{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.390857527938451,
  "eval_steps": 3000,
  "global_step": 48000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 4.799633407592774,
      "epoch": 0.0005615769079575448,
      "grad_norm": 14.6875,
      "learning_rate": 2e-06,
      "loss": 14.5557,
      "mean_token_accuracy": 0.0,
      "num_tokens": 11659.0,
      "step": 5
    },
    {
      "entropy": 4.788970327377319,
      "epoch": 0.0011231538159150896,
      "grad_norm": 14.625,
      "learning_rate": 4.5e-06,
      "loss": 14.4346,
      "mean_token_accuracy": 0.0,
      "num_tokens": 21794.0,
      "step": 10
    },
    {
      "entropy": 4.845954370498657,
      "epoch": 0.0016847307238726343,
      "grad_norm": 16.5,
      "learning_rate": 7e-06,
      "loss": 14.2561,
      "mean_token_accuracy": 0.0,
      "num_tokens": 33206.0,
      "step": 15
    },
    {
      "entropy": 4.982930469512939,
      "epoch": 0.0022463076318301792,
      "grad_norm": 22.75,
      "learning_rate": 9.5e-06,
      "loss": 13.7895,
      "mean_token_accuracy": 9.009009227156639e-05,
      "num_tokens": 44487.0,
      "step": 20
    },
    {
      "entropy": 6.338765239715576,
      "epoch": 0.002807884539787724,
      "grad_norm": 28.25,
      "learning_rate": 1.2e-05,
      "loss": 12.4105,
      "mean_token_accuracy": 0.0,
      "num_tokens": 56062.0,
      "step": 25
    },
    {
      "entropy": 9.909645938873291,
      "epoch": 0.0033694614477452686,
      "grad_norm": 3.625,
      "learning_rate": 1.4500000000000002e-05,
      "loss": 10.9952,
      "mean_token_accuracy": 0.00023980815894901751,
      "num_tokens": 67453.0,
      "step": 30
    },
    {
      "entropy": 10.690417957305907,
      "epoch": 0.003931038355702813,
      "grad_norm": 3.203125,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 10.6677,
      "mean_token_accuracy": 0.012192816147580743,
      "num_tokens": 78527.0,
      "step": 35
    },
    {
      "entropy": 10.696582508087157,
      "epoch": 0.0044926152636603585,
      "grad_norm": 2.859375,
      "learning_rate": 1.95e-05,
      "loss": 10.3258,
      "mean_token_accuracy": 0.01598402885720134,
      "num_tokens": 89267.0,
      "step": 40
    },
    {
      "entropy": 10.660813236236573,
      "epoch": 0.005054192171617903,
      "grad_norm": 2.734375,
      "learning_rate": 2.2e-05,
      "loss": 9.9837,
      "mean_token_accuracy": 0.0142933520488441,
      "num_tokens": 102035.0,
      "step": 45
    },
    {
      "entropy": 10.618993759155273,
      "epoch": 0.005615769079575448,
      "grad_norm": 2.46875,
      "learning_rate": 2.4500000000000003e-05,
      "loss": 9.7405,
      "mean_token_accuracy": 0.03635566085577011,
      "num_tokens": 112734.0,
      "step": 50
    },
    {
      "entropy": 10.588664054870605,
      "epoch": 0.006177345987532993,
      "grad_norm": 2.171875,
      "learning_rate": 2.7e-05,
      "loss": 9.5631,
      "mean_token_accuracy": 0.03734342660754919,
      "num_tokens": 123854.0,
      "step": 55
    },
    {
      "entropy": 10.573542404174805,
      "epoch": 0.006738922895490537,
      "grad_norm": 2.140625,
      "learning_rate": 2.95e-05,
      "loss": 9.4769,
      "mean_token_accuracy": 0.036952446401119235,
      "num_tokens": 133444.0,
      "step": 60
    },
    {
      "entropy": 10.567419147491455,
      "epoch": 0.007300499803448083,
      "grad_norm": 2.25,
      "learning_rate": 3.2e-05,
      "loss": 9.3825,
      "mean_token_accuracy": 0.03926227204501629,
      "num_tokens": 144678.0,
      "step": 65
    },
    {
      "entropy": 10.560663509368897,
      "epoch": 0.007862076711405626,
      "grad_norm": 2.25,
      "learning_rate": 3.4500000000000005e-05,
      "loss": 9.3132,
      "mean_token_accuracy": 0.040116249769926074,
      "num_tokens": 155866.0,
      "step": 70
    },
    {
      "entropy": 10.53321475982666,
      "epoch": 0.008423653619363172,
      "grad_norm": 2.109375,
      "learning_rate": 3.7e-05,
      "loss": 9.2394,
      "mean_token_accuracy": 0.03705750666558742,
      "num_tokens": 166965.0,
      "step": 75
    },
    {
      "entropy": 10.505771923065186,
      "epoch": 0.008985230527320717,
      "grad_norm": 2.015625,
      "learning_rate": 3.95e-05,
      "loss": 9.148,
      "mean_token_accuracy": 0.038686181604862216,
      "num_tokens": 177423.0,
      "step": 80
    },
    {
      "entropy": 10.492299270629882,
      "epoch": 0.00954680743527826,
      "grad_norm": 1.9765625,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 9.0453,
      "mean_token_accuracy": 0.03904180806130171,
      "num_tokens": 188687.0,
      "step": 85
    },
    {
      "entropy": 10.44793472290039,
      "epoch": 0.010108384343235806,
      "grad_norm": 1.8828125,
      "learning_rate": 4.45e-05,
      "loss": 9.0036,
      "mean_token_accuracy": 0.03891269974410534,
      "num_tokens": 200173.0,
      "step": 90
    },
    {
      "entropy": 10.397690582275391,
      "epoch": 0.010669961251193351,
      "grad_norm": 1.9375,
      "learning_rate": 4.7000000000000004e-05,
      "loss": 8.877,
      "mean_token_accuracy": 0.04189660474658012,
      "num_tokens": 210877.0,
      "step": 95
    },
    {
      "entropy": 10.355337524414063,
      "epoch": 0.011231538159150897,
      "grad_norm": 1.859375,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 8.7438,
      "mean_token_accuracy": 0.04301192462444305,
      "num_tokens": 221770.0,
      "step": 100
    },
    {
      "entropy": 10.287545108795166,
      "epoch": 0.01179311506710844,
      "grad_norm": 1.765625,
      "learning_rate": 5.2e-05,
      "loss": 8.6419,
      "mean_token_accuracy": 0.03942227344959974,
      "num_tokens": 232140.0,
      "step": 105
    },
    {
      "entropy": 10.178714561462403,
      "epoch": 0.012354691975065986,
      "grad_norm": 1.5703125,
      "learning_rate": 5.45e-05,
      "loss": 8.5694,
      "mean_token_accuracy": 0.0399751354008913,
      "num_tokens": 244186.0,
      "step": 110
    },
    {
      "entropy": 10.056532287597657,
      "epoch": 0.012916268883023531,
      "grad_norm": 1.5234375,
      "learning_rate": 5.7e-05,
      "loss": 8.3694,
      "mean_token_accuracy": 0.04193062111735344,
      "num_tokens": 254805.0,
      "step": 115
    },
    {
      "entropy": 9.897430610656738,
      "epoch": 0.013477845790981074,
      "grad_norm": 1.4609375,
      "learning_rate": 5.9499999999999996e-05,
      "loss": 8.1667,
      "mean_token_accuracy": 0.05483008846640587,
      "num_tokens": 265055.0,
      "step": 120
    },
    {
      "entropy": 9.6635272026062,
      "epoch": 0.01403942269893862,
      "grad_norm": 1.34375,
      "learning_rate": 6.2e-05,
      "loss": 8.0303,
      "mean_token_accuracy": 0.05366372875869274,
      "num_tokens": 275850.0,
      "step": 125
    },
    {
      "entropy": 9.444734287261962,
      "epoch": 0.014600999606896165,
      "grad_norm": 1.125,
      "learning_rate": 6.450000000000001e-05,
      "loss": 7.9378,
      "mean_token_accuracy": 0.04993940331041813,
      "num_tokens": 286802.0,
      "step": 130
    },
    {
      "entropy": 9.225319862365723,
      "epoch": 0.015162576514853709,
      "grad_norm": 1.0625,
      "learning_rate": 6.7e-05,
      "loss": 7.8838,
      "mean_token_accuracy": 0.05432286374270916,
      "num_tokens": 297070.0,
      "step": 135
    },
    {
      "entropy": 8.997181510925293,
      "epoch": 0.015724153422811252,
      "grad_norm": 0.9921875,
      "learning_rate": 6.950000000000001e-05,
      "loss": 7.7723,
      "mean_token_accuracy": 0.049000997096300125,
      "num_tokens": 309617.0,
      "step": 140
    },
    {
      "entropy": 8.685122966766357,
      "epoch": 0.0162857303307688,
      "grad_norm": 0.953125,
      "learning_rate": 7.2e-05,
      "loss": 7.6868,
      "mean_token_accuracy": 0.05469490811228752,
      "num_tokens": 321310.0,
      "step": 145
    },
    {
      "entropy": 8.375407505035401,
      "epoch": 0.016847307238726343,
      "grad_norm": 0.8828125,
      "learning_rate": 7.45e-05,
      "loss": 7.5755,
      "mean_token_accuracy": 0.060717220231890676,
      "num_tokens": 332243.0,
      "step": 150
    },
    {
      "entropy": 8.259093475341796,
      "epoch": 0.017408884146683887,
      "grad_norm": 0.76171875,
      "learning_rate": 7.7e-05,
      "loss": 7.5338,
      "mean_token_accuracy": 0.054727142304182054,
      "num_tokens": 342847.0,
      "step": 155
    },
    {
      "entropy": 8.08473482131958,
      "epoch": 0.017970461054641434,
      "grad_norm": 0.97265625,
      "learning_rate": 7.950000000000001e-05,
      "loss": 7.4611,
      "mean_token_accuracy": 0.05515093542635441,
      "num_tokens": 353570.0,
      "step": 160
    },
    {
      "entropy": 7.938348913192749,
      "epoch": 0.018532037962598977,
      "grad_norm": 0.8828125,
      "learning_rate": 8.2e-05,
      "loss": 7.4358,
      "mean_token_accuracy": 0.06007346101105213,
      "num_tokens": 364330.0,
      "step": 165
    },
    {
      "entropy": 7.877144718170166,
      "epoch": 0.01909361487055652,
      "grad_norm": 1.265625,
      "learning_rate": 8.450000000000001e-05,
      "loss": 7.4554,
      "mean_token_accuracy": 0.0588134877383709,
      "num_tokens": 374930.0,
      "step": 170
    },
    {
      "entropy": 7.905826044082642,
      "epoch": 0.019655191778514068,
      "grad_norm": 1.15625,
      "learning_rate": 8.7e-05,
      "loss": 7.4192,
      "mean_token_accuracy": 0.06769459694623947,
      "num_tokens": 384690.0,
      "step": 175
    },
    {
      "entropy": 7.834839868545532,
      "epoch": 0.020216768686471612,
      "grad_norm": 1.7265625,
      "learning_rate": 8.95e-05,
      "loss": 7.4338,
      "mean_token_accuracy": 0.06734748221933842,
      "num_tokens": 395753.0,
      "step": 180
    },
    {
      "entropy": 7.7875371932983395,
      "epoch": 0.020778345594429155,
      "grad_norm": 1.8828125,
      "learning_rate": 9.2e-05,
      "loss": 7.3766,
      "mean_token_accuracy": 0.070906263589859,
      "num_tokens": 407049.0,
      "step": 185
    },
    {
      "entropy": 7.763374948501587,
      "epoch": 0.021339922502386702,
      "grad_norm": 1.6328125,
      "learning_rate": 9.45e-05,
      "loss": 7.3517,
      "mean_token_accuracy": 0.08063211217522621,
      "num_tokens": 417942.0,
      "step": 190
    },
    {
      "entropy": 7.7808509349823,
      "epoch": 0.021901499410344246,
      "grad_norm": 1.1796875,
      "learning_rate": 9.7e-05,
      "loss": 7.3703,
      "mean_token_accuracy": 0.0755468975752592,
      "num_tokens": 428151.0,
      "step": 195
    },
    {
      "entropy": 7.772652101516724,
      "epoch": 0.022463076318301793,
      "grad_norm": 1.0234375,
      "learning_rate": 9.95e-05,
      "loss": 7.4319,
      "mean_token_accuracy": 0.07449556142091751,
      "num_tokens": 439837.0,
      "step": 200
    },
    {
      "entropy": 7.81195764541626,
      "epoch": 0.023024653226259337,
      "grad_norm": 1.0625,
      "learning_rate": 0.000102,
      "loss": 7.2708,
      "mean_token_accuracy": 0.08094515725970268,
      "num_tokens": 450357.0,
      "step": 205
    },
    {
      "entropy": 7.7834734439849855,
      "epoch": 0.02358623013421688,
      "grad_norm": 1.1875,
      "learning_rate": 0.00010449999999999999,
      "loss": 7.259,
      "mean_token_accuracy": 0.08583413660526276,
      "num_tokens": 461174.0,
      "step": 210
    },
    {
      "entropy": 7.716898488998413,
      "epoch": 0.024147807042174427,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000107,
      "loss": 7.3163,
      "mean_token_accuracy": 0.08057258948683739,
      "num_tokens": 472785.0,
      "step": 215
    },
    {
      "entropy": 7.6768852233886715,
      "epoch": 0.02470938395013197,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0001095,
      "loss": 7.2606,
      "mean_token_accuracy": 0.0886429488658905,
      "num_tokens": 483248.0,
      "step": 220
    },
    {
      "entropy": 7.674954032897949,
      "epoch": 0.025270960858089515,
      "grad_norm": 1.046875,
      "learning_rate": 0.000112,
      "loss": 7.1547,
      "mean_token_accuracy": 0.08543359860777855,
      "num_tokens": 494690.0,
      "step": 225
    },
    {
      "entropy": 7.694804096221924,
      "epoch": 0.025832537766047062,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0001145,
      "loss": 7.3437,
      "mean_token_accuracy": 0.0837646782398224,
      "num_tokens": 506192.0,
      "step": 230
    },
    {
      "entropy": 7.681186056137085,
      "epoch": 0.026394114674004605,
      "grad_norm": 1.3125,
      "learning_rate": 0.00011700000000000001,
      "loss": 7.2056,
      "mean_token_accuracy": 0.08629952967166901,
      "num_tokens": 517152.0,
      "step": 235
    },
    {
      "entropy": 7.538256502151489,
      "epoch": 0.02695569158196215,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00011949999999999999,
      "loss": 7.1298,
      "mean_token_accuracy": 0.09271039292216302,
      "num_tokens": 527836.0,
      "step": 240
    },
    {
      "entropy": 7.712969779968262,
      "epoch": 0.027517268489919696,
      "grad_norm": 1.3125,
      "learning_rate": 0.000122,
      "loss": 7.2466,
      "mean_token_accuracy": 0.08624056577682496,
      "num_tokens": 539491.0,
      "step": 245
    },
    {
      "entropy": 7.584847354888916,
      "epoch": 0.02807884539787724,
      "grad_norm": 1.171875,
      "learning_rate": 0.0001245,
      "loss": 7.219,
      "mean_token_accuracy": 0.08856328502297402,
      "num_tokens": 550816.0,
      "step": 250
    },
    {
      "entropy": 7.612125205993652,
      "epoch": 0.028640422305834783,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000127,
      "loss": 7.249,
      "mean_token_accuracy": 0.08498690947890282,
      "num_tokens": 561566.0,
      "step": 255
    },
    {
      "entropy": 7.637190675735473,
      "epoch": 0.02920199921379233,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0001295,
      "loss": 7.1759,
      "mean_token_accuracy": 0.09114351570606231,
      "num_tokens": 573032.0,
      "step": 260
    },
    {
      "entropy": 7.649462938308716,
      "epoch": 0.029763576121749874,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000132,
      "loss": 7.1046,
      "mean_token_accuracy": 0.09311182051897049,
      "num_tokens": 584140.0,
      "step": 265
    },
    {
      "entropy": 7.557653856277466,
      "epoch": 0.030325153029707418,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00013450000000000002,
      "loss": 7.1759,
      "mean_token_accuracy": 0.09324203431606293,
      "num_tokens": 595097.0,
      "step": 270
    },
    {
      "entropy": 7.557107162475586,
      "epoch": 0.030886729937664965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00013700000000000002,
      "loss": 7.1836,
      "mean_token_accuracy": 0.09248357862234116,
      "num_tokens": 606797.0,
      "step": 275
    },
    {
      "entropy": 7.505317544937133,
      "epoch": 0.031448306845622505,
      "grad_norm": 1.15625,
      "learning_rate": 0.0001395,
      "loss": 7.1357,
      "mean_token_accuracy": 0.09568876847624778,
      "num_tokens": 618187.0,
      "step": 280
    },
    {
      "entropy": 7.587315082550049,
      "epoch": 0.032009883753580055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00014199999999999998,
      "loss": 7.0694,
      "mean_token_accuracy": 0.09733563140034676,
      "num_tokens": 629829.0,
      "step": 285
    },
    {
      "entropy": 7.454817771911621,
      "epoch": 0.0325714606615376,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0001445,
      "loss": 7.0751,
      "mean_token_accuracy": 0.09884163439273834,
      "num_tokens": 640914.0,
      "step": 290
    },
    {
      "entropy": 7.540893125534057,
      "epoch": 0.03313303756949514,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000147,
      "loss": 7.0809,
      "mean_token_accuracy": 0.09153685793280601,
      "num_tokens": 653057.0,
      "step": 295
    },
    {
      "entropy": 7.441942119598389,
      "epoch": 0.033694614477452686,
      "grad_norm": 1.0625,
      "learning_rate": 0.0001495,
      "loss": 7.0141,
      "mean_token_accuracy": 0.10246249213814736,
      "num_tokens": 663437.0,
      "step": 300
    },
    {
      "entropy": 7.427733898162842,
      "epoch": 0.03425619138541023,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000152,
      "loss": 7.0828,
      "mean_token_accuracy": 0.09505958333611489,
      "num_tokens": 675416.0,
      "step": 305
    },
    {
      "entropy": 7.327917146682739,
      "epoch": 0.03481776829336777,
      "grad_norm": 1.6015625,
      "learning_rate": 0.00015450000000000001,
      "loss": 6.8761,
      "mean_token_accuracy": 0.10240658894181251,
      "num_tokens": 686633.0,
      "step": 310
    },
    {
      "entropy": 7.421519470214844,
      "epoch": 0.035379345201325324,
      "grad_norm": 1.25,
      "learning_rate": 0.000157,
      "loss": 6.9702,
      "mean_token_accuracy": 0.10307649448513985,
      "num_tokens": 698065.0,
      "step": 315
    },
    {
      "entropy": 7.40921802520752,
      "epoch": 0.03594092210928287,
      "grad_norm": 1.8125,
      "learning_rate": 0.0001595,
