": 5.4742,
      "mean_token_accuracy": 0.17682809382677078,
      "num_tokens": 6979624.0,
      "step": 3085
    },
    {
      "entropy": 5.733098745346069,
      "epoch": 0.1988672930879135,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00014485386761584773,
      "loss": 5.5196,
      "mean_token_accuracy": 0.17878370732069016,
      "num_tokens": 6990673.0,
      "step": 3090
    },
    {
      "entropy": 5.669073295593262,
      "epoch": 0.1991890848243017,
      "grad_norm": 0.96875,
      "learning_rate": 0.00014389464207457042,
      "loss": 5.4721,
      "mean_token_accuracy": 0.18245527744293213,
      "num_tokens": 7002412.0,
      "step": 3095
    },
    {
      "entropy": 5.74008150100708,
      "epoch": 0.19951087656068991,
      "grad_norm": 1.03125,
      "learning_rate": 0.00014293901085290795,
      "loss": 5.58,
      "mean_token_accuracy": 0.17535691112279891,
      "num_tokens": 7013909.0,
      "step": 3100
    },
    {
      "entropy": 5.671574306488037,
      "epoch": 0.19983266829707813,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00014198700014997307,
      "loss": 5.535,
      "mean_token_accuracy": 0.1856901690363884,
      "num_tokens": 7026270.0,
      "step": 3105
    },
    {
      "entropy": 5.690511417388916,
      "epoch": 0.20015446003346635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00014103863606562016,
      "loss": 5.4816,
      "mean_token_accuracy": 0.182679383456707,
      "num_tokens": 7038255.0,
      "step": 3110
    },
    {
      "entropy": 5.72007737159729,
      "epoch": 0.20047625176985456,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00014009394459972964,
      "loss": 5.4935,
      "mean_token_accuracy": 0.19080989062786102,
      "num_tokens": 7049601.0,
      "step": 3115
    },
    {
      "entropy": 5.662781763076782,
      "epoch": 0.20079804350624275,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00013915295165149513,
      "loss": 5.4249,
      "mean_token_accuracy": 0.18327374756336212,
      "num_tokens": 7061352.0,
      "step": 3120
    },
    {
      "entropy": 5.660637235641479,
      "epoch": 0.20111983524263097,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00013821568301871384,
      "loss": 5.53,
      "mean_token_accuracy": 0.17459167689085006,
      "num_tokens": 7072555.0,
      "step": 3125
    },
    {
      "entropy": 5.69699501991272,
      "epoch": 0.20144162697901918,
      "grad_norm": 1.046875,
      "learning_rate": 0.00013728216439707862,
      "loss": 5.4997,
      "mean_token_accuracy": 0.17946283221244813,
      "num_tokens": 7083198.0,
      "step": 3130
    },
    {
      "entropy": 5.5951892852783205,
      "epoch": 0.2017634187154074,
      "grad_norm": 0.96875,
      "learning_rate": 0.00013635242137947419,
      "loss": 5.3927,
      "mean_token_accuracy": 0.18453539460897445,
      "num_tokens": 7094039.0,
      "step": 3135
    },
    {
      "entropy": 5.626377773284912,
      "epoch": 0.2020852104517956,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00013542647945527498,
      "loss": 5.4677,
      "mean_token_accuracy": 0.18502026945352554,
      "num_tokens": 7105343.0,
      "step": 3140
    },
    {
      "entropy": 5.680795907974243,
      "epoch": 0.2024070021881838,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001345043640096465,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17568370699882507,
      "num_tokens": 7118957.0,
      "step": 3145
    },
    {
      "entropy": 5.774059247970581,
      "epoch": 0.20272879392457202,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00013358610032284957,
      "loss": 5.5571,
      "mean_token_accuracy": 0.17776385843753814,
      "num_tokens": 7131129.0,
      "step": 3150
    },
    {
      "entropy": 5.695317554473877,
      "epoch": 0.20305058566096024,
      "grad_norm": 1.0,
      "learning_rate": 0.000132671713569547,
      "loss": 5.5671,