      "loss": 6.9708,
      "mean_token_accuracy": 0.1034545697271824,
      "num_tokens": 708652.0,
      "step": 320
    },
    {
      "entropy": 7.3969964504241945,
      "epoch": 0.03650249901724041,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000162,
      "loss": 7.1225,
      "mean_token_accuracy": 0.09705277755856515,
      "num_tokens": 720199.0,
      "step": 325
    },
    {
      "entropy": 7.317982196807861,
      "epoch": 0.037064075925197955,
      "grad_norm": 1.5625,
      "learning_rate": 0.00016450000000000001,
      "loss": 7.1133,
      "mean_token_accuracy": 0.09536896646022797,
      "num_tokens": 732964.0,
      "step": 330
    },
    {
      "entropy": 7.344883298873901,
      "epoch": 0.0376256528331555,
      "grad_norm": 1.203125,
      "learning_rate": 0.00016700000000000002,
      "loss": 6.8741,
      "mean_token_accuracy": 0.10452906191349029,
      "num_tokens": 743528.0,
      "step": 335
    },
    {
      "entropy": 7.326264238357544,
      "epoch": 0.03818722974111304,
      "grad_norm": 1.421875,
      "learning_rate": 0.00016950000000000003,
      "loss": 6.9363,
      "mean_token_accuracy": 0.10507969111204148,
      "num_tokens": 754365.0,
      "step": 340
    },
    {
      "entropy": 7.219662475585937,
      "epoch": 0.03874880664907059,
      "grad_norm": 1.5546875,
      "learning_rate": 0.00017199999999999998,
      "loss": 6.8577,
      "mean_token_accuracy": 0.10085783675312995,
      "num_tokens": 766094.0,
      "step": 345
    },
    {
      "entropy": 7.265485715866089,
      "epoch": 0.039310383557028136,
      "grad_norm": 1.453125,
      "learning_rate": 0.00017449999999999999,
      "loss": 6.9116,
      "mean_token_accuracy": 0.10725476443767548,
      "num_tokens": 777275.0,
      "step": 350
    },
    {
      "entropy": 7.303876209259033,
      "epoch": 0.03987196046498568,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000177,
      "loss": 6.973,
      "mean_token_accuracy": 0.10475650802254677,
      "num_tokens": 789436.0,
      "step": 355
    },
    {
      "entropy": 7.291001510620117,
      "epoch": 0.040433537372943223,
      "grad_norm": 1.125,
      "learning_rate": 0.0001795,
      "loss": 6.988,
      "mean_token_accuracy": 0.10399318337440491,
      "num_tokens": 800164.0,
      "step": 360
    },
    {
      "entropy": 7.15556755065918,
      "epoch": 0.04099511428090077,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000182,
      "loss": 6.8667,
      "mean_token_accuracy": 0.10801982134580612,
      "num_tokens": 811797.0,
      "step": 365
    },
    {
      "entropy": 7.239162635803223,
      "epoch": 0.04155669118885831,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0001845,
      "loss": 6.864,
      "mean_token_accuracy": 0.1076484501361847,
      "num_tokens": 824158.0,
      "step": 370
    },
    {
      "entropy": 7.216827726364135,
      "epoch": 0.04211826809681586,
      "grad_norm": 1.046875,
      "learning_rate": 0.000187,
      "loss": 6.8616,
      "mean_token_accuracy": 0.1121969074010849,
      "num_tokens": 835803.0,
      "step": 375
    },
    {
      "entropy": 7.260309410095215,
      "epoch": 0.042679845004773405,
      "grad_norm": 1.125,
      "learning_rate": 0.0001895,
      "loss": 6.9411,
      "mean_token_accuracy": 0.10338237062096596,
      "num_tokens": 848106.0,
      "step": 380
    },
    {
      "entropy": 7.208122730255127,
      "epoch": 0.04324142191273095,
      "grad_norm": 1.15625,
      "learning_rate": 0.000192,
      "loss": 6.9009,
      "mean_token_accuracy": 0.10842320248484612,
      "num_tokens": 859101.0,
      "step": 385
    },
    {
      "entropy": 7.150190019607544,
      "epoch": 0.04380299882068849,
      "grad_norm": 1.25,
      "learning_rate": 0.0001945,
      "loss": 6.7826,
      "mean_token_accuracy": 0.1109951414167881,
      "num_tokens": 870309.0,
      "step": 390
    },
    {
      "entropy": 7.0933675289154055,
      "epoch": 0.044364575728646036,
      "grad_norm": 1.28125,
      "learning_rate": 0.00019700000000000002,
      "loss": 6.8608,
      "mean_token_accuracy": 0.10988331884145737,
      "num_tokens": 881515.0,
      "step": 395
    },
    {
      "entropy": 7.202591133117676,
      "epoch": 0.044926152636603586,
      "grad_norm": 1.359375,
      "learning_rate": 0.00019950000000000002,
      "loss": 6.8292,
      "mean_token_accuracy": 0.10998395457863808,
      "num_tokens": 892129.0,
      "step": 400
    },
    {
      "entropy": 7.136573028564453,
      "epoch": 0.04548772954456113,
      "grad_norm": 1.1875,
      "learning_rate": 0.000202,
      "loss": 6.8465,
      "mean_token_accuracy": 0.10337187349796295,
      "num_tokens": 903186.0,
      "step": 405
    },
    {
      "entropy": 7.1220067024230955,
      "epoch": 0.046049306452518673,
      "grad_norm": 1.6484375,
      "learning_rate": 0.00020449999999999998,
      "loss": 6.8047,
      "mean_token_accuracy": 0.11569347232580185,
      "num_tokens": 913471.0,
      "step": 410
    },
    {
      "entropy": 7.141632843017578,
      "epoch": 0.04661088336047622,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000207,
      "loss": 6.8051,
      "mean_token_accuracy": 0.10749553069472313,
      "num_tokens": 924283.0,
      "step": 415
    },
    {
      "entropy": 7.063813877105713,
      "epoch": 0.04717246026843376,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002095,
      "loss": 6.8408,
      "mean_token_accuracy": 0.11442757770419121,
      "num_tokens": 934975.0,
      "step": 420
    },
    {
      "entropy": 7.101827573776245,
      "epoch": 0.047734037176391304,
      "grad_norm": 1.8515625,
      "learning_rate": 0.000212,
      "loss": 6.705,
      "mean_token_accuracy": 0.11826195791363717,
      "num_tokens": 945587.0,
      "step": 425
    },
    {
      "entropy": 7.012417364120483,
      "epoch": 0.048295614084348855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002145,
      "loss": 6.7701,
      "mean_token_accuracy": 0.11151043251156807,
      "num_tokens": 955935.0,
      "step": 430
    },
    {
      "entropy": 7.108279705047607,
      "epoch": 0.0488571909923064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00021700000000000002,
      "loss": 6.7326,
      "mean_token_accuracy": 0.11344726905226707,
      "num_tokens": 967960.0,
      "step": 435
    },
    {
      "entropy": 7.063254356384277,
      "epoch": 0.04941876790026394,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002195,
      "loss": 6.7808,
      "mean_token_accuracy": 0.11375493034720421,
      "num_tokens": 979680.0,
      "step": 440
    },
    {
      "entropy": 7.0870774269104,
      "epoch": 0.049980344808221486,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000222,
      "loss": 6.8463,
      "mean_token_accuracy": 0.11385233402252197,
      "num_tokens": 990600.0,
      "step": 445
    },
    {
      "entropy": 7.130812168121338,
      "epoch": 0.05054192171617903,
      "grad_norm": 1.328125,
      "learning_rate": 0.0002245,
      "loss": 6.8797,
      "mean_token_accuracy": 0.11623142063617706,
      "num_tokens": 1002331.0,
      "step": 450
    },
    {
      "entropy": 6.960196018218994,
      "epoch": 0.05110349862413657,
      "grad_norm": 1.609375,
      "learning_rate": 0.00022700000000000002,
      "loss": 6.6839,
      "mean_token_accuracy": 0.11727181151509285,
      "num_tokens": 1013473.0,
      "step": 455
    },
    {
      "entropy": 7.0530540466308596,
      "epoch": 0.051665075532094124,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00022950000000000002,
      "loss": 6.7589,
      "mean_token_accuracy": 0.11695582792162895,
      "num_tokens": 1024434.0,
      "step": 460
    },
    {
      "entropy": 7.068104553222656,
      "epoch": 0.05222665244005167,
      "grad_norm": 1.09375,
      "learning_rate": 0.00023200000000000003,
      "loss": 6.7928,
      "mean_token_accuracy": 0.10744224488735199,
      "num_tokens": 1035452.0,
      "step": 465
    },
    {
      "entropy": 7.012246704101562,
      "epoch": 0.05278822934800921,
      "grad_norm": 1.1875,
      "learning_rate": 0.00023449999999999998,
      "loss": 6.6691,
      "mean_token_accuracy": 0.11236618980765342,
      "num_tokens": 1046213.0,
      "step": 470
    },
    {
      "entropy": 6.907792472839356,
      "epoch": 0.053349806255966754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000237,
      "loss": 6.69,
      "mean_token_accuracy": 0.11790019795298576,
      "num_tokens": 1057469.0,
      "step": 475
    },
    {
      "entropy": 7.105744743347168,
      "epoch": 0.0539113831639243,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0002395,
      "loss": 6.8301,
      "mean_token_accuracy": 0.11361000612378121,
      "num_tokens": 1069543.0,
      "step": 480
    },
    {
      "entropy": 6.806726551055908,
      "epoch": 0.05447296007188184,
      "grad_norm": 1.2734375,
      "learning_rate": 0.000242,
      "loss": 6.7071,
      "mean_token_accuracy": 0.11520676463842391,
      "num_tokens": 1080476.0,
      "step": 485
    },
    {
      "entropy": 7.111533737182617,
      "epoch": 0.05503453697983939,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002445,
      "loss": 6.8287,
      "mean_token_accuracy": 0.11254625618457795,
      "num_tokens": 1091700.0,
      "step": 490
    },
    {
      "entropy": 6.96989221572876,
      "epoch": 0.055596113887796936,
      "grad_norm": 1.078125,
      "learning_rate": 0.000247,
      "loss": 6.7132,
      "mean_token_accuracy": 0.11217723339796067,
      "num_tokens": 1103056.0,
      "step": 495
    },
    {
      "entropy": 6.942185974121093,
      "epoch": 0.05615769079575448,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002495,
      "loss": 6.7678,
      "mean_token_accuracy": 0.11992822587490082,
      "num_tokens": 1114836.0,
      "step": 500
    },
    {
      "entropy": 6.93104133605957,
      "epoch": 0.05671926770371202,
      "grad_norm": 1.328125,
      "learning_rate": 0.000252,
      "loss": 6.6666,
      "mean_token_accuracy": 0.11608512848615646,
      "num_tokens": 1125993.0,
      "step": 505
    },
    {
      "entropy": 7.019130229949951,
      "epoch": 0.05728084461166957,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002545,
      "loss": 6.7356,
      "mean_token_accuracy": 0.12077120542526246,
      "num_tokens": 1136719.0,
      "step": 510
    },
    {
      "entropy": 6.877520227432251,
      "epoch": 0.05784242151962711,
      "grad_norm": 1.359375,
      "learning_rate": 0.000257,
      "loss": 6.605,
      "mean_token_accuracy": 0.12040217965841293,
      "num_tokens": 1146947.0,
      "step": 515
    },
    {
      "entropy": 6.971859741210937,
      "epoch": 0.05840399842758466,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0002595,
      "loss": 6.736,
      "mean_token_accuracy": 0.11773492619395257,
      "num_tokens": 1157974.0,
      "step": 520
    },
    {
      "entropy": 6.8482812404632565,
      "epoch": 0.058965575335542204,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000262,
      "loss": 6.6668,
      "mean_token_accuracy": 0.11813957914710045,
      "num_tokens": 1168573.0,
      "step": 525
    },
    {
      "entropy": 6.854047727584839,
      "epoch": 0.05952715224349975,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026450000000000003,
      "loss": 6.5609,
      "mean_token_accuracy": 0.11992745697498322,
      "num_tokens": 1178446.0,
      "step": 530
    },
    {
      "entropy": 6.924405860900879,
      "epoch": 0.06008872915145729,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026700000000000004,
      "loss": 6.5966,
      "mean_token_accuracy": 0.12243785858154296,
      "num_tokens": 1190558.0,
      "step": 535
    },
    {
      "entropy": 6.876425790786743,
      "epoch": 0.060650306059414835,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00026950000000000005,
      "loss": 6.7667,
      "mean_token_accuracy": 0.11598537936806678,
      "num_tokens": 1201929.0,
      "step": 540
    },
    {
      "entropy": 6.789496803283692,
      "epoch": 0.06121188296737238,
      "grad_norm": 1.40625,
      "learning_rate": 0.00027200000000000005,
      "loss": 6.6055,
      "mean_token_accuracy": 0.12160162925720215,
      "num_tokens": 1213155.0,
      "step": 545
    },
    {
      "entropy": 6.950420570373535,
      "epoch": 0.06177345987532993,
      "grad_norm": 1.265625,
      "learning_rate": 0.0002745,
      "loss": 6.6558,
      "mean_token_accuracy": 0.12115484923124313,
      "num_tokens": 1223961.0,
      "step": 550
    },
    {
      "entropy": 6.740339994430542,
      "epoch": 0.06233503678328747,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000277,
      "loss": 6.5575,
      "mean_token_accuracy": 0.12820586189627647,
      "num_tokens": 1233888.0,
      "step": 555
    },
    {
      "entropy": 6.817170190811157,
      "epoch": 0.06289661369124501,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002795,
      "loss": 6.6744,
      "mean_token_accuracy": 0.11722169443964958,
      "num_tokens": 1244405.0,
      "step": 560
    },
    {
      "entropy": 6.862272119522094,
      "epoch": 0.06345819059920256,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028199999999999997,
      "loss": 6.6067,
      "mean_token_accuracy": 0.12411838248372078,
      "num_tokens": 1257099.0,
      "step": 565
    },
    {
      "entropy": 6.882076740264893,
      "epoch": 0.06401976750716011,
      "grad_norm": 1.3125,
      "learning_rate": 0.0002845,
      "loss": 6.7212,
      "mean_token_accuracy": 0.12341399341821671,
      "num_tokens": 1267965.0,
      "step": 570
    },
    {
      "entropy": 6.852020120620727,
      "epoch": 0.06458134441511765,
      "grad_norm": 1.4609375,
      "learning_rate": 0.000287,
      "loss": 6.6231,
      "mean_token_accuracy": 0.12452883943915367,
      "num_tokens": 1280075.0,
      "step": 575
    },
    {
      "entropy": 6.82808027267456,
      "epoch": 0.0651429213230752,
      "grad_norm": 1.28125,
      "learning_rate": 0.0002895,
      "loss": 6.6336,
      "mean_token_accuracy": 0.12447365745902061,
      "num_tokens": 1290574.0,
      "step": 580
    },
    {
      "entropy": 6.803577661514282,
      "epoch": 0.06570449823103273,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000292,
      "loss": 6.5981,
      "mean_token_accuracy": 0.12608856037259103,
      "num_tokens": 1302680.0,
      "step": 585
    },
    {
      "entropy": 6.73208703994751,
      "epoch": 0.06626607513899029,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002945,
      "loss": 6.591,
      "mean_token_accuracy": 0.12613771557807923,
      "num_tokens": 1314577.0,
      "step": 590
    },
    {
      "entropy": 6.796487188339233,
      "epoch": 0.06682765204694784,
      "grad_norm": 1.25,
      "learning_rate": 0.000297,
      "loss": 6.7192,
      "mean_token_accuracy": 0.12332894131541253,
      "num_tokens": 1326560.0,
      "step": 595
    },
    {
      "entropy": 6.990584993362427,
      "epoch": 0.06738922895490537,
      "grad_norm": 1.125,
      "learning_rate": 0.0002995,
      "loss": 6.7842,
      "mean_token_accuracy": 0.11380501464009285,
      "num_tokens": 1339105.0,
      "step": 600
    },
    {
      "entropy": 6.57991738319397,
      "epoch": 0.06795080586286292,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000302,
      "loss": 6.4443,
      "mean_token_accuracy": 0.13007400184869766,
      "num_tokens": 1349914.0,
      "step": 605
    },
    {
      "entropy": 6.823658180236817,
      "epoch": 0.06851238277082046,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003045,
      "loss": 6.6728,
      "mean_token_accuracy": 0.12215135470032693,
      "num_tokens": 1361700.0,
      "step": 610
    },
    {
      "entropy": 6.7814453125,
      "epoch": 0.06907395967877801,
      "grad_norm": 1.078125,
      "learning_rate": 0.000307,
      "loss": 6.6743,
      "mean_token_accuracy": 0.12424325123429299,
      "num_tokens": 1373103.0,
      "step": 615
    },
    {
      "entropy": 6.68872766494751,
      "epoch": 0.06963553658673555,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0003095,
      "loss": 6.4991,
      "mean_token_accuracy": 0.1282663404941559,
      "num_tokens": 1383742.0,
      "step": 620
    },
    {
      "entropy": 6.695285606384277,
      "epoch": 0.0701971134946931,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000312,
      "loss": 6.6344,
      "mean_token_accuracy": 0.1173564150929451,
      "num_tokens": 1396656.0,
      "step": 625
    },
    {
      "entropy": 6.7927330493927,
      "epoch": 0.07075869040265065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003145,
      "loss": 6.5057,
      "mean_token_accuracy": 0.13664172142744063,
      "num_tokens": 1406709.0,
      "step": 630
    },
    {
      "entropy": 6.760857725143433,
      "epoch": 0.07132026731060818,
      "grad_norm": 1.046875,
      "learning_rate": 0.000317,
      "loss": 6.5346,
      "mean_token_accuracy": 0.13074129149317743,
      "num_tokens": 1417793.0,
      "step": 635
    },
    {
      "entropy": 6.764607715606689,
      "epoch": 0.07188184421856574,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003195,
      "loss": 6.6279,
      "mean_token_accuracy": 0.12794213369488716,
      "num_tokens": 1429093.0,
      "step": 640
    },
    {
      "entropy": 6.65194787979126,
      "epoch": 0.07244342112652327,
      "grad_norm": 1.078125,
      "learning_rate": 0.000322,
      "loss": 6.5454,
      "mean_token_accuracy": 0.13121496513485909,
      "num_tokens": 1440907.0,
      "step": 645
    },
    {
      "entropy": 6.690778017044067,
      "epoch": 0.07300499803448082,
      "grad_norm": 1.15625,
      "learning_rate": 0.00032450000000000003,
      "loss": 6.6135,
      "mean_token_accuracy": 0.1266121596097946,
      "num_tokens": 1452089.0,
      "step": 650
    },
    {
      "entropy": 6.808844184875488,
      "epoch": 0.07356657494243837,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032700000000000003,
      "loss": 6.6337,
      "mean_token_accuracy": 0.1225229762494564,
      "num_tokens": 1464378.0,
      "step": 655
    },
    {
      "entropy": 6.704527997970581,
      "epoch": 0.07412815185039591,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00032950000000000004,
      "loss": 6.5677,
      "mean_token_accuracy": 0.12628930658102036,
      "num_tokens": 1476678.0,
      "step": 660
    },
    {
      "entropy": 6.637593460083008,
      "epoch": 0.07468972875835346,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00033200000000000005,
      "loss": 6.4052,
      "mean_token_accuracy": 0.13140960410237312,
      "num_tokens": 1486808.0,
      "step": 665
    },
    {
      "entropy": 6.757417011260986,
      "epoch": 0.075251305666311,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033450000000000005,
      "loss": 6.6957,
      "mean_token_accuracy": 0.12566189095377922,
      "num_tokens": 1499281.0,
      "step": 670
    },
    {
      "entropy": 6.700873708724975,
      "epoch": 0.07581288257426855,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000337,
      "loss": 6.5875,
      "mean_token_accuracy": 0.12513456344604493,
      "num_tokens": 1510202.0,
      "step": 675
    },
    {
      "entropy": 6.68131456375122,
      "epoch": 0.07637445948222608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003395,
      "loss": 6.4684,
      "mean_token_accuracy": 0.13036980405449866,
      "num_tokens": 1521492.0,
      "step": 680
    },
    {
      "entropy": 6.599414730072022,
      "epoch": 0.07693603639018363,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000342,
      "loss": 6.4806,
      "mean_token_accuracy": 0.13861154913902282,
      "num_tokens": 1532856.0,
      "step": 685
    },
    {
      "entropy": 6.614651203155518,
      "epoch": 0.07749761329814119,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034449999999999997,
      "loss": 6.4823,
      "mean_token_accuracy": 0.1309678554534912,
      "num_tokens": 1543408.0,
      "step": 690
    },
    {
      "entropy": 6.596418333053589,
      "epoch": 0.07805919020609872,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000347,
      "loss": 6.5167,
      "mean_token_accuracy": 0.12967425882816314,
      "num_tokens": 1555474.0,
      "step": 695
    },
    {
      "entropy": 6.689687299728393,
      "epoch": 0.07862076711405627,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003495,
      "loss": 6.6084,
      "mean_token_accuracy": 0.1304874062538147,
      "num_tokens": 1567484.0,
      "step": 700
    },
    {
      "entropy": 6.589553880691528,
      "epoch": 0.07918234402201381,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000352,
      "loss": 6.4484,
      "mean_token_accuracy": 0.13632451072335244,
      "num_tokens": 1578204.0,
      "step": 705
    },
    {
      "entropy": 6.66821985244751,
      "epoch": 0.07974392092997136,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0003545,
      "loss": 6.4521,
      "mean_token_accuracy": 0.13238313496112825,
      "num_tokens": 1589012.0,
      "step": 710
    },
    {
      "entropy": 6.522986221313476,
      "epoch": 0.08030549783792891,
      "grad_norm": 1.109375,
      "learning_rate": 0.000357,
      "loss": 6.4476,
      "mean_token_accuracy": 0.1307396411895752,
      "num_tokens": 1601099.0,
      "step": 715
    },
    {
      "entropy": 6.685805368423462,
      "epoch": 0.08086707474588645,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003595,
      "loss": 6.6162,
      "mean_token_accuracy": 0.12508113980293273,
      "num_tokens": 1612814.0,
      "step": 720
    },
    {
      "entropy": 6.628611469268799,
      "epoch": 0.081428651653844,
      "grad_norm": 1.140625,
      "learning_rate": 0.000362,
      "loss": 6.4967,
      "mean_token_accuracy": 0.1275397375226021,
      "num_tokens": 1624953.0,
      "step": 725
    },
    {
      "entropy": 6.6380833148956295,
      "epoch": 0.08199022856180153,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003645,
      "loss": 6.5367,
      "mean_token_accuracy": 0.13109428957104682,
      "num_tokens": 1635673.0,
      "step": 730
    },
    {
      "entropy": 6.542795658111572,
      "epoch": 0.08255180546975908,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000367,
      "loss": 6.4406,
      "mean_token_accuracy": 0.1324732281267643,
      "num_tokens": 1647282.0,
      "step": 735
    },
    {
      "entropy": 6.581196880340576,
      "epoch": 0.08311338237771662,
      "grad_norm": 1.296875,
      "learning_rate": 0.0003695,
      "loss": 6.4511,
      "mean_token_accuracy": 0.1369396723806858,
      "num_tokens": 1658426.0,
      "step": 740
    },
    {
      "entropy": 6.588415479660034,
      "epoch": 0.08367495928567417,
      "grad_norm": 1.109375,
      "learning_rate": 0.000372,
      "loss": 6.4821,
      "mean_token_accuracy": 0.13475994616746903,
      "num_tokens": 1669958.0,
      "step": 745
    },
    {
      "entropy": 6.565198183059692,
      "epoch": 0.08423653619363172,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003745,
      "loss": 6.5368,
      "mean_token_accuracy": 0.13080626279115676,
      "num_tokens": 1682409.0,
      "step": 750
    },
    {
      "entropy": 6.551465749740601,
      "epoch": 0.08479811310158926,
      "grad_norm": 1.15625,
      "learning_rate": 0.000377,
      "loss": 6.4584,
      "mean_token_accuracy": 0.13542011380195618,
      "num_tokens": 1692409.0,
      "step": 755
    },
    {
      "entropy": 6.6324629306793215,
      "epoch": 0.08535969000954681,
      "grad_norm": 1.125,
      "learning_rate": 0.0003795,
      "loss": 6.5569,
      "mean_token_accuracy": 0.12928430885076522,
      "num_tokens": 1705018.0,
      "step": 760
    },
    {
      "entropy": 6.529924488067627,
      "epoch": 0.08592126691750435,
      "grad_norm": 1.078125,
      "learning_rate": 0.000382,
      "loss": 6.4759,
      "mean_token_accuracy": 0.12947854325175284,
      "num_tokens": 1715625.0,
      "step": 765
    },
    {
      "entropy": 6.624479961395264,
      "epoch": 0.0864828438254619,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003845,
      "loss": 6.5534,
      "mean_token_accuracy": 0.12158236727118492,
      "num_tokens": 1727045.0,
      "step": 770
    },
    {
      "entropy": 6.64137897491455,
      "epoch": 0.08704442073341945,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038700000000000003,
      "loss": 6.6148,
      "mean_token_accuracy": 0.13262347802519797,
      "num_tokens": 1737374.0,
      "step": 775
    },
    {
      "entropy": 6.616215944290161,
      "epoch": 0.08760599764137698,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038950000000000003,
      "loss": 6.6163,
      "mean_token_accuracy": 0.13496674597263336,
      "num_tokens": 1747788.0,
      "step": 780
    },
    {
      "entropy": 6.421903228759765,
      "epoch": 0.08816757454933453,
      "grad_norm": 1.140625,
      "learning_rate": 0.00039200000000000004,
      "loss": 6.3084,
      "mean_token_accuracy": 0.14094914942979814,
      "num_tokens": 1757769.0,
      "step": 785
    },
    {
      "entropy": 6.548354005813598,
      "epoch": 0.08872915145729207,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039450000000000005,
      "loss": 6.4146,
      "mean_token_accuracy": 0.13216090872883796,
      "num_tokens": 1769246.0,
      "step": 790
    },
    {
      "entropy": 6.5623010158538815,
      "epoch": 0.08929072836524962,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039700000000000005,
      "loss": 6.554,
      "mean_token_accuracy": 0.12859445959329605,
      "num_tokens": 1781306.0,
      "step": 795
    },
    {
      "entropy": 6.527663707733154,
      "epoch": 0.08985230527320717,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003995,
      "loss": 6.4493,
      "mean_token_accuracy": 0.1310065895318985,
      "num_tokens": 1792888.0,
      "step": 800
    },
    {
      "entropy": 6.516381120681762,
      "epoch": 0.09041388218116471,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000402,
      "loss": 6.4648,
      "mean_token_accuracy": 0.13185214772820472,
      "num_tokens": 1802938.0,
      "step": 805
    },
    {
      "entropy": 6.583385372161866,
      "epoch": 0.09097545908912226,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004045,
      "loss": 6.4746,
      "mean_token_accuracy": 0.13560603633522988,
      "num_tokens": 1814738.0,
      "step": 810
    },
    {
      "entropy": 6.595263147354126,
      "epoch": 0.0915370359970798,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040699999999999997,
      "loss": 6.5531,
      "mean_token_accuracy": 0.1318923979997635,
      "num_tokens": 1825100.0,
      "step": 815
    },
    {
      "entropy": 6.3817685604095455,
      "epoch": 0.09209861290503735,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004095,
      "loss": 6.3926,
      "mean_token_accuracy": 0.13854454085230827,
      "num_tokens": 1835879.0,
      "step": 820
    },
    {
      "entropy": 6.663535404205322,
      "epoch": 0.09266018981299488,
      "grad_norm": 1.125,
      "learning_rate": 0.000412,
      "loss": 6.5323,
      "mean_token_accuracy": 0.12780267298221587,
      "num_tokens": 1848352.0,
      "step": 825
    },
    {
      "entropy": 6.477813053131103,
      "epoch": 0.09322176672095243,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004145,
      "loss": 6.4859,
      "mean_token_accuracy": 0.13047706484794616,
      "num_tokens": 1859687.0,
      "step": 830
    },
    {
      "entropy": 6.532577610015869,
      "epoch": 0.09378334362890998,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000417,
      "loss": 6.4749,
      "mean_token_accuracy": 0.13406161218881607,
      "num_tokens": 1871438.0,
      "step": 835
    },
    {
      "entropy": 6.455235624313355,
      "epoch": 0.09434492053686752,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004195,
      "loss": 6.3513,
      "mean_token_accuracy": 0.1425087571144104,
      "num_tokens": 1881519.0,
      "step": 840
    },
    {
      "entropy": 6.609141397476196,
      "epoch": 0.09490649744482507,
      "grad_norm": 1.03125,
      "learning_rate": 0.000422,
      "loss": 6.5414,
      "mean_token_accuracy": 0.1294293776154518,
      "num_tokens": 1892428.0,
      "step": 845
    },
    {
      "entropy": 6.550697469711304,
      "epoch": 0.09546807435278261,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004245,
      "loss": 6.4251,
      "mean_token_accuracy": 0.13496895208954812,
      "num_tokens": 1903760.0,
      "step": 850
    },
    {
      "entropy": 6.508064842224121,
      "epoch": 0.09602965126074016,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000427,
      "loss": 6.4953,
      "mean_token_accuracy": 0.12665094286203385,
      "num_tokens": 1915430.0,
      "step": 855
    },
    {
      "entropy": 6.478051090240479,
      "epoch": 0.09659122816869771,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004295,
      "loss": 6.4524,
      "mean_token_accuracy": 0.12965421453118325,
      "num_tokens": 1925709.0,
      "step": 860
    },
    {
      "entropy": 6.517416477203369,
      "epoch": 0.09715280507665525,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000432,
      "loss": 6.4937,
      "mean_token_accuracy": 0.13162111937999726,
      "num_tokens": 1938148.0,
      "step": 865
    },
    {
      "entropy": 6.323285913467407,
      "epoch": 0.0977143819846128,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004345,
      "loss": 6.3699,
      "mean_token_accuracy": 0.13734893202781678,
      "num_tokens": 1949028.0,
      "step": 870
    },
    {
      "entropy": 6.510279512405395,
      "epoch": 0.09827595889257033,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000437,
      "loss": 6.4184,
      "mean_token_accuracy": 0.13602244704961777,
      "num_tokens": 1960353.0,
      "step": 875
    },
    {
      "entropy": 6.504331731796265,
      "epoch": 0.09883753580052788,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004395,
      "loss": 6.5146,
      "mean_token_accuracy": 0.13366802632808686,
      "num_tokens": 1972335.0,
      "step": 880
    },
    {
      "entropy": 6.485915279388427,
      "epoch": 0.09939911270848542,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000442,
      "loss": 6.3532,
      "mean_token_accuracy": 0.1334136687219143,
      "num_tokens": 1982851.0,
      "step": 885
    },
    {
      "entropy": 6.433965969085693,
      "epoch": 0.09996068961644297,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004445,
      "loss": 6.3522,
      "mean_token_accuracy": 0.1376023381948471,
      "num_tokens": 1993931.0,
      "step": 890
    },
    {
      "entropy": 6.434508419036865,
      "epoch": 0.10052226652440052,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000447,
      "loss": 6.4802,
      "mean_token_accuracy": 0.13329798057675363,
      "num_tokens": 2004838.0,
      "step": 895
    },
    {
      "entropy": 6.3818906307220455,
      "epoch": 0.10108384343235806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044950000000000003,
      "loss": 6.2804,
      "mean_token_accuracy": 0.1390576109290123,
      "num_tokens": 2015620.0,
      "step": 900
    },
    {
      "entropy": 6.456867647171021,
      "epoch": 0.10164542034031561,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045200000000000004,
      "loss": 6.4092,
      "mean_token_accuracy": 0.1347552627325058,
      "num_tokens": 2027489.0,
      "step": 905
    },
    {
      "entropy": 6.344513940811157,
      "epoch": 0.10220699724827315,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045450000000000004,
      "loss": 6.2881,
      "mean_token_accuracy": 0.14329898431897165,
      "num_tokens": 2038475.0,
      "step": 910
    },
    {
      "entropy": 6.311835908889771,
      "epoch": 0.1027685741562307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045700000000000005,
      "loss": 6.2934,
      "mean_token_accuracy": 0.1361900806427002,
      "num_tokens": 2048938.0,
      "step": 915
    },
    {
      "entropy": 6.3546638011932375,
      "epoch": 0.10333015106418825,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045950000000000006,
      "loss": 6.3841,
      "mean_token_accuracy": 0.1331220120191574,
      "num_tokens": 2060372.0,
      "step": 920
    },
    {
      "entropy": 6.428075313568115,
      "epoch": 0.10389172797214578,
      "grad_norm": 1.03125,
      "learning_rate": 0.000462,
      "loss": 6.3677,
      "mean_token_accuracy": 0.13431201800704,
      "num_tokens": 2071558.0,
      "step": 925
    },
    {
      "entropy": 6.419263553619385,
      "epoch": 0.10445330488010333,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004645,
      "loss": 6.4142,
      "mean_token_accuracy": 0.13359070122241973,
      "num_tokens": 2082350.0,
      "step": 930
    },
    {
      "entropy": 6.386594343185425,
      "epoch": 0.10501488178806087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000467,
      "loss": 6.4635,
      "mean_token_accuracy": 0.13304929211735725,
      "num_tokens": 2093448.0,
      "step": 935
    },
    {
      "entropy": 6.608812284469605,
      "epoch": 0.10557645869601842,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004695,
      "loss": 6.5099,
      "mean_token_accuracy": 0.13017828315496444,
      "num_tokens": 2104171.0,
      "step": 940
    },
    {
      "entropy": 6.355282163619995,
      "epoch": 0.10613803560397596,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000472,
      "loss": 6.3814,
      "mean_token_accuracy": 0.13445304855704307,
      "num_tokens": 2114496.0,
      "step": 945
    },
    {
      "entropy": 6.452987957000732,
      "epoch": 0.10669961251193351,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004745,
      "loss": 6.379,
      "mean_token_accuracy": 0.13855226784944535,
      "num_tokens": 2125930.0,
      "step": 950
    },
    {
      "entropy": 6.360589504241943,
      "epoch": 0.10726118941989106,
      "grad_norm": 1.0625,
      "learning_rate": 0.000477,
      "loss": 6.3377,
      "mean_token_accuracy": 0.13913045451045036,
      "num_tokens": 2136452.0,
      "step": 955
    },
    {
      "entropy": 6.399038314819336,
      "epoch": 0.1078227663278486,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004795,
      "loss": 6.4486,