      "mean_token_accuracy": 0.18454511761665343,
      "num_tokens": 7142735.0,
      "step": 3155
    },
    {
      "entropy": 5.746871376037598,
      "epoch": 0.20337237739734843,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0001317612288181136,
      "loss": 5.6096,
      "mean_token_accuracy": 0.17044817805290222,
      "num_tokens": 7154658.0,
      "step": 3160
    },
    {
      "entropy": 5.630042362213135,
      "epoch": 0.20369416913373664,
      "grad_norm": 1.015625,
      "learning_rate": 0.00013085467102994864,
      "loss": 5.4589,
      "mean_token_accuracy": 0.18516594171524048,
      "num_tokens": 7165127.0,
      "step": 3165
    },
    {
      "entropy": 5.690448999404907,
      "epoch": 0.20401596087012486,
      "grad_norm": 1.109375,
      "learning_rate": 0.00012995206505879198,
      "loss": 5.4718,
      "mean_token_accuracy": 0.18847323954105377,
      "num_tokens": 7176355.0,
      "step": 3170
    },
    {
      "entropy": 5.710048961639404,
      "epoch": 0.20433775260651307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001290534356500421,
      "loss": 5.5361,
      "mean_token_accuracy": 0.17686813175678254,
      "num_tokens": 7188056.0,
      "step": 3175
    },
    {
      "entropy": 5.649774217605591,
      "epoch": 0.20465954434290126,
      "grad_norm": 1.140625,
      "learning_rate": 0.00012815880744007827,
      "loss": 5.4608,
      "mean_token_accuracy": 0.1936315268278122,
      "num_tokens": 7197816.0,
      "step": 3180
    },
    {
      "entropy": 5.734100008010865,
      "epoch": 0.20498133607928948,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00012726820495558483,
      "loss": 5.6635,
      "mean_token_accuracy": 0.17429392635822297,
      "num_tokens": 7208354.0,
      "step": 3185
    },
    {
      "entropy": 5.723574447631836,
      "epoch": 0.2053031278156777,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00012638165261287868,
      "loss": 5.4728,
      "mean_token_accuracy": 0.1796579271554947,
      "num_tokens": 7218932.0,
      "step": 3190
    },
    {
      "entropy": 5.774714851379395,
      "epoch": 0.2056249195520659,
      "grad_norm": 1.0,
      "learning_rate": 0.0001254991747172402,
      "loss": 5.6235,
      "mean_token_accuracy": 0.1639267086982727,
      "num_tokens": 7229197.0,
      "step": 3195
    },
    {
      "entropy": 5.687872886657715,
      "epoch": 0.20594671128845413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00012462079546224662,
      "loss": 5.4341,
      "mean_token_accuracy": 0.18549608439207077,
      "num_tokens": 7239336.0,
      "step": 3200
    },
    {
      "entropy": 5.704177951812744,
      "epoch": 0.20626850302484231,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00012374653892910896,
      "loss": 5.5501,
      "mean_token_accuracy": 0.17581791430711746,
      "num_tokens": 7250354.0,
      "step": 3205
    },
    {
      "entropy": 5.771119451522827,
      "epoch": 0.20659029476123053,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00012287642908601166,
      "loss": 5.5455,
      "mean_token_accuracy": 0.18164872974157334,
      "num_tokens": 7261537.0,
      "step": 3210
    },
    {
      "entropy": 5.67866735458374,
      "epoch": 0.20691208649761875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00012201048978745569,
      "loss": 5.5029,
      "mean_token_accuracy": 0.18419386744499205,
      "num_tokens": 7272017.0,
      "step": 3215
    },
    {
      "entropy": 5.781367492675781,
      "epoch": 0.20723387823400696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00012114874477360427,
      "loss": 5.5905,
      "mean_token_accuracy": 0.17322648614645003,
      "num_tokens": 7283979.0,
      "step": 3220
    },
    {
      "entropy": 5.7428923606872555,
      "epoch": 0.20755566997039515,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00012029121766963236,
      "loss": 5.5248,