      "mean_token_accuracy": 0.13280693516135217,
      "num_tokens": 2148233.0,
      "step": 960
    },
    {
      "entropy": 6.3246174335479735,
      "epoch": 0.10838434323580615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000482,
      "loss": 6.2221,
      "mean_token_accuracy": 0.14445948973298073,
      "num_tokens": 2159475.0,
      "step": 965
    },
    {
      "entropy": 6.350338172912598,
      "epoch": 0.10894592014376368,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004845,
      "loss": 6.3454,
      "mean_token_accuracy": 0.13676348254084586,
      "num_tokens": 2170042.0,
      "step": 970
    },
    {
      "entropy": 6.358055448532104,
      "epoch": 0.10950749705172123,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000487,
      "loss": 6.3533,
      "mean_token_accuracy": 0.13602220565080642,
      "num_tokens": 2182234.0,
      "step": 975
    },
    {
      "entropy": 6.287888669967652,
      "epoch": 0.11006907395967878,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004895,
      "loss": 6.3348,
      "mean_token_accuracy": 0.13971530720591546,
      "num_tokens": 2192477.0,
      "step": 980
    },
    {
      "entropy": 6.348295879364014,
      "epoch": 0.11063065086763632,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000492,
      "loss": 6.3239,
      "mean_token_accuracy": 0.1356002353131771,
      "num_tokens": 2203082.0,
      "step": 985
    },
    {
      "entropy": 6.3176719665527346,
      "epoch": 0.11119222777559387,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945,
      "loss": 6.29,
      "mean_token_accuracy": 0.14067138880491256,
      "num_tokens": 2213684.0,
      "step": 990
    },
    {
      "entropy": 6.312493705749512,
      "epoch": 0.11175380468355141,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497,
      "loss": 6.3034,
      "mean_token_accuracy": 0.1373144119977951,
      "num_tokens": 2225524.0,
      "step": 995
    },
    {
      "entropy": 6.38683066368103,
      "epoch": 0.11231538159150896,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995,
      "loss": 6.2591,
      "mean_token_accuracy": 0.14032192155718803,
      "num_tokens": 2236004.0,
      "step": 1000
    },
    {
      "entropy": 6.363068675994873,
      "epoch": 0.1128769584994665,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499999997707492,
      "loss": 6.4203,
      "mean_token_accuracy": 0.12891897335648536,
      "num_tokens": 2247730.0,
      "step": 1005
    },
    {
      "entropy": 6.390230703353882,
      "epoch": 0.11343853540742405,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999999883941787,
      "loss": 6.3767,
      "mean_token_accuracy": 0.13985844999551772,
      "num_tokens": 2260132.0,
      "step": 1010
    },
    {
      "entropy": 6.254923486709595,
      "epoch": 0.1140001123153816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999999719167785,
      "loss": 6.3404,
      "mean_token_accuracy": 0.1435151532292366,
      "num_tokens": 2270060.0,
      "step": 1015
    },
    {
      "entropy": 6.382527685165405,
      "epoch": 0.11456168922333913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999999482752917,
      "loss": 6.3631,
      "mean_token_accuracy": 0.13998973742127419,
      "num_tokens": 2281568.0,
      "step": 1020
    },
    {
      "entropy": 6.354326391220093,
      "epoch": 0.11512326613129668,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999999174697195,
      "loss": 6.3815,
      "mean_token_accuracy": 0.13222115933895112,
      "num_tokens": 2293325.0,
      "step": 1025
    },
    {
      "entropy": 6.438601779937744,
      "epoch": 0.11568484303925422,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999998795000627,
      "loss": 6.3794,
      "mean_token_accuracy": 0.1392307199537754,
      "num_tokens": 2304780.0,
      "step": 1030
    },
    {
      "entropy": 6.276957988739014,
      "epoch": 0.11624641994721177,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999998343663223,
      "loss": 6.3166,
      "mean_token_accuracy": 0.1334988869726658,
      "num_tokens": 2315999.0,
      "step": 1035
    },
    {
      "entropy": 6.314444875717163,
      "epoch": 0.11680799685516932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999997820685,
      "loss": 6.4675,
      "mean_token_accuracy": 0.13674540892243386,
      "num_tokens": 2326695.0,
      "step": 1040
    },
    {
      "entropy": 6.406942319869995,
      "epoch": 0.11736957376312686,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999997226065974,
      "loss": 6.3016,
      "mean_token_accuracy": 0.1462261736392975,
      "num_tokens": 2336925.0,
      "step": 1045
    },
    {
      "entropy": 6.263730525970459,
      "epoch": 0.11793115067108441,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999996559806163,
      "loss": 6.1958,
      "mean_token_accuracy": 0.14441620260477067,
      "num_tokens": 2347431.0,
      "step": 1050
    },
    {
      "entropy": 6.244428634643555,
      "epoch": 0.11849272757904195,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999995821905589,
      "loss": 6.3172,
      "mean_token_accuracy": 0.1421760179102421,
      "num_tokens": 2359894.0,
      "step": 1055
    },
    {
      "entropy": 6.303659820556641,
      "epoch": 0.1190543044869995,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999995012364276,
      "loss": 6.337,
      "mean_token_accuracy": 0.13856007382273675,
      "num_tokens": 2371181.0,
      "step": 1060
    },
    {
      "entropy": 6.324774694442749,
      "epoch": 0.11961588139495703,
      "grad_norm": 1.25,
      "learning_rate": 0.0004999994131182248,
      "loss": 6.2517,
      "mean_token_accuracy": 0.1395735837519169,
      "num_tokens": 2381531.0,
      "step": 1065
    },
    {
      "entropy": 6.371315765380859,
      "epoch": 0.12017745830291458,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999993178359535,
      "loss": 6.3653,
      "mean_token_accuracy": 0.13924674913287163,
      "num_tokens": 2391573.0,
      "step": 1070
    },
    {
      "entropy": 6.3575451374053955,
      "epoch": 0.12073903521087213,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999992153896166,
      "loss": 6.3265,
      "mean_token_accuracy": 0.1408253751695156,
      "num_tokens": 2402822.0,
      "step": 1075
    },
    {
      "entropy": 6.25819411277771,
      "epoch": 0.12130061211882967,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999991057792175,
      "loss": 6.2388,
      "mean_token_accuracy": 0.14113450944423675,
      "num_tokens": 2414322.0,
      "step": 1080
    },
    {
      "entropy": 6.331667757034301,
      "epoch": 0.12186218902678722,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999989890047596,
      "loss": 6.4214,
      "mean_token_accuracy": 0.14067383036017417,
      "num_tokens": 2426560.0,
      "step": 1085
    },
    {
      "entropy": 6.356629800796509,
      "epoch": 0.12242376593474476,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004999988650662467,
      "loss": 6.3375,
      "mean_token_accuracy": 0.13867293000221254,
      "num_tokens": 2437723.0,
      "step": 1090
    },
    {
      "entropy": 6.203175830841064,
      "epoch": 0.12298534284270231,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004999987339636825,
      "loss": 6.2935,
      "mean_token_accuracy": 0.14306040480732918,
      "num_tokens": 2448497.0,
      "step": 1095
    },
    {
      "entropy": 6.248790836334228,
      "epoch": 0.12354691975065986,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999985956970713,
      "loss": 6.1472,
      "mean_token_accuracy": 0.15073004588484765,
      "num_tokens": 2459089.0,
      "step": 1100
    },
    {
      "entropy": 6.2033332824707035,
      "epoch": 0.1241084966586174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999984502664178,
      "loss": 6.2462,
      "mean_token_accuracy": 0.1456376627087593,
      "num_tokens": 2469882.0,
      "step": 1105
    },
    {
      "entropy": 6.280349397659302,
      "epoch": 0.12467007356657495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999982976717261,
      "loss": 6.294,
      "mean_token_accuracy": 0.1401812620460987,
      "num_tokens": 2482120.0,
      "step": 1110
    },
    {
      "entropy": 6.249101734161377,
      "epoch": 0.1252316504745325,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999981379130016,
      "loss": 6.2899,
      "mean_token_accuracy": 0.14231890588998794,
      "num_tokens": 2492626.0,
      "step": 1115
    },
    {
      "entropy": 6.41325273513794,
      "epoch": 0.12579322738249002,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499997970990249,
      "loss": 6.3502,
      "mean_token_accuracy": 0.13953726515173911,
      "num_tokens": 2506347.0,
      "step": 1120
    },
    {
      "entropy": 6.118116569519043,
      "epoch": 0.12635480429044757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999977969034737,
      "loss": 6.231,
      "mean_token_accuracy": 0.14573983773589133,
      "num_tokens": 2516584.0,
      "step": 1125
    },
    {
      "entropy": 6.293763732910156,
      "epoch": 0.12691638119840512,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999976156526814,
      "loss": 6.2808,
      "mean_token_accuracy": 0.14250900223851204,
      "num_tokens": 2526731.0,
      "step": 1130
    },
    {
      "entropy": 6.255453729629517,
      "epoch": 0.12747795810636267,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999974272378776,
      "loss": 6.1823,
      "mean_token_accuracy": 0.14683502316474914,
      "num_tokens": 2537180.0,
      "step": 1135
    },
    {
      "entropy": 6.250594902038574,
      "epoch": 0.12803953501432022,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999972316590685,
      "loss": 6.2542,
      "mean_token_accuracy": 0.14406149387359618,
      "num_tokens": 2547617.0,
      "step": 1140
    },
    {
      "entropy": 6.249514532089234,
      "epoch": 0.12860111192227774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999970289162604,
      "loss": 6.1485,
      "mean_token_accuracy": 0.15137399435043336,
      "num_tokens": 2557743.0,
      "step": 1145
    },
    {
      "entropy": 6.165639209747314,
      "epoch": 0.1291626888302353,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999968190094595,
      "loss": 6.2457,
      "mean_token_accuracy": 0.14083524644374848,
      "num_tokens": 2568351.0,
      "step": 1150
    },
    {
      "entropy": 6.251227760314942,
      "epoch": 0.12972426573819285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999966019386727,
      "loss": 6.2939,
      "mean_token_accuracy": 0.14459053725004195,
      "num_tokens": 2579703.0,
      "step": 1155
    },
    {
      "entropy": 6.3433530807495115,
      "epoch": 0.1302858426461504,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999963777039067,
      "loss": 6.2756,
      "mean_token_accuracy": 0.14344643652439118,
      "num_tokens": 2591197.0,
      "step": 1160
    },
    {
      "entropy": 6.212835693359375,
      "epoch": 0.13084741955410795,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499996146305169,
      "loss": 6.2425,
      "mean_token_accuracy": 0.14024166837334634,
      "num_tokens": 2601836.0,
      "step": 1165
    },
    {
      "entropy": 6.174261999130249,
      "epoch": 0.13140899646206547,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999959077424665,
      "loss": 6.1914,
      "mean_token_accuracy": 0.14582822620868682,
      "num_tokens": 2613513.0,
      "step": 1170
    },
    {
      "entropy": 6.215035343170166,
      "epoch": 0.13197057337002302,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999956620158072,
      "loss": 6.2188,
      "mean_token_accuracy": 0.1405838191509247,
      "num_tokens": 2625246.0,
      "step": 1175
    },
    {
      "entropy": 6.209453153610229,
      "epoch": 0.13253215027798057,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999954091251987,
      "loss": 6.1998,
      "mean_token_accuracy": 0.145721872150898,
      "num_tokens": 2636375.0,
      "step": 1180
    },
    {
      "entropy": 6.258368968963623,
      "epoch": 0.13309372718593812,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999951490706491,
      "loss": 6.2813,
      "mean_token_accuracy": 0.1400119759142399,
      "num_tokens": 2648586.0,
      "step": 1185
    },
    {
      "entropy": 6.228626394271851,
      "epoch": 0.13365530409389567,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999948818521667,
      "loss": 6.2922,
      "mean_token_accuracy": 0.14909495860338212,
      "num_tokens": 2660847.0,
      "step": 1190
    },
    {
      "entropy": 6.27654242515564,
      "epoch": 0.1342168810018532,
      "grad_norm": 0.953125,
      "learning_rate": 0.00049999460746976,
      "loss": 6.1995,
      "mean_token_accuracy": 0.15112631320953368,
      "num_tokens": 2671160.0,
      "step": 1195
    },
    {
      "entropy": 6.2761036396026615,
      "epoch": 0.13477845790981074,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004999943259234377,
      "loss": 6.3332,
      "mean_token_accuracy": 0.1402974992990494,
      "num_tokens": 2681622.0,
      "step": 1200
    },
    {
      "entropy": 6.138793230056763,
      "epoch": 0.1353400348177683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999940372132089,
      "loss": 6.1344,
      "mean_token_accuracy": 0.15093236565589904,
      "num_tokens": 2691503.0,
      "step": 1205
    },
    {
      "entropy": 6.240342044830323,
      "epoch": 0.13590161172572585,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999937413390826,
      "loss": 6.2271,
      "mean_token_accuracy": 0.1419438488781452,
      "num_tokens": 2702207.0,
      "step": 1210
    },
    {
      "entropy": 6.202216148376465,
      "epoch": 0.1364631886336834,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999934383010683,
      "loss": 6.2241,
      "mean_token_accuracy": 0.14322525262832642,
      "num_tokens": 2713178.0,
      "step": 1215
    },
    {
      "entropy": 6.297481489181519,
      "epoch": 0.13702476554164092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999931280991757,
      "loss": 6.2245,
      "mean_token_accuracy": 0.14976111873984338,
      "num_tokens": 2723717.0,
      "step": 1220
    },
    {
      "entropy": 6.216017818450927,
      "epoch": 0.13758634244959847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999928107334146,
      "loss": 6.2951,
      "mean_token_accuracy": 0.136559446901083,
      "num_tokens": 2735669.0,
      "step": 1225
    },
    {
      "entropy": 6.167074728012085,
      "epoch": 0.13814791935755602,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999924862037952,
      "loss": 6.1884,
      "mean_token_accuracy": 0.14789925068616866,
      "num_tokens": 2746087.0,
      "step": 1230
    },
    {
      "entropy": 6.147751665115356,
      "epoch": 0.13870949626551357,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999921545103278,
      "loss": 6.2034,
      "mean_token_accuracy": 0.14419680163264276,
      "num_tokens": 2757143.0,
      "step": 1235
    },
    {
      "entropy": 6.358134937286377,
      "epoch": 0.1392710731734711,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999918156530229,
      "loss": 6.3213,
      "mean_token_accuracy": 0.13962023183703423,
      "num_tokens": 2769340.0,
      "step": 1240
    },
    {
      "entropy": 6.162968301773072,
      "epoch": 0.13983265008142864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999914696318913,
      "loss": 6.2232,
      "mean_token_accuracy": 0.14598722010850906,
      "num_tokens": 2779514.0,
      "step": 1245
    },
    {
      "entropy": 6.283104515075683,
      "epoch": 0.1403942269893862,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004999911164469441,
      "loss": 6.1679,
      "mean_token_accuracy": 0.1477931559085846,
      "num_tokens": 2790801.0,
      "step": 1250
    },
    {
      "entropy": 6.126214218139649,
      "epoch": 0.14095580389734375,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999907560981925,
      "loss": 6.1673,
      "mean_token_accuracy": 0.14310122951865195,
      "num_tokens": 2801082.0,
      "step": 1255
    },
    {
      "entropy": 6.248990869522094,
      "epoch": 0.1415173808053013,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999903885856479,
      "loss": 6.1766,
      "mean_token_accuracy": 0.14449185878038406,
      "num_tokens": 2812719.0,
      "step": 1260
    },
    {
      "entropy": 6.088176202774048,
      "epoch": 0.14207895771325882,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000499990013909322,
      "loss": 6.1826,
      "mean_token_accuracy": 0.1402534618973732,
      "num_tokens": 2823766.0,
      "step": 1265
    },
    {
      "entropy": 6.18324728012085,
      "epoch": 0.14264053462121637,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499989632069227,
      "loss": 6.2595,
      "mean_token_accuracy": 0.14025890976190566,
      "num_tokens": 2834801.0,
      "step": 1270
    },
    {
      "entropy": 6.208938646316528,
      "epoch": 0.14320211152917392,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999892430653747,
      "loss": 6.075,
      "mean_token_accuracy": 0.14267505779862405,
      "num_tokens": 2845769.0,
      "step": 1275
    },
    {
      "entropy": 6.056669235229492,
      "epoch": 0.14376368843713147,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999888468977776,
      "loss": 6.1233,
      "mean_token_accuracy": 0.15102226734161378,
      "num_tokens": 2856001.0,
      "step": 1280
    },
    {
      "entropy": 6.231183815002441,
      "epoch": 0.14432526534508902,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999884435664485,
      "loss": 6.1527,
      "mean_token_accuracy": 0.1492895245552063,
      "num_tokens": 2867703.0,
      "step": 1285
    },
    {
      "entropy": 6.038867378234864,
      "epoch": 0.14488684225304654,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004999880330714,
      "loss": 6.1031,
      "mean_token_accuracy": 0.14959244281053544,
      "num_tokens": 2878671.0,
      "step": 1290
    },
    {
      "entropy": 6.206944847106934,
      "epoch": 0.1454484191610041,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004999876154126453,
      "loss": 6.218,
      "mean_token_accuracy": 0.139216335862875,
      "num_tokens": 2890137.0,
      "step": 1295
    },
    {
      "entropy": 6.149424123764038,
      "epoch": 0.14600999606896164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004999871905901977,
      "loss": 6.1208,
      "mean_token_accuracy": 0.14582841619849204,
      "num_tokens": 2900344.0,
      "step": 1300
    },
    {
      "entropy": 6.129654502868652,
      "epoch": 0.1465715729769192,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004999867586040706,
      "loss": 6.1297,
      "mean_token_accuracy": 0.1448382779955864,
      "num_tokens": 2912270.0,
      "step": 1305
    },
    {
      "entropy": 6.205135250091553,
      "epoch": 0.14713314988487675,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999863194542779,
      "loss": 6.2111,
      "mean_token_accuracy": 0.13794636726379395,
      "num_tokens": 2924677.0,
      "step": 1310
    },
    {
      "entropy": 6.095851850509644,
      "epoch": 0.14769472679283427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999858731408336,
      "loss": 6.1288,
      "mean_token_accuracy": 0.14932863563299179,
      "num_tokens": 2935794.0,
      "step": 1315
    },
    {
      "entropy": 6.201195335388183,
      "epoch": 0.14825630370079182,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004999854196637518,
      "loss": 6.2348,
      "mean_token_accuracy": 0.14271686151623725,
      "num_tokens": 2945958.0,
      "step": 1320
    },
    {
      "entropy": 6.227007293701172,
      "epoch": 0.14881788060874937,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999849590230469,
      "loss": 6.191,
      "mean_token_accuracy": 0.14383039027452468,
      "num_tokens": 2957536.0,
      "step": 1325
    },
    {
      "entropy": 6.134780406951904,
      "epoch": 0.14937945751670692,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999844912187336,
      "loss": 6.1875,
      "mean_token_accuracy": 0.1435810834169388,
      "num_tokens": 2967787.0,
      "step": 1330
    },
    {
      "entropy": 6.221927547454834,
      "epoch": 0.14994103442466447,
      "grad_norm": 1.0,
      "learning_rate": 0.000499984016250827,
      "loss": 6.2061,
      "mean_token_accuracy": 0.14361153468489646,
      "num_tokens": 2980522.0,
      "step": 1335
    },
    {
      "entropy": 6.042116403579712,
      "epoch": 0.150502611332622,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000499983534119342,
      "loss": 6.1357,
      "mean_token_accuracy": 0.14881859421730043,
      "num_tokens": 2990497.0,
      "step": 1340
    },
    {
      "entropy": 6.260757350921631,
      "epoch": 0.15106418824057954,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004999830448242941,
      "loss": 6.2239,
      "mean_token_accuracy": 0.14589421674609185,
      "num_tokens": 3002931.0,
      "step": 1345
    },
    {
      "entropy": 6.185908365249634,
      "epoch": 0.1516257651485371,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999825483656986,
      "loss": 6.1498,
      "mean_token_accuracy": 0.15437110215425492,
      "num_tokens": 3014137.0,
      "step": 1350
    },
    {
      "entropy": 6.2453696727752686,
      "epoch": 0.15218734205649465,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999820447435716,
      "loss": 6.3187,
      "mean_token_accuracy": 0.13783042505383492,
      "num_tokens": 3026145.0,
      "step": 1355
    },
    {
      "entropy": 6.102747440338135,
      "epoch": 0.15274891896445217,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999815339579291,
      "loss": 6.1756,
      "mean_token_accuracy": 0.14163123667240143,
      "num_tokens": 3037068.0,
      "step": 1360
    },
    {
      "entropy": 6.226761770248413,
      "epoch": 0.15331049587240972,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999810160087872,
      "loss": 6.0993,
      "mean_token_accuracy": 0.1480005696415901,
      "num_tokens": 3047836.0,
      "step": 1365
    },
    {
      "entropy": 6.135635757446289,
      "epoch": 0.15387207278036727,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004999804908961626,
      "loss": 6.2293,
      "mean_token_accuracy": 0.1483537584543228,
      "num_tokens": 3060309.0,
      "step": 1370
    },
    {
      "entropy": 6.175763988494873,
      "epoch": 0.15443364968832482,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999799586200719,
      "loss": 6.1913,
      "mean_token_accuracy": 0.15013613551855087,
      "num_tokens": 3071789.0,
      "step": 1375
    },
    {
      "entropy": 6.160604238510132,
      "epoch": 0.15499522659628237,
      "grad_norm": 1.109375,
      "learning_rate": 0.000499979419180532,
      "loss": 6.0957,
      "mean_token_accuracy": 0.15357528924942015,
      "num_tokens": 3081756.0,
      "step": 1380
    },
    {
      "entropy": 6.104126453399658,
      "epoch": 0.1555568035042399,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999788725775601,
      "loss": 6.2207,
      "mean_token_accuracy": 0.1406393028795719,
      "num_tokens": 3093456.0,
      "step": 1385
    },
    {
      "entropy": 6.2420367240905765,
      "epoch": 0.15611838041219744,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999783188111737,
      "loss": 6.1335,
      "mean_token_accuracy": 0.1524800807237625,
      "num_tokens": 3104565.0,
      "step": 1390
    },
    {
      "entropy": 6.143574810028076,
      "epoch": 0.156679957320155,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004999777578813904,
      "loss": 6.1554,
      "mean_token_accuracy": 0.14844489991664886,
      "num_tokens": 3115577.0,
      "step": 1395
    },
    {
      "entropy": 6.128584003448486,
      "epoch": 0.15724153422811254,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499977189788228,
      "loss": 6.0467,
      "mean_token_accuracy": 0.14999896585941314,
      "num_tokens": 3126019.0,
      "step": 1400
    },
    {
      "entropy": 6.168298864364624,
      "epoch": 0.1578031111360701,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999766145317046,
      "loss": 6.2184,
      "mean_token_accuracy": 0.13864450231194497,
      "num_tokens": 3136191.0,
      "step": 1405
    },
    {
      "entropy": 6.245380020141601,
      "epoch": 0.15836468804402762,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999760321118385,
      "loss": 6.0875,
      "mean_token_accuracy": 0.1503506787121296,
      "num_tokens": 3147737.0,
      "step": 1410
    },
    {
      "entropy": 6.0368874073028564,
      "epoch": 0.15892626495198517,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999754425286483,
      "loss": 6.1608,
      "mean_token_accuracy": 0.14620319604873658,
      "num_tokens": 3158494.0,
      "step": 1415
    },
    {
      "entropy": 6.1492870330810545,
      "epoch": 0.15948784185994272,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999748457821529,
      "loss": 6.1833,
      "mean_token_accuracy": 0.1447380617260933,
      "num_tokens": 3170550.0,
      "step": 1420
    },
    {
      "entropy": 6.0979039669036865,
      "epoch": 0.16004941876790027,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499974241872371,
      "loss": 6.1609,
      "mean_token_accuracy": 0.15229749754071237,
      "num_tokens": 3182860.0,
      "step": 1425
    },
    {
      "entropy": 6.11085262298584,
      "epoch": 0.16061099567585782,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999736307993221,
      "loss": 6.1698,
      "mean_token_accuracy": 0.1435732513666153,
      "num_tokens": 3193807.0,
      "step": 1430
    },
    {
      "entropy": 6.072397756576538,
      "epoch": 0.16117257258381534,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999730125630255,
      "loss": 6.0308,
      "mean_token_accuracy": 0.15132891982793809,
      "num_tokens": 3203959.0,
      "step": 1435
    },
    {
      "entropy": 6.119855213165283,
      "epoch": 0.1617341494917729,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999723871635008,
      "loss": 6.0406,
      "mean_token_accuracy": 0.15461250096559526,
      "num_tokens": 3215416.0,
      "step": 1440
    },
    {
      "entropy": 6.184870100021362,
      "epoch": 0.16229572639973044,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999717546007681,
      "loss": 6.1993,
      "mean_token_accuracy": 0.14106006175279617,
      "num_tokens": 3226938.0,
      "step": 1445
    },
    {
      "entropy": 5.983025455474854,
      "epoch": 0.162857303307688,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999711148748476,
      "loss": 6.0115,
      "mean_token_accuracy": 0.15607472509145737,
      "num_tokens": 3238406.0,
      "step": 1450
    },
    {
      "entropy": 6.070747804641724,
      "epoch": 0.16341888021564555,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004999704679857594,
      "loss": 6.0323,
      "mean_token_accuracy": 0.14664915949106216,
      "num_tokens": 3249695.0,
      "step": 1455
    },
    {
      "entropy": 6.213888835906983,
      "epoch": 0.16398045712360307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999698139335243,
      "loss": 6.1773,
      "mean_token_accuracy": 0.15213444456458092,
      "num_tokens": 3259656.0,
      "step": 1460
    },
    {
      "entropy": 6.050896024703979,
      "epoch": 0.16454203403156062,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004999691527181632,
      "loss": 6.015,
      "mean_token_accuracy": 0.1554547891020775,
      "num_tokens": 3269712.0,
      "step": 1465
    },
    {
      "entropy": 5.951676607131958,
      "epoch": 0.16510361093951817,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999684843396968,
      "loss": 6.0044,
      "mean_token_accuracy": 0.1552644908428192,
      "num_tokens": 3280343.0,
      "step": 1470
    },
    {
      "entropy": 6.096579122543335,
      "epoch": 0.16566518784747572,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999678087981467,
      "loss": 6.0393,
      "mean_token_accuracy": 0.15125902444124223,
      "num_tokens": 3291649.0,
      "step": 1475
    },
    {
      "entropy": 6.175355672836304,
      "epoch": 0.16622676475543324,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004999671260935343,
      "loss": 6.2151,
      "mean_token_accuracy": 0.14770857915282248,
      "num_tokens": 3304190.0,
      "step": 1480
    },
    {
      "entropy": 6.013831090927124,
      "epoch": 0.1667883416633908,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999664362258814,
      "loss": 5.9392,
      "mean_token_accuracy": 0.16093746125698088,
      "num_tokens": 3313966.0,
      "step": 1485
    },
    {
      "entropy": 6.109157514572144,
      "epoch": 0.16734991857134834,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999657391952099,
      "loss": 6.1968,
      "mean_token_accuracy": 0.1453491024672985,
      "num_tokens": 3326017.0,
      "step": 1490
    },
    {
      "entropy": 6.128896427154541,
      "epoch": 0.1679114954793059,
      "grad_norm": 0.88671875,
      "learning_rate": 0.000499965035001542,
      "loss": 6.0754,
      "mean_token_accuracy": 0.14839625358581543,
      "num_tokens": 3338264.0,
      "step": 1495
    },
    {
      "entropy": 6.084127426147461,
      "epoch": 0.16847307238726345,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999643236449001,
      "loss": 6.1332,
      "mean_token_accuracy": 0.1588374301791191,
      "num_tokens": 3348901.0,
      "step": 1500
    },
    {
      "entropy": 6.147513389587402,
      "epoch": 0.16903464929522097,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999636051253069,
      "loss": 6.0628,
      "mean_token_accuracy": 0.1494102582335472,
      "num_tokens": 3358441.0,
      "step": 1505
    },
    {
      "entropy": 6.085050630569458,
      "epoch": 0.16959622620317852,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999628794427854,
      "loss": 6.147,
      "mean_token_accuracy": 0.1559033215045929,
      "num_tokens": 3369334.0,
      "step": 1510
    },
    {
      "entropy": 6.1317285060882565,
      "epoch": 0.17015780311113607,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004999621465973583,
      "loss": 6.1442,
      "mean_token_accuracy": 0.1466472938656807,
      "num_tokens": 3380698.0,
      "step": 1515
    },
    {
      "entropy": 6.239275121688843,
      "epoch": 0.17071938001909362,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004999614065890495,
      "loss": 6.1503,
      "mean_token_accuracy": 0.14520846977829932,
      "num_tokens": 3391061.0,
      "step": 1520
    },
    {
      "entropy": 6.090221786499024,
      "epoch": 0.17128095692705117,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999606594178821,
      "loss": 6.2056,
      "mean_token_accuracy": 0.1438593976199627,
      "num_tokens": 3403983.0,
      "step": 1525
    },
    {
      "entropy": 6.175626039505005,
      "epoch": 0.1718425338350087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00049995990508388,
      "loss": 6.1397,
      "mean_token_accuracy": 0.15281769931316375,
      "num_tokens": 3413838.0,
      "step": 1530
    },
    {
      "entropy": 5.981231355667115,
      "epoch": 0.17240411074296624,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004999591435870674,
      "loss": 5.961,
      "mean_token_accuracy": 0.15355922728776933,
      "num_tokens": 3423978.0,
      "step": 1535
    },
    {
      "entropy": 6.136894989013672,
      "epoch": 0.1729656876509238,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004999583749274684,
      "loss": 6.1312,
      "mean_token_accuracy": 0.15456196144223214,
      "num_tokens": 3433416.0,
      "step": 1540
    },
    {
      "entropy": 5.982147359848023,
      "epoch": 0.17352726455888134,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004999575991051074,
      "loss": 5.9838,
      "mean_token_accuracy": 0.1558251142501831,
      "num_tokens": 3444983.0,
      "step": 1545
    },
    {
      "entropy": 6.021946763992309,
      "epoch": 0.1740888414668389,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999568161200093,
      "loss": 5.9757,
      "mean_token_accuracy": 0.156097112596035,
      "num_tokens": 3455808.0,
      "step": 1550
    },
    {
      "entropy": 5.985035371780396,
      "epoch": 0.17465041837479642,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999560259721989,
      "loss": 6.1258,
      "mean_token_accuracy": 0.14893127232789993,
      "num_tokens": 3467503.0,
      "step": 1555
    },
    {
      "entropy": 6.338688659667969,
      "epoch": 0.17521199528275397,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999552286617014,
      "loss": 6.3017,
      "mean_token_accuracy": 0.14476528465747834,
      "num_tokens": 3479941.0,
      "step": 1560
    },
    {
      "entropy": 6.006549453735351,
      "epoch": 0.17577357219071152,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999544241885421,
      "loss": 5.9989,
      "mean_token_accuracy": 0.16019308269023896,
      "num_tokens": 3491069.0,
      "step": 1565
    },
    {
      "entropy": 6.127226495742798,
      "epoch": 0.17633514909866907,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004999536125527467,
      "loss": 6.1753,
      "mean_token_accuracy": 0.14448272734880446,
      "num_tokens": 3502065.0,
      "step": 1570
    },
    {
      "entropy": 6.176832151412964,
      "epoch": 0.17689672600662662,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499952793754341,
      "loss": 6.1282,
      "mean_token_accuracy": 0.15281856656074524,
      "num_tokens": 3514228.0,
      "step": 1575
    },
    {
      "entropy": 6.0846302032470705,
      "epoch": 0.17745830291458414,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999519677933512,
      "loss": 6.0535,
      "mean_token_accuracy": 0.1507752999663353,
      "num_tokens": 3525801.0,
      "step": 1580
    },
    {
      "entropy": 6.053277921676636,
      "epoch": 0.1780198798225417,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999511346698035,
      "loss": 5.9995,
      "mean_token_accuracy": 0.15059795528650283,
      "num_tokens": 3536431.0,
      "step": 1585
    },
    {