      "mean_token_accuracy": 0.1830981805920601,
      "num_tokens": 7295494.0,
      "step": 3225
    },
    {
      "entropy": 5.712482929229736,
      "epoch": 0.20787746170678337,
      "grad_norm": 1.125,
      "learning_rate": 0.00011943793198507858,
      "loss": 5.5048,
      "mean_token_accuracy": 0.1833385020494461,
      "num_tokens": 7305537.0,
      "step": 3230
    },
    {
      "entropy": 5.649680185317993,
      "epoch": 0.20819925344317158,
      "grad_norm": 0.96875,
      "learning_rate": 0.00011858891111320104,
      "loss": 5.4139,
      "mean_token_accuracy": 0.18263983577489853,
      "num_tokens": 7316627.0,
      "step": 3235
    },
    {
      "entropy": 5.721251440048218,
      "epoch": 0.2085210451795598,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0001177441783303359,
      "loss": 5.5126,
      "mean_token_accuracy": 0.18288784176111222,
      "num_tokens": 7326208.0,
      "step": 3240
    },
    {
      "entropy": 5.6964263916015625,
      "epoch": 0.208842836915948,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00011690375679525896,
      "loss": 5.5162,
      "mean_token_accuracy": 0.17970404028892517,
      "num_tokens": 7337096.0,
      "step": 3245
    },
    {
      "entropy": 5.671505928039551,
      "epoch": 0.2091646286523362,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00011606766954855124,
      "loss": 5.4874,
      "mean_token_accuracy": 0.18251161873340607,
      "num_tokens": 7348054.0,
      "step": 3250
    },
    {
      "entropy": 5.660500144958496,
      "epoch": 0.20948642038872442,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00011523593951196702,
      "loss": 5.4746,
      "mean_token_accuracy": 0.18285904824733734,
      "num_tokens": 7359499.0,
      "step": 3255
    },
    {
      "entropy": 5.676732158660888,
      "epoch": 0.20980821212511264,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00011440858948780523,
      "loss": 5.4623,
      "mean_token_accuracy": 0.17779761850833892,
      "num_tokens": 7370092.0,
      "step": 3260
    },
    {
      "entropy": 5.6439310073852536,
      "epoch": 0.21013000386150082,
      "grad_norm": 1.046875,
      "learning_rate": 0.00011358564215828484,
      "loss": 5.5176,
      "mean_token_accuracy": 0.17610785812139512,
      "num_tokens": 7381298.0,
      "step": 3265
    },
    {
      "entropy": 5.660351896286011,
      "epoch": 0.21045179559788904,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00011276712008492254,
      "loss": 5.4981,
      "mean_token_accuracy": 0.17899704724550247,
      "num_tokens": 7394703.0,
      "step": 3270
    },
    {
      "entropy": 5.68933482170105,
      "epoch": 0.21077358733427726,
      "grad_norm": 1.015625,
      "learning_rate": 0.00011195304570791451,
      "loss": 5.5958,
      "mean_token_accuracy": 0.1734357014298439,
      "num_tokens": 7407061.0,
      "step": 3275
    },
    {
      "entropy": 5.659317493438721,
      "epoch": 0.21109537907066547,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00011114344134552094,
      "loss": 5.4822,
      "mean_token_accuracy": 0.1914387345314026,
      "num_tokens": 7419503.0,
      "step": 3280
    },
    {
      "entropy": 5.689638662338257,
      "epoch": 0.21141717080705366,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0001103383291934545,
      "loss": 5.4705,
      "mean_token_accuracy": 0.18759380429983138,
      "num_tokens": 7430364.0,
      "step": 3285
    },
    {
      "entropy": 5.73215069770813,
      "epoch": 0.21173896254344188,
      "grad_norm": 0.921875,
      "learning_rate": 0.00010953773132427141,
      "loss": 5.5562,
      "mean_token_accuracy": 0.17441725879907607,
      "num_tokens": 7444001.0,
      "step": 3290
    },
    {
      "entropy": 5.6462126731872555,
      "epoch": 0.2120607542798301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00010874166968676677,
      "loss": 5.45,
      "mean_token_accuracy": 0.18846633285284042,
      "num_tokens": 7454508.0,
      "step": 3295
    },
    {
      "entropy": 5.653306198120117,
      "epoch": 0.2123825460162183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00010795016610537251,
      "loss": 5.4163,
      "mean_token_accuracy": 0.19095997512340546,
      "num_tokens": 7465157.0,
      "step": 3300
    },
    {
      "entropy": 5.684601783752441,
      "epoch": 0.21270433775260653,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00010716324227955904,
      "loss": 5.4874,
      "mean_token_accuracy": 0.17782267332077026,
      "num_tokens": 7477722.0,
      "step": 3305
    },
    {
      "entropy": 5.74354658126831,
      "epoch": 0.21302612948899471,
      "grad_norm": 1.125,
      "learning_rate": 0.0001063809197832406,
      "loss": 5.5618,
      "mean_token_accuracy": 0.1757574662566185,
      "num_tokens": 7489024.0,
      "step": 3310
    },
    {
      "entropy": 5.679970693588257,
      "epoch": 0.21334792122538293,
      "grad_norm": 0.859375,
      "learning_rate": 0.00010560322006418368,
      "loss": 5.5289,
      "mean_token_accuracy": 0.18203138262033464,
      "num_tokens": 7501245.0,
      "step": 3315
    },
    {
      "entropy": 5.720022583007813,
      "epoch": 0.21366971296177115,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00010483016444341887,
      "loss": 5.601,
      "mean_token_accuracy": 0.17927283197641372,
      "num_tokens": 7513044.0,
      "step": 3320
    },
    {
      "entropy": 5.743742179870606,
      "epoch": 0.21399150469815936,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00010406177411465654,
      "loss": 5.6288,
      "mean_token_accuracy": 0.17428549081087114,
      "num_tokens": 7524675.0,
      "step": 3325
    },
    {
      "entropy": 5.625386142730713,
      "epoch": 0.21431329643454755,
      "grad_norm": 1.03125,
      "learning_rate": 0.00010329807014370562,
      "loss": 5.406,
      "mean_token_accuracy": 0.19135176986455918,
      "num_tokens": 7535721.0,
      "step": 3330
    },
    {
      "entropy": 5.727019214630127,
      "epoch": 0.21463508817093577,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00010253907346789632,
      "loss": 5.4711,
      "mean_token_accuracy": 0.18167847245931626,
      "num_tokens": 7547056.0,
      "step": 3335
    },
    {
      "entropy": 5.669043064117432,
      "epoch": 0.21495687990732398,
      "grad_norm": 1.03125,
      "learning_rate": 0.00010178480489550596,
      "loss": 5.5082,
      "mean_token_accuracy": 0.17800174355506898,
      "num_tokens": 7559792.0,
      "step": 3340
    },
    {
      "entropy": 5.63152847290039,
      "epoch": 0.2152786716437122,
      "grad_norm": 1.078125,
      "learning_rate": 0.00010103528510518836,
      "loss": 5.5123,
      "mean_token_accuracy": 0.18170783668756485,
      "num_tokens": 7570995.0,
      "step": 3345
    },
    {
      "entropy": 5.715353727340698,
      "epoch": 0.2156004633801004,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0001002905346454073,
      "loss": 5.571,
      "mean_token_accuracy": 0.1725604072213173,
      "num_tokens": 7582502.0,
      "step": 3350
    },
    {
      "entropy": 5.646735763549804,
      "epoch": 0.2159222551164886,
      "grad_norm": 0.97265625,
      "learning_rate": 9.955057393387285e-05,
      "loss": 5.4019,
      "mean_token_accuracy": 0.19154874086380005,
      "num_tokens": 7594006.0,
      "step": 3355
    },
    {
      "entropy": 5.672197055816651,
      "epoch": 0.21624404685287682,
      "grad_norm": 1.0078125,
      "learning_rate": 9.88154232569816e-05,
      "loss": 5.4422,
      "mean_token_accuracy": 0.1860278606414795,
      "num_tokens": 7605172.0,
      "step": 3360
    },
    {
      "entropy": 5.630903816223144,
      "epoch": 0.21656583858926504,
      "grad_norm": 1.1875,
      "learning_rate": 9.808510276926075e-05,
      "loss": 5.3921,