      "entropy": 5.99881501197815,
      "epoch": 0.17858145673049924,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004999502943837244,
      "loss": 6.1027,
      "mean_token_accuracy": 0.14881094694137573,
      "num_tokens": 3549098.0,
      "step": 1590
    },
    {
      "entropy": 6.1661224365234375,
      "epoch": 0.1791430336384568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999494469351406,
      "loss": 6.0682,
      "mean_token_accuracy": 0.14940569475293158,
      "num_tokens": 3560532.0,
      "step": 1595
    },
    {
      "entropy": 6.003274202346802,
      "epoch": 0.17970461054641435,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999485923240793,
      "loss": 6.0716,
      "mean_token_accuracy": 0.14933864623308182,
      "num_tokens": 3572175.0,
      "step": 1600
    },
    {
      "entropy": 6.1079789161682125,
      "epoch": 0.18026618745437187,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004999477305505675,
      "loss": 5.9988,
      "mean_token_accuracy": 0.1632366694509983,
      "num_tokens": 3582862.0,
      "step": 1605
    },
    {
      "entropy": 6.1130759716033936,
      "epoch": 0.18082776436232942,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004999468616146328,
      "loss": 6.1636,
      "mean_token_accuracy": 0.15272294878959655,
      "num_tokens": 3594685.0,
      "step": 1610
    },
    {
      "entropy": 5.947771835327148,
      "epoch": 0.18138934127028697,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004999459855163026,
      "loss": 5.9689,
      "mean_token_accuracy": 0.1606626659631729,
      "num_tokens": 3604847.0,
      "step": 1615
    },
    {
      "entropy": 6.14026780128479,
      "epoch": 0.18195091817824452,
      "grad_norm": 0.953125,
      "learning_rate": 0.000499945102255605,
      "loss": 6.0664,
      "mean_token_accuracy": 0.15268882662057875,
      "num_tokens": 3615549.0,
      "step": 1620
    },
    {
      "entropy": 5.954753828048706,
      "epoch": 0.18251249508620204,
      "grad_norm": 1.0,
      "learning_rate": 0.0004999442118325682,
      "loss": 5.967,
      "mean_token_accuracy": 0.16526584029197694,
      "num_tokens": 3626300.0,
      "step": 1625
    },
    {
      "entropy": 6.00643219947815,
      "epoch": 0.1830740719941596,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004999433142472205,
      "loss": 5.9874,
      "mean_token_accuracy": 0.15639984607696533,
      "num_tokens": 3637832.0,
      "step": 1630
    },
    {
      "entropy": 6.080643796920777,
      "epoch": 0.18363564890211714,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004999424094995902,
      "loss": 6.0342,
      "mean_token_accuracy": 0.14727310091257095,
      "num_tokens": 3649179.0,
      "step": 1635
    },
    {
      "entropy": 6.008195066452027,
      "epoch": 0.1841972258100747,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999414975897066,
      "loss": 6.0708,
      "mean_token_accuracy": 0.15119689106941223,
      "num_tokens": 3660011.0,
      "step": 1640
    },
    {
      "entropy": 6.162057495117187,
      "epoch": 0.18475880271803224,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999405785175983,
      "loss": 6.1501,
      "mean_token_accuracy": 0.14309106916189193,
      "num_tokens": 3672791.0,
      "step": 1645
    },
    {
      "entropy": 5.958169221878052,
      "epoch": 0.18532037962598977,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004999396522832949,
      "loss": 5.9647,
      "mean_token_accuracy": 0.14754443913698195,
      "num_tokens": 3684709.0,
      "step": 1650
    },
    {
      "entropy": 5.964678478240967,
      "epoch": 0.18588195653394732,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004999387188868256,
      "loss": 5.9426,
      "mean_token_accuracy": 0.1558021366596222,
      "num_tokens": 3694675.0,
      "step": 1655
    },
    {
      "entropy": 5.890969276428223,
      "epoch": 0.18644353344190487,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999377783282203,
      "loss": 5.9461,
      "mean_token_accuracy": 0.15189632922410964,
      "num_tokens": 3706384.0,
      "step": 1660
    },
    {
      "entropy": 6.1524893283844,
      "epoch": 0.18700511034986242,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004999368306075089,
      "loss": 6.0175,
      "mean_token_accuracy": 0.1526872456073761,
      "num_tokens": 3717419.0,
      "step": 1665
    },
    {
      "entropy": 6.060198020935059,
      "epoch": 0.18756668725781997,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004999358757247215,
      "loss": 6.0517,
      "mean_token_accuracy": 0.1526139944791794,
      "num_tokens": 3727306.0,
      "step": 1670
    },
    {
      "entropy": 6.124920654296875,
      "epoch": 0.1881282641657775,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999349136798886,
      "loss": 6.1765,
      "mean_token_accuracy": 0.14685777351260185,
      "num_tokens": 3739476.0,
      "step": 1675
    },
    {
      "entropy": 6.046180200576782,
      "epoch": 0.18868984107373504,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004999339444730409,
      "loss": 5.9431,
      "mean_token_accuracy": 0.15702874660491944,
      "num_tokens": 3751035.0,
      "step": 1680
    },
    {
      "entropy": 6.048432970046997,
      "epoch": 0.1892514179816926,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999329681042092,
      "loss": 6.0281,
      "mean_token_accuracy": 0.1569201871752739,
      "num_tokens": 3762881.0,
      "step": 1685
    },
    {
      "entropy": 6.009223794937133,
      "epoch": 0.18981299488965014,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999319845734243,
      "loss": 6.0435,
      "mean_token_accuracy": 0.15313177406787873,
      "num_tokens": 3774180.0,
      "step": 1690
    },
    {
      "entropy": 6.0793358325958256,
      "epoch": 0.1903745717976077,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999309938807181,
      "loss": 6.0413,
      "mean_token_accuracy": 0.15854722559452056,
      "num_tokens": 3784864.0,
      "step": 1695
    },
    {
      "entropy": 6.063125991821289,
      "epoch": 0.19093614870556522,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004999299960261216,
      "loss": 6.0034,
      "mean_token_accuracy": 0.15233902782201766,
      "num_tokens": 3797693.0,
      "step": 1700
    },
    {
      "entropy": 6.026073360443116,
      "epoch": 0.19149772561352277,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004999289910096669,
      "loss": 6.0425,
      "mean_token_accuracy": 0.1526545524597168,
      "num_tokens": 3809379.0,
      "step": 1705
    },
    {
      "entropy": 5.922689151763916,
      "epoch": 0.19205930252148032,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999279788313858,
      "loss": 5.8735,
      "mean_token_accuracy": 0.1640583574771881,
      "num_tokens": 3819473.0,
      "step": 1710
    },
    {
      "entropy": 5.850899267196655,
      "epoch": 0.19262087942943787,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004999269594913107,
      "loss": 5.8711,
      "mean_token_accuracy": 0.16054081916809082,
      "num_tokens": 3831037.0,
      "step": 1715
    },
    {
      "entropy": 5.991314029693603,
      "epoch": 0.19318245633739542,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004999259329894739,
      "loss": 6.018,
      "mean_token_accuracy": 0.151178976893425,
      "num_tokens": 3841767.0,
      "step": 1720
    },
    {
      "entropy": 5.9119984149932865,
      "epoch": 0.19374403324535294,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004999248993259081,
      "loss": 5.8597,
      "mean_token_accuracy": 0.16238378435373307,
      "num_tokens": 3851629.0,
      "step": 1725
    },
    {
      "entropy": 5.94813723564148,
      "epoch": 0.1943056101533105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004999238585006463,
      "loss": 5.94,
      "mean_token_accuracy": 0.15244577899575235,
      "num_tokens": 3863247.0,
      "step": 1730
    },
    {
      "entropy": 6.0112062931060795,
      "epoch": 0.19486718706126804,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999228105137218,
      "loss": 5.9176,
      "mean_token_accuracy": 0.15697233527898788,
      "num_tokens": 3873949.0,
      "step": 1735
    },
    {
      "entropy": 5.9939721584320065,
      "epoch": 0.1954287639692256,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004999217553651676,
      "loss": 6.0223,
      "mean_token_accuracy": 0.15648642033338547,
      "num_tokens": 3884992.0,
      "step": 1740
    },
    {
      "entropy": 6.029172420501709,
      "epoch": 0.19599034087718312,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004999206930550175,
      "loss": 6.1497,
      "mean_token_accuracy": 0.158347637206316,
      "num_tokens": 3897029.0,
      "step": 1745
    },
    {
      "entropy": 6.016314172744751,
      "epoch": 0.19655191778514067,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004999196235833054,
      "loss": 6.001,
      "mean_token_accuracy": 0.15252462327480315,
      "num_tokens": 3909510.0,
      "step": 1750
    },
    {
      "entropy": 5.9791820526123045,
      "epoch": 0.19711349469309822,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004999185469500651,
      "loss": 5.8895,
      "mean_token_accuracy": 0.15988631993532182,
      "num_tokens": 3920593.0,
      "step": 1755
    },
    {
      "entropy": 6.043954753875733,
      "epoch": 0.19767507160105577,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000499917463155331,
      "loss": 6.0383,
      "mean_token_accuracy": 0.15022043734788895,
      "num_tokens": 3932215.0,
      "step": 1760
    },
    {
      "entropy": 5.971577739715576,
      "epoch": 0.19823664850901332,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004999163721991378,
      "loss": 5.9892,
      "mean_token_accuracy": 0.15806906223297118,
      "num_tokens": 3943876.0,
      "step": 1765
    },
    {
      "entropy": 5.915837717056275,
      "epoch": 0.19879822541697084,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049991527408152,
      "loss": 5.9785,
      "mean_token_accuracy": 0.1582863599061966,
      "num_tokens": 3954533.0,
      "step": 1770
    },
    {
      "entropy": 6.1025553226470945,
      "epoch": 0.1993598023249284,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999141688025127,
      "loss": 6.0337,
      "mean_token_accuracy": 0.15851334780454635,
      "num_tokens": 3965722.0,
      "step": 1775
    },
    {
      "entropy": 5.983110094070435,
      "epoch": 0.19992137923288594,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499913056362151,
      "loss": 5.9503,
      "mean_token_accuracy": 0.1551786482334137,
      "num_tokens": 3976737.0,
      "step": 1780
    },
    {
      "entropy": 6.036009645462036,
      "epoch": 0.2004829561408435,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004999119367604703,
      "loss": 6.0433,
      "mean_token_accuracy": 0.15762924998998643,
      "num_tokens": 3986102.0,
      "step": 1785
    },
    {
      "entropy": 6.072134399414063,
      "epoch": 0.20104453304880104,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004999108099975064,
      "loss": 6.0061,
      "mean_token_accuracy": 0.1530447334051132,
      "num_tokens": 3996534.0,
      "step": 1790
    },
    {
      "entropy": 5.904342555999756,
      "epoch": 0.20160610995675857,
      "grad_norm": 1.0,
      "learning_rate": 0.000499909676073295,
      "loss": 5.9275,
      "mean_token_accuracy": 0.1523033708333969,
      "num_tokens": 4006947.0,
      "step": 1795
    },
    {
      "entropy": 5.950516653060913,
      "epoch": 0.20216768686471612,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004999085349878723,
      "loss": 5.9585,
      "mean_token_accuracy": 0.16358500719070435,
      "num_tokens": 4017723.0,
      "step": 1800
    },
    {
      "entropy": 5.962766361236572,
      "epoch": 0.20272926377267367,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999073867412746,
      "loss": 6.0335,
      "mean_token_accuracy": 0.1502598688006401,
      "num_tokens": 4028347.0,
      "step": 1805
    },
    {
      "entropy": 6.021455144882202,
      "epoch": 0.20329084068063122,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004999062313335385,
      "loss": 5.8942,
      "mean_token_accuracy": 0.16167644560337066,
      "num_tokens": 4039648.0,
      "step": 1810
    },
    {
      "entropy": 5.977356767654419,
      "epoch": 0.20385241758858877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004999050687647008,
      "loss": 5.9553,
      "mean_token_accuracy": 0.15589284300804138,
      "num_tokens": 4050786.0,
      "step": 1815
    },
    {
      "entropy": 6.015071296691895,
      "epoch": 0.2044139944965463,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004999038990347984,
      "loss": 5.9929,
      "mean_token_accuracy": 0.16174751073122023,
      "num_tokens": 4061184.0,
      "step": 1820
    },
    {
      "entropy": 6.110763120651245,
      "epoch": 0.20497557140450384,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004999027221438688,
      "loss": 6.1276,
      "mean_token_accuracy": 0.15206350088119508,
      "num_tokens": 4073607.0,
      "step": 1825
    },
    {
      "entropy": 5.954882335662842,
      "epoch": 0.2055371483124614,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499901538091949,
      "loss": 5.9348,
      "mean_token_accuracy": 0.15849609076976776,
      "num_tokens": 4083924.0,
      "step": 1830
    },
    {
      "entropy": 5.934993076324463,
      "epoch": 0.20609872522041894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004999003468790773,
      "loss": 5.9227,
      "mean_token_accuracy": 0.15979374349117278,
      "num_tokens": 4094083.0,
      "step": 1835
    },
    {
      "entropy": 6.010126113891602,
      "epoch": 0.2066603021283765,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998991485052911,
      "loss": 5.9711,
      "mean_token_accuracy": 0.16147021800279618,
      "num_tokens": 4105124.0,
      "step": 1840
    },
    {
      "entropy": 6.109958982467651,
      "epoch": 0.20722187903633402,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000499897942970629,
      "loss": 6.1275,
      "mean_token_accuracy": 0.14718503132462502,
      "num_tokens": 4116792.0,
      "step": 1845
    },
    {
      "entropy": 6.114938068389892,
      "epoch": 0.20778345594429157,
      "grad_norm": 0.87890625,
      "learning_rate": 0.000499896730275129,
      "loss": 6.1146,
      "mean_token_accuracy": 0.1512462429702282,
      "num_tokens": 4127441.0,
      "step": 1850
    },
    {
      "entropy": 5.993119430541992,
      "epoch": 0.20834503285224912,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049989551041883,
      "loss": 6.0981,
      "mean_token_accuracy": 0.14911631792783736,
      "num_tokens": 4138696.0,
      "step": 1855
    },
    {
      "entropy": 6.001512622833252,
      "epoch": 0.20890660976020667,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004998942834017707,
      "loss": 5.9664,
      "mean_token_accuracy": 0.15618520528078078,
      "num_tokens": 4151254.0,
      "step": 1860
    },
    {
      "entropy": 5.903667259216308,
      "epoch": 0.2094681866681642,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998930492239901,
      "loss": 5.9333,
      "mean_token_accuracy": 0.155420657992363,
      "num_tokens": 4162046.0,
      "step": 1865
    },
    {
      "entropy": 5.980195140838623,
      "epoch": 0.21002976357612174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004998918078855277,
      "loss": 5.8815,
      "mean_token_accuracy": 0.16040604412555695,
      "num_tokens": 4172409.0,
      "step": 1870
    },
    {
      "entropy": 5.9213951587677,
      "epoch": 0.2105913404840793,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004998905593864227,
      "loss": 6.0383,
      "mean_token_accuracy": 0.16037157773971558,
      "num_tokens": 4182949.0,
      "step": 1875
    },
    {
      "entropy": 6.040307188034058,
      "epoch": 0.21115291739203684,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004998893037267153,
      "loss": 5.9975,
      "mean_token_accuracy": 0.15625440329313278,
      "num_tokens": 4194164.0,
      "step": 1880
    },
    {
      "entropy": 6.089320707321167,
      "epoch": 0.2117144942999944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998880409064452,
      "loss": 5.9505,
      "mean_token_accuracy": 0.1591552823781967,
      "num_tokens": 4206056.0,
      "step": 1885
    },
    {
      "entropy": 5.8916090965271,
      "epoch": 0.21227607120795192,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998867709256527,
      "loss": 6.0341,
      "mean_token_accuracy": 0.1535590335726738,
      "num_tokens": 4216660.0,
      "step": 1890
    },
    {
      "entropy": 6.002785873413086,
      "epoch": 0.21283764811590947,
      "grad_norm": 0.890625,
      "learning_rate": 0.000499885493784378,
      "loss": 5.9187,
      "mean_token_accuracy": 0.15967616736888884,
      "num_tokens": 4228495.0,
      "step": 1895
    },
    {
      "entropy": 5.95042576789856,