      "mean_token_accuracy": 0.18900154083967208,
      "num_tokens": 7616169.0,
      "step": 3365
    },
    {
      "entropy": 5.641145801544189,
      "epoch": 0.21688763032565322,
      "grad_norm": 0.97265625,
      "learning_rate": 9.735963249281549e-05,
      "loss": 5.5257,
      "mean_token_accuracy": 0.17962203174829483,
      "num_tokens": 7627515.0,
      "step": 3370
    },
    {
      "entropy": 5.6899768829345705,
      "epoch": 0.21720942206204144,
      "grad_norm": 0.99609375,
      "learning_rate": 9.663903231677974e-05,
      "loss": 5.4922,
      "mean_token_accuracy": 0.17624151110649108,
      "num_tokens": 7639585.0,
      "step": 3375
    },
    {
      "entropy": 5.692270660400391,
      "epoch": 0.21753121379842966,
      "grad_norm": 0.97265625,
      "learning_rate": 9.592332199677145e-05,
      "loss": 5.42,
      "mean_token_accuracy": 0.18190095871686934,
      "num_tokens": 7650905.0,
      "step": 3380
    },
    {
      "entropy": 5.644244956970215,
      "epoch": 0.21785300553481787,
      "grad_norm": 1.03125,
      "learning_rate": 9.521252115435061e-05,
      "loss": 5.3955,
      "mean_token_accuracy": 0.18672704100608825,
      "num_tokens": 7662558.0,
      "step": 3385
    },
    {
      "entropy": 5.6287590026855465,
      "epoch": 0.2181747972712061,
      "grad_norm": 0.90625,
      "learning_rate": 9.450664927648126e-05,
      "loss": 5.513,
      "mean_token_accuracy": 0.17950384467840194,
      "num_tokens": 7674124.0,
      "step": 3390
    },
    {
      "entropy": 5.6026817798614506,
      "epoch": 0.21849658900759428,
      "grad_norm": 1.0234375,
      "learning_rate": 9.380572571499758e-05,
      "loss": 5.4205,
      "mean_token_accuracy": 0.18300425559282302,
      "num_tokens": 7684596.0,
      "step": 3395
    },
    {
      "entropy": 5.705789279937744,
      "epoch": 0.2188183807439825,
      "grad_norm": 0.9296875,
      "learning_rate": 9.310976968607307e-05,
      "loss": 5.5586,
      "mean_token_accuracy": 0.18176765888929367,
      "num_tokens": 7696356.0,
      "step": 3400
    },
    {
      "entropy": 5.63036208152771,
      "epoch": 0.2191401724803707,
      "grad_norm": 1.0078125,
      "learning_rate": 9.241880026969381e-05,
      "loss": 5.4223,
      "mean_token_accuracy": 0.18291159421205522,
      "num_tokens": 7707970.0,
      "step": 3405
    },
    {
      "entropy": 5.735514307022095,
      "epoch": 0.21946196421675893,
      "grad_norm": 1.015625,
      "learning_rate": 9.173283640913537e-05,
      "loss": 5.5037,
      "mean_token_accuracy": 0.18661999106407165,
      "num_tokens": 7718777.0,
      "step": 3410
    },
    {
      "entropy": 5.645886325836182,
      "epoch": 0.21978375595314711,
      "grad_norm": 1.0703125,
      "learning_rate": 9.10518969104436e-05,
      "loss": 5.4772,
      "mean_token_accuracy": 0.1833420976996422,
      "num_tokens": 7730870.0,
      "step": 3415
    },
    {
      "entropy": 5.687615871429443,
      "epoch": 0.22010554768953533,
      "grad_norm": 1.046875,
      "learning_rate": 9.037600044191868e-05,
      "loss": 5.4984,
      "mean_token_accuracy": 0.1802757829427719,
      "num_tokens": 7741459.0,
      "step": 3420
    },
    {
      "entropy": 5.705859279632568,
      "epoch": 0.22042733942592355,
      "grad_norm": 1.0390625,
      "learning_rate": 8.970516553360383e-05,
      "loss": 5.5259,
      "mean_token_accuracy": 0.18371005952358246,
      "num_tokens": 7752561.0,
      "step": 3425
    },
    {
      "entropy": 5.632897424697876,
      "epoch": 0.22074913116231176,
      "grad_norm": 1.1015625,
      "learning_rate": 8.903941057677692e-05,
      "loss": 5.5002,
      "mean_token_accuracy": 0.18794768750667573,
      "num_tokens": 7764409.0,
      "step": 3430
    },
    {
      "entropy": 5.678521251678466,
      "epoch": 0.22107092289869995,
      "grad_norm": 1.0703125,
      "learning_rate": 8.837875382344635e-05,
      "loss": 5.4496,
      "mean_token_accuracy": 0.18130633383989334,
      "num_tokens": 7775348.0,
      "step": 3435
    },
    {
      "entropy": 5.66662073135376,
      "epoch": 0.22139271463508817,
      "grad_norm": 0.98828125,
      "learning_rate": 8.772321338585076e-05,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18357299119234086,
      "num_tokens": 7787564.0,
      "step": 3440
    },
    {
      "entropy": 5.7291735172271725,
      "epoch": 0.22171450637147638,
      "grad_norm": 1.1484375,
      "learning_rate": 8.707280723596242e-05,
      "loss": 5.5575,
      "mean_token_accuracy": 0.1795124188065529,
      "num_tokens": 7798285.0,
      "step": 3445
    },
    {
      "entropy": 5.62374472618103,
      "epoch": 0.2220362981078646,
      "grad_norm": 0.9453125,
      "learning_rate": 8.64275532049944e-05,
      "loss": 5.4699,
      "mean_token_accuracy": 0.1861945241689682,
      "num_tokens": 7810085.0,
      "step": 3450
    },
    {
      "entropy": 5.7821448802947994,
      "epoch": 0.2223580898442528,
      "grad_norm": 0.96875,
      "learning_rate": 8.578746898291198e-05,
      "loss": 5.6248,
      "mean_token_accuracy": 0.18239967674016952,
      "num_tokens": 7821392.0,
      "step": 3455
    },
    {
      "entropy": 5.709903621673584,
      "epoch": 0.222679881580641,
      "grad_norm": 1.0546875,
      "learning_rate": 8.515257211794742e-05,
      "loss": 5.5206,
      "mean_token_accuracy": 0.18002507090568542,
      "num_tokens": 7831695.0,
      "step": 3460
    },
    {
      "entropy": 5.75536732673645,
      "epoch": 0.22300167331702922,
      "grad_norm": 1.046875,
      "learning_rate": 8.452288001611896e-05,
      "loss": 5.5439,
      "mean_token_accuracy": 0.1809769243001938,
      "num_tokens": 7844648.0,
      "step": 3465
    },
    {
      "entropy": 5.682497310638428,
      "epoch": 0.22332346505341744,
      "grad_norm": 0.9140625,
      "learning_rate": 8.389840994075379e-05,
      "loss": 5.4994,
      "mean_token_accuracy": 0.18335730731487274,
      "num_tokens": 7856473.0,
      "step": 3470
    },
    {
      "entropy": 5.710560464859009,
      "epoch": 0.22364525678980562,
      "grad_norm": 1.0234375,
      "learning_rate": 8.327917901201435e-05,
      "loss": 5.4373,
      "mean_token_accuracy": 0.1875859409570694,
      "num_tokens": 7865972.0,
      "step": 3475
    },
    {
      "entropy": 5.709625625610352,
      "epoch": 0.22396704852619384,
      "grad_norm": 1.09375,
      "learning_rate": 8.266520420642931e-05,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18544894456863403,
      "num_tokens": 7874846.0,
      "step": 3480
    },
    {
      "entropy": 5.67598671913147,
      "epoch": 0.22428884026258206,
      "grad_norm": 1.5,
      "learning_rate": 8.205650235642828e-05,
      "loss": 5.5609,
      "mean_token_accuracy": 0.18824456483125687,
      "num_tokens": 7885963.0,
      "step": 3485
    },
    {
      "entropy": 5.696454572677612,
      "epoch": 0.22461063199897027,
      "grad_norm": 0.95703125,
      "learning_rate": 8.145309014987978e-05,
      "loss": 5.546,
      "mean_token_accuracy": 0.18367184549570084,
      "num_tokens": 7896898.0,
      "step": 3490
    },
    {
      "entropy": 5.662345886230469,
      "epoch": 0.2249324237353585,
      "grad_norm": 1.078125,
      "learning_rate": 8.085498412963437e-05,
      "loss": 5.4528,
      "mean_token_accuracy": 0.18317982405424119,
      "num_tokens": 7907969.0,
      "step": 3495
    },
    {
      "entropy": 5.677160358428955,
      "epoch": 0.22525421547174668,
      "grad_norm": 0.98828125,
      "learning_rate": 8.026220069307078e-05,
      "loss": 5.3502,
      "mean_token_accuracy": 0.18638567477464676,
      "num_tokens": 7917357.0,
      "step": 3500
    }
  ],
  "logging_steps": 5,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2511960786944e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}