      "epoch": 0.21339922502386702,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004998842094826621,
      "loss": 5.916,
      "mean_token_accuracy": 0.15873922109603883,
      "num_tokens": 4238560.0,
      "step": 1900
    },
    {
      "entropy": 6.099523401260376,
      "epoch": 0.21396080193182457,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998829180205458,
      "loss": 6.1178,
      "mean_token_accuracy": 0.153577783703804,
      "num_tokens": 4250553.0,
      "step": 1905
    },
    {
      "entropy": 5.987069082260132,
      "epoch": 0.21452237883978212,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998816193980701,
      "loss": 6.0338,
      "mean_token_accuracy": 0.15206430703401566,
      "num_tokens": 4262050.0,
      "step": 1910
    },
    {
      "entropy": 6.141225290298462,
      "epoch": 0.21508395574773964,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004998803136152764,
      "loss": 5.9872,
      "mean_token_accuracy": 0.15555329471826554,
      "num_tokens": 4273019.0,
      "step": 1915
    },
    {
      "entropy": 5.866562271118164,
      "epoch": 0.2156455326556972,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998790006722063,
      "loss": 5.8698,
      "mean_token_accuracy": 0.1652226209640503,
      "num_tokens": 4283585.0,
      "step": 1920
    },
    {
      "entropy": 5.903815793991089,
      "epoch": 0.21620710956365474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004998776805689016,
      "loss": 5.9505,
      "mean_token_accuracy": 0.15552265867590903,
      "num_tokens": 4294905.0,
      "step": 1925
    },
    {
      "entropy": 6.123217678070068,
      "epoch": 0.2167686864716123,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004998763533054045,
      "loss": 6.0647,
      "mean_token_accuracy": 0.16015672981739043,
      "num_tokens": 4305926.0,
      "step": 1930
    },
    {
      "entropy": 5.9090136051177975,
      "epoch": 0.21733026337956984,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998750188817568,
      "loss": 6.0258,
      "mean_token_accuracy": 0.15124135613441467,
      "num_tokens": 4318297.0,
      "step": 1935
    },
    {
      "entropy": 5.996935319900513,
      "epoch": 0.21789184028752737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998736772980015,
      "loss": 5.8483,
      "mean_token_accuracy": 0.16015774011611938,
      "num_tokens": 4328795.0,
      "step": 1940
    },
    {
      "entropy": 5.8646955490112305,
      "epoch": 0.21845341719548492,
      "grad_norm": 0.8671875,
      "learning_rate": 0.000499872328554181,
      "loss": 5.862,
      "mean_token_accuracy": 0.16194941848516464,
      "num_tokens": 4340279.0,
      "step": 1945
    },
    {
      "entropy": 6.046488285064697,
      "epoch": 0.21901499410344247,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998709726503383,
      "loss": 6.0497,
      "mean_token_accuracy": 0.1511320725083351,
      "num_tokens": 4352157.0,
      "step": 1950
    },
    {
      "entropy": 5.998716926574707,
      "epoch": 0.21957657101140002,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004998696095865168,
      "loss": 5.8847,
      "mean_token_accuracy": 0.16323180198669435,
      "num_tokens": 4361909.0,
      "step": 1955
    },
    {
      "entropy": 5.7217999458312985,
      "epoch": 0.22013814791935757,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998682393627595,
      "loss": 5.799,
      "mean_token_accuracy": 0.1613177850842476,
      "num_tokens": 4372177.0,
      "step": 1960
    },
    {
      "entropy": 5.92481837272644,
      "epoch": 0.2206997248273151,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998668619791104,
      "loss": 5.8547,
      "mean_token_accuracy": 0.16234170347452165,
      "num_tokens": 4382872.0,
      "step": 1965
    },
    {
      "entropy": 5.908624505996704,
      "epoch": 0.22126130173527264,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004998654774356131,
      "loss": 5.9559,
      "mean_token_accuracy": 0.1552143692970276,
      "num_tokens": 4393623.0,
      "step": 1970
    },
    {
      "entropy": 5.96333270072937,
      "epoch": 0.2218228786432302,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004998640857323118,
      "loss": 5.8898,
      "mean_token_accuracy": 0.15621398091316224,
      "num_tokens": 4404215.0,
      "step": 1975
    },
    {
      "entropy": 6.029521846771241,
      "epoch": 0.22238445555118774,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998626868692509,
      "loss": 6.1434,
      "mean_token_accuracy": 0.15685454457998277,
      "num_tokens": 4415978.0,
      "step": 1980
    },
    {
      "entropy": 6.012211847305298,
      "epoch": 0.2229460324591453,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998612808464748,
      "loss": 5.8877,
      "mean_token_accuracy": 0.15981114357709886,
      "num_tokens": 4427060.0,
      "step": 1985
    },
    {
      "entropy": 5.914674568176269,
      "epoch": 0.22350760936710282,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004998598676640283,
      "loss": 5.971,
      "mean_token_accuracy": 0.156388621032238,
      "num_tokens": 4438145.0,
      "step": 1990
    },
    {
      "entropy": 6.117107343673706,
      "epoch": 0.22406918627506037,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004998584473219563,
      "loss": 6.0468,
      "mean_token_accuracy": 0.1594668760895729,
      "num_tokens": 4449324.0,
      "step": 1995
    },
    {
      "entropy": 5.848163843154907,
      "epoch": 0.22463076318301792,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004998570198203043,
      "loss": 5.7684,
      "mean_token_accuracy": 0.16067076921463014,
      "num_tokens": 4459728.0,
      "step": 2000
    },
    {
      "entropy": 5.908514308929443,
      "epoch": 0.22519234009097547,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998555851591175,
      "loss": 5.9773,
      "mean_token_accuracy": 0.16138978004455568,
      "num_tokens": 4471164.0,
      "step": 2005
    },
    {
      "entropy": 6.021873474121094,
      "epoch": 0.225753916998933,
      "grad_norm": 0.85546875,
      "learning_rate": 0.0004998541433384416,
      "loss": 6.0352,
      "mean_token_accuracy": 0.15729770064353943,
      "num_tokens": 4483349.0,
      "step": 2010
    },
    {
      "entropy": 6.085103368759155,
      "epoch": 0.22631549390689054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998526943583226,
      "loss": 6.0037,
      "mean_token_accuracy": 0.15304621160030366,
      "num_tokens": 4495311.0,
      "step": 2015
    },
    {
      "entropy": 5.900832271575927,
      "epoch": 0.2268770708148481,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004998512382188066,
      "loss": 5.9293,
      "mean_token_accuracy": 0.16477998793125154,
      "num_tokens": 4507017.0,
      "step": 2020
    },
    {
      "entropy": 5.960407686233521,
      "epoch": 0.22743864772280564,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004998497749199401,
      "loss": 5.9688,
      "mean_token_accuracy": 0.15732726454734802,
      "num_tokens": 4517746.0,
      "step": 2025
    },
    {
      "entropy": 6.0614457607269285,
      "epoch": 0.2280002246307632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998483044617694,
      "loss": 6.0633,
      "mean_token_accuracy": 0.15260435193777083,
      "num_tokens": 4529149.0,
      "step": 2030
    },
    {
      "entropy": 5.929168272018432,
      "epoch": 0.22856180153872072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004998468268443416,
      "loss": 5.9371,
      "mean_token_accuracy": 0.1589831829071045,
      "num_tokens": 4540630.0,
      "step": 2035
    },
    {
      "entropy": 5.85460319519043,
      "epoch": 0.22912337844667827,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004998453420677034,
      "loss": 5.8672,
      "mean_token_accuracy": 0.1566688224673271,
      "num_tokens": 4552877.0,
      "step": 2040
    },
    {
      "entropy": 6.016587114334106,
      "epoch": 0.22968495535463582,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998438501319025,
      "loss": 5.9635,
      "mean_token_accuracy": 0.16212720274925232,
      "num_tokens": 4563634.0,
      "step": 2045
    },
    {
      "entropy": 5.904137372970581,
      "epoch": 0.23024653226259337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004998423510369862,
      "loss": 5.8992,
      "mean_token_accuracy": 0.16215125918388368,
      "num_tokens": 4574007.0,
      "step": 2050
    },
    {
      "entropy": 5.954766511917114,
      "epoch": 0.23080810917055092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004998408447830022,
      "loss": 5.911,
      "mean_token_accuracy": 0.1593327283859253,
      "num_tokens": 4584146.0,
      "step": 2055
    },
    {
      "entropy": 5.965726757049561,
      "epoch": 0.23136968607850844,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004998393313699985,
      "loss": 5.9217,
      "mean_token_accuracy": 0.15980084091424943,
      "num_tokens": 4594839.0,
      "step": 2060
    },
    {
      "entropy": 6.105428791046142,
      "epoch": 0.231931262986466,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998378107980233,
      "loss": 6.0798,
      "mean_token_accuracy": 0.14904378205537797,
      "num_tokens": 4608229.0,
      "step": 2065
    },
    {
      "entropy": 5.952469778060913,
      "epoch": 0.23249283989442354,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004998362830671249,
      "loss": 5.8784,
      "mean_token_accuracy": 0.15615165382623672,
      "num_tokens": 4619206.0,
      "step": 2070
    },
    {
      "entropy": 5.863278198242187,
      "epoch": 0.2330544168023811,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998347481773522,
      "loss": 5.9249,
      "mean_token_accuracy": 0.16034539192914962,
      "num_tokens": 4628966.0,
      "step": 2075
    },
    {
      "entropy": 5.915069961547852,
      "epoch": 0.23361599371033864,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004998332061287538,
      "loss": 5.8112,
      "mean_token_accuracy": 0.15875162035226822,
      "num_tokens": 4639935.0,
      "step": 2080
    },
    {
      "entropy": 6.008445787429809,
      "epoch": 0.23417757061829617,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499831656921379,
      "loss": 6.0489,
      "mean_token_accuracy": 0.1555440604686737,
      "num_tokens": 4651998.0,
      "step": 2085
    },
    {
      "entropy": 6.036960220336914,
      "epoch": 0.23473914752625372,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000499830100555277,
      "loss": 6.0192,
      "mean_token_accuracy": 0.1560396894812584,
      "num_tokens": 4664002.0,
      "step": 2090
    },
    {
      "entropy": 5.941565656661988,
      "epoch": 0.23530072443421127,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004998285370304974,
      "loss": 5.9088,
      "mean_token_accuracy": 0.1645731508731842,
      "num_tokens": 4675343.0,
      "step": 2095
    },
    {
      "entropy": 5.974581003189087,
      "epoch": 0.23586230134216882,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00049982696634709,
      "loss": 5.9546,
      "mean_token_accuracy": 0.16146257668733596,
      "num_tokens": 4686410.0,
      "step": 2100
    },
    {
      "entropy": 5.937360239028931,
      "epoch": 0.23642387825012637,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004998253885051048,
      "loss": 5.984,
      "mean_token_accuracy": 0.1578327551484108,
      "num_tokens": 4697768.0,
      "step": 2105
    },
    {
      "entropy": 6.069368505477906,
      "epoch": 0.2369854551580839,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000499823803504592,
      "loss": 5.993,
      "mean_token_accuracy": 0.15976879000663757,
      "num_tokens": 4709933.0,
      "step": 2110
    },
    {
      "entropy": 5.981991481781006,
      "epoch": 0.23754703206604144,
      "grad_norm": 1.0,
      "learning_rate": 0.0004998222113456022,
      "loss": 6.0305,
      "mean_token_accuracy": 0.15242783725261688,
      "num_tokens": 4721385.0,
      "step": 2115
    },
    {
      "entropy": 5.920117902755737,
      "epoch": 0.238108608973999,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004998206120281858,
      "loss": 5.8442,
      "mean_token_accuracy": 0.1651250496506691,
      "num_tokens": 4732562.0,
      "step": 2120
    },
    {
      "entropy": 5.845587635040284,
      "epoch": 0.23867018588195654,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499819005552394,
      "loss": 5.8581,
      "mean_token_accuracy": 0.15949287563562392,
      "num_tokens": 4742794.0,
      "step": 2125
    },
    {
      "entropy": 5.866573333740234,
      "epoch": 0.23923176278991407,
      "grad_norm": 0.8515625,
      "learning_rate": 0.0004998173919182779,
      "loss": 5.8202,
      "mean_token_accuracy": 0.16423655897378922,
      "num_tokens": 4754903.0,
      "step": 2130
    },
    {
      "entropy": 5.91752257347107,
      "epoch": 0.23979333969787162,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000499815771125889,
      "loss": 5.8741,
      "mean_token_accuracy": 0.1597582370042801,
      "num_tokens": 4766165.0,
      "step": 2135
    },
    {
      "entropy": 5.860339212417602,
      "epoch": 0.24035491660582917,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998141431752785,
      "loss": 5.8838,
      "mean_token_accuracy": 0.16476210951805115,
      "num_tokens": 4776551.0,
      "step": 2140
    },
    {
      "entropy": 6.054014873504639,
      "epoch": 0.24091649351378672,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004998125080664987,
      "loss": 6.067,
      "mean_token_accuracy": 0.1543096624314785,
      "num_tokens": 4788344.0,
      "step": 2145
    },
    {
      "entropy": 5.984238958358764,
      "epoch": 0.24147807042174427,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004998108657996012,
      "loss": 5.8624,
      "mean_token_accuracy": 0.1560610830783844,
      "num_tokens": 4799314.0,
      "step": 2150
    },
    {
      "entropy": 5.885314512252807,
      "epoch": 0.2420396473297018,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004998092163746387,
      "loss": 5.9116,
      "mean_token_accuracy": 0.16344153732061387,
      "num_tokens": 4809588.0,
      "step": 2155
    },
    {
      "entropy": 5.870743465423584,
      "epoch": 0.24260122423765934,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004998075597916636,
      "loss": 5.7357,
      "mean_token_accuracy": 0.16858597844839096,
      "num_tokens": 4820076.0,
      "step": 2160
    },
    {
      "entropy": 5.9259528636932375,
      "epoch": 0.2431628011456169,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998058960507286,
      "loss": 6.0033,
      "mean_token_accuracy": 0.1558015376329422,
      "num_tokens": 4831504.0,
      "step": 2165
    },
    {
      "entropy": 5.880033540725708,
      "epoch": 0.24372437805357444,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998042251518866,
      "loss": 5.8649,
      "mean_token_accuracy": 0.1617542490363121,
      "num_tokens": 4842826.0,
      "step": 2170
    },
    {
      "entropy": 5.8402345180511475,
      "epoch": 0.244285954961532,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004998025470951908,
      "loss": 5.8465,
      "mean_token_accuracy": 0.1644066721200943,
      "num_tokens": 4853279.0,
      "step": 2175
    },
    {
      "entropy": 5.966151905059815,
      "epoch": 0.24484753186948952,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004998008618806949,
      "loss": 5.8979,
      "mean_token_accuracy": 0.16503124088048934,
      "num_tokens": 4863332.0,
      "step": 2180
    },
    {
      "entropy": 5.958868360519409,
      "epoch": 0.24540910877744707,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997991695084523,
      "loss": 6.0339,
      "mean_token_accuracy": 0.1506052479147911,
      "num_tokens": 4874217.0,
      "step": 2185
    },
    {
      "entropy": 6.0297637462615965,
      "epoch": 0.24597068568540462,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997974699785169,
      "loss": 5.881,
      "mean_token_accuracy": 0.162646321952343,
      "num_tokens": 4885322.0,
      "step": 2190
    },
    {
      "entropy": 5.715388011932373,
      "epoch": 0.24653226259336217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997957632909429,
      "loss": 5.8452,
      "mean_token_accuracy": 0.15895508527755736,
      "num_tokens": 4895546.0,
      "step": 2195
    },
    {
      "entropy": 6.107920455932617,
      "epoch": 0.24709383950131972,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997940494457846,
      "loss": 6.0395,
      "mean_token_accuracy": 0.15601962208747863,
      "num_tokens": 4906957.0,
      "step": 2200
    },
    {
      "entropy": 6.023680686950684,
      "epoch": 0.24765541640927724,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997923284430967,
      "loss": 5.9557,
      "mean_token_accuracy": 0.15208095461130142,
      "num_tokens": 4917965.0,
      "step": 2205
    },
    {
      "entropy": 5.857148885726929,
      "epoch": 0.2482169933172348,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004997906002829338,
      "loss": 5.8617,
      "mean_token_accuracy": 0.1620682030916214,
      "num_tokens": 4928585.0,
      "step": 2210
    },
    {
      "entropy": 5.891238260269165,
      "epoch": 0.24877857022519234,
      "grad_norm": 0.87109375,
      "learning_rate": 0.000499788864965351,
      "loss": 5.8717,
      "mean_token_accuracy": 0.15956535786390305,
      "num_tokens": 4939818.0,
      "step": 2215
    },
    {
      "entropy": 5.967899799346924,
      "epoch": 0.2493401471331499,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004997871224904036,
      "loss": 5.9513,
      "mean_token_accuracy": 0.15834061205387115,
      "num_tokens": 4951752.0,
      "step": 2220
    },
    {
      "entropy": 5.793693161010742,
      "epoch": 0.24990172404110744,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004997853728581468,
      "loss": 5.7895,
      "mean_token_accuracy": 0.1681654319167137,
      "num_tokens": 4961846.0,
      "step": 2225
    },
    {
      "entropy": 5.945404481887818,
      "epoch": 0.250463300949065,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997836160686368,
      "loss": 5.9361,
      "mean_token_accuracy": 0.15449708849191665,
      "num_tokens": 4973693.0,
      "step": 2230
    },
    {
      "entropy": 5.915171194076538,
      "epoch": 0.25102487785702254,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004997818521219291,
      "loss": 5.8589,
      "mean_token_accuracy": 0.1630162000656128,
      "num_tokens": 4986156.0,
      "step": 2235
    },
    {
      "entropy": 5.906194400787354,
      "epoch": 0.25158645476498004,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004997800810180803,
      "loss": 5.9333,
      "mean_token_accuracy": 0.1587043434381485,
      "num_tokens": 4997615.0,
      "step": 2240
    },
    {
      "entropy": 5.911606407165527,
      "epoch": 0.2521480316729376,
      "grad_norm": 1.0,
      "learning_rate": 0.0004997783027571464,
      "loss": 5.8738,
      "mean_token_accuracy": 0.1615534693002701,
      "num_tokens": 5007655.0,
      "step": 2245
    },
    {
      "entropy": 5.969314908981323,
      "epoch": 0.25270960858089514,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004997765173391841,
      "loss": 5.9367,
      "mean_token_accuracy": 0.16165156960487365,
      "num_tokens": 5018933.0,
      "step": 2250
    },
    {
      "entropy": 6.040049314498901,
      "epoch": 0.2532711854888527,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004997747247642505,
      "loss": 5.9383,
      "mean_token_accuracy": 0.1602029874920845,
      "num_tokens": 5030495.0,
      "step": 2255
    },
    {
      "entropy": 5.842279529571533,
      "epoch": 0.25383276239681024,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004997729250324023,
      "loss": 5.9816,
      "mean_token_accuracy": 0.15450992733240126,
      "num_tokens": 5041553.0,
      "step": 2260
    },
    {
      "entropy": 5.963804531097412,
      "epoch": 0.2543943393047678,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997711181436971,
      "loss": 5.9114,
      "mean_token_accuracy": 0.16113370209932326,
      "num_tokens": 5052312.0,
      "step": 2265
    },
    {
      "entropy": 6.014153528213501,
      "epoch": 0.25495591621272534,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997693040981922,
      "loss": 6.0888,
      "mean_token_accuracy": 0.1531001329421997,
      "num_tokens": 5063445.0,
      "step": 2270
    },
    {
      "entropy": 6.010663318634033,
      "epoch": 0.2555174931206829,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997674828959456,
      "loss": 5.8763,
      "mean_token_accuracy": 0.16219639629125596,
      "num_tokens": 5073052.0,
      "step": 2275
    },
    {
      "entropy": 5.857589530944824,
      "epoch": 0.25607907002864044,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997656545370151,
      "loss": 5.9201,
      "mean_token_accuracy": 0.16253976821899413,
      "num_tokens": 5084174.0,
      "step": 2280
    },
    {
      "entropy": 6.022715759277344,
      "epoch": 0.256640646936598,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499763819021459,
      "loss": 5.9051,
      "mean_token_accuracy": 0.16299960166215896,
      "num_tokens": 5094106.0,
      "step": 2285
    },
    {
      "entropy": 5.972675132751465,
      "epoch": 0.2572022238445555,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004997619763493357,
      "loss": 5.9997,
      "mean_token_accuracy": 0.16121359914541245,
      "num_tokens": 5106963.0,
      "step": 2290
    },
    {
      "entropy": 5.864818668365478,
      "epoch": 0.25776380075251304,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499760126520704,
      "loss": 5.7925,
      "mean_token_accuracy": 0.15583763271570206,
      "num_tokens": 5119476.0,
      "step": 2295
    },
    {
      "entropy": 5.973382377624512,
      "epoch": 0.2583253776604706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997582695356225,
      "loss": 6.0256,
      "mean_token_accuracy": 0.15469253659248353,
      "num_tokens": 5132169.0,
      "step": 2300
    },
    {
      "entropy": 5.877126216888428,
      "epoch": 0.25888695456842814,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997564053941507,
      "loss": 5.8854,
      "mean_token_accuracy": 0.16001738011837005,
      "num_tokens": 5143354.0,
      "step": 2305
    },
    {
      "entropy": 5.895409488677979,
      "epoch": 0.2594485314763857,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997545340963478,
      "loss": 5.9458,
      "mean_token_accuracy": 0.15308494120836258,
      "num_tokens": 5154624.0,
      "step": 2310
    },
    {
      "entropy": 5.987952375411988,
      "epoch": 0.26001010838434324,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004997526556422733,
      "loss": 5.8418,
      "mean_token_accuracy": 0.16011188477277755,
      "num_tokens": 5165349.0,
      "step": 2315
    },
    {
      "entropy": 5.847837829589844,
      "epoch": 0.2605716852923008,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004997507700319872,
      "loss": 5.9174,
      "mean_token_accuracy": 0.16073694676160813,
      "num_tokens": 5175874.0,
      "step": 2320
    },
    {
      "entropy": 5.864932250976563,
      "epoch": 0.26113326220025834,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004997488772655492,
      "loss": 5.8153,
      "mean_token_accuracy": 0.16911125481128692,
      "num_tokens": 5186591.0,
      "step": 2325
    },
    {
      "entropy": 5.867122650146484,
      "epoch": 0.2616948391082159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004997469773430199,
      "loss": 5.9228,
      "mean_token_accuracy": 0.16698685735464097,
      "num_tokens": 5197307.0,
      "step": 2330
    },
    {
      "entropy": 5.876765441894531,
      "epoch": 0.2622564160161734,
      "grad_norm": 0.8359375,
      "learning_rate": 0.0004997450702644596,
      "loss": 5.8734,
      "mean_token_accuracy": 0.16059717684984207,
      "num_tokens": 5209868.0,
      "step": 2335
    },
    {
      "entropy": 5.920298910140991,
      "epoch": 0.26281799292413094,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997431560299292,
      "loss": 5.8907,
      "mean_token_accuracy": 0.16422481536865235,
      "num_tokens": 5220365.0,
      "step": 2340
    },
    {
      "entropy": 5.885955953598023,
      "epoch": 0.2633795698320885,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997412346394894,
      "loss": 5.8033,
      "mean_token_accuracy": 0.16235794126987457,
      "num_tokens": 5231575.0,
      "step": 2345
    },
    {
      "entropy": 5.947313213348389,
      "epoch": 0.26394114674004604,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004997393060932017,
      "loss": 5.887,
      "mean_token_accuracy": 0.16234982758760452,
      "num_tokens": 5241852.0,
      "step": 2350
    },
    {
      "entropy": 5.920132827758789,
      "epoch": 0.2645027236480036,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004997373703911273,
      "loss": 5.9284,
      "mean_token_accuracy": 0.16154667288064956,
      "num_tokens": 5254536.0,
      "step": 2355
    },
    {
      "entropy": 5.923627662658691,
      "epoch": 0.26506430055596114,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997354275333277,
      "loss": 5.8557,
      "mean_token_accuracy": 0.1607941135764122,
      "num_tokens": 5264946.0,
      "step": 2360
    },
    {
      "entropy": 5.686444616317749,
      "epoch": 0.2656258774639187,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499733477519865,
      "loss": 5.6798,
      "mean_token_accuracy": 0.16499558091163635,
      "num_tokens": 5275527.0,
      "step": 2365
    },
    {
      "entropy": 5.865219974517823,
      "epoch": 0.26618745437187624,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004997315203508013,
      "loss": 5.8695,
      "mean_token_accuracy": 0.15666571706533433,
      "num_tokens": 5285471.0,
      "step": 2370
    },
    {
      "entropy": 5.798485612869262,
      "epoch": 0.2667490312798338,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004997295560261988,
      "loss": 5.7753,
      "mean_token_accuracy": 0.1605561688542366,
      "num_tokens": 5295743.0,
      "step": 2375
    },
    {
      "entropy": 5.981579542160034,
      "epoch": 0.26731060818779134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00049972758454612,
      "loss": 5.9192,
      "mean_token_accuracy": 0.15332525223493576,
      "num_tokens": 5307963.0,
      "step": 2380
    },
    {
      "entropy": 5.880096340179444,
      "epoch": 0.26787218509574884,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004997256059106279,
      "loss": 5.8074,
      "mean_token_accuracy": 0.16239653378725052,
      "num_tokens": 5318394.0,
      "step": 2385
    },
    {
      "entropy": 5.912300062179566,
      "epoch": 0.2684337620037064,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004997236201197853,
      "loss": 6.037,
      "mean_token_accuracy": 0.14811042100191116,
      "num_tokens": 5331728.0,
      "step": 2390
    },
    {
      "entropy": 5.964923286437989,
      "epoch": 0.26899533891166394,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004997216271736554,
      "loss": 5.922,
      "mean_token_accuracy": 0.1677936553955078,
      "num_tokens": 5342825.0,
      "step": 2395
    },
    {
      "entropy": 5.907764720916748,
      "epoch": 0.2695569158196215,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004997196270723018,
      "loss": 5.8239,
      "mean_token_accuracy": 0.1624791741371155,
      "num_tokens": 5354232.0,
      "step": 2400
    },
    {
      "entropy": 5.892585563659668,
      "epoch": 0.27011849272757904,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004997176198157881,
      "loss": 5.9215,
      "mean_token_accuracy": 0.15687112659215927,
      "num_tokens": 5364807.0,
      "step": 2405
    },
    {
      "entropy": 5.903974103927612,
      "epoch": 0.2706800696355366,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997156054041781,
      "loss": 5.8953,
      "mean_token_accuracy": 0.15242768228054046,
      "num_tokens": 5375593.0,
      "step": 2410
    },
    {
      "entropy": 5.877764558792114,
      "epoch": 0.27124164654349414,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004997135838375362,
      "loss": 5.8406,
      "mean_token_accuracy": 0.16519979536533355,
      "num_tokens": 5386740.0,
      "step": 2415
    },
    {
      "entropy": 5.813501453399658,
      "epoch": 0.2718032234514517,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004997115551159267,
      "loss": 5.8211,
      "mean_token_accuracy": 0.1626619353890419,
      "num_tokens": 5397732.0,
      "step": 2420
    },
    {
      "entropy": 5.987624835968018,
      "epoch": 0.27236480035940924,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000499709519239414,
      "loss": 5.886,
      "mean_token_accuracy": 0.15518247038125993,
      "num_tokens": 5409392.0,
      "step": 2425
    },
    {
      "entropy": 5.722806978225708,
      "epoch": 0.2729263772673668,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004997074762080632,
      "loss": 5.8547,
      "mean_token_accuracy": 0.16847770363092424,
      "num_tokens": 5420903.0,
      "step": 2430
    },
    {
      "entropy": 5.803295469284057,
      "epoch": 0.2734879541753243,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499705426021939,
      "loss": 5.7793,
      "mean_token_accuracy": 0.16043397933244705,
      "num_tokens": 5432631.0,
      "step": 2435
    },
    {
      "entropy": 5.865436649322509,
      "epoch": 0.27404953108328184,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499703368681107,
      "loss": 5.7477,
      "mean_token_accuracy": 0.16816219240427016,
      "num_tokens": 5443463.0,
      "step": 2440
    },
    {
      "entropy": 5.954450368881226,
      "epoch": 0.2746111079912394,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004997013041856326,
      "loss": 6.0624,
      "mean_token_accuracy": 0.15737709179520606,
      "num_tokens": 5455903.0,
      "step": 2445
    },
    {
      "entropy": 5.9812792301177975,
      "epoch": 0.27517268489919694,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004996992325355815,
      "loss": 5.8839,
      "mean_token_accuracy": 0.15456786453723909,
      "num_tokens": 5467201.0,
      "step": 2450
    },
    {
      "entropy": 5.89325475692749,
      "epoch": 0.2757342618071545,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004996971537310198,
      "loss": 5.7582,
      "mean_token_accuracy": 0.16121137291193008,
      "num_tokens": 5478109.0,
      "step": 2455
    },
    {
      "entropy": 5.869486522674561,
      "epoch": 0.27629583871511204,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004996950677720133,
      "loss": 5.8984,
      "mean_token_accuracy": 0.1576277181506157,
      "num_tokens": 5489621.0,
      "step": 2460
    },
    {
      "entropy": 5.821279716491699,
      "epoch": 0.2768574156230696,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499692974658629,
      "loss": 5.8327,
      "mean_token_accuracy": 0.16300647109746932,
      "num_tokens": 5500606.0,
      "step": 2465
    },
    {
      "entropy": 5.920242738723755,
      "epoch": 0.27741899253102714,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004996908743909331,
      "loss": 5.7813,
      "mean_token_accuracy": 0.1680440992116928,
      "num_tokens": 5510768.0,
      "step": 2470
    },
    {
      "entropy": 5.775556421279907,
      "epoch": 0.2779805694389847,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004996887669689926,
      "loss": 5.7973,
      "mean_token_accuracy": 0.16250731498003007,
      "num_tokens": 5522210.0,
      "step": 2475
    },
    {
      "entropy": 5.893093776702881,
      "epoch": 0.2785421463469422,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996866523928747,
      "loss": 5.931,
      "mean_token_accuracy": 0.1568371132016182,
      "num_tokens": 5533055.0,
      "step": 2480
    },
    {
      "entropy": 5.997058773040772,
      "epoch": 0.27910372325489974,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996845306626467,
      "loss": 5.9507,
      "mean_token_accuracy": 0.15390594601631163,
      "num_tokens": 5544345.0,
      "step": 2485
    },
    {
      "entropy": 5.914350700378418,
      "epoch": 0.2796653001628573,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996824017783759,
      "loss": 5.8473,
      "mean_token_accuracy": 0.15839995592832565,
      "num_tokens": 5556734.0,
      "step": 2490
    },
    {
      "entropy": 5.818574714660644,
      "epoch": 0.28022687707081484,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996802657401304,
      "loss": 5.8816,
      "mean_token_accuracy": 0.16247260347008705,
      "num_tokens": 5566874.0,
      "step": 2495
    },
    {
      "entropy": 5.952364873886109,
      "epoch": 0.2807884539787724,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004996781225479781,
      "loss": 5.8857,
      "mean_token_accuracy": 0.1558615043759346,
      "num_tokens": 5578812.0,
      "step": 2500
    },
    {
      "entropy": 5.894041633605957,
      "epoch": 0.28135003088672994,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004996759722019872,
      "loss": 5.9373,
      "mean_token_accuracy": 0.1599496230483055,
      "num_tokens": 5590021.0,
      "step": 2505
    },
    {
      "entropy": 5.947154235839844,
      "epoch": 0.2819116077946875,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996738147022262,
      "loss": 5.9428,
      "mean_token_accuracy": 0.15261465460062026,
      "num_tokens": 5600921.0,
      "step": 2510
    },
    {
      "entropy": 5.871411657333374,
      "epoch": 0.28247318470264504,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004996716500487638,
      "loss": 5.8606,
      "mean_token_accuracy": 0.1596810534596443,
      "num_tokens": 5610615.0,
      "step": 2515
    },
    {
      "entropy": 5.800993251800537,
      "epoch": 0.2830347616106026,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000499669478241669,
      "loss": 5.7878,
      "mean_token_accuracy": 0.16247710585594177,
      "num_tokens": 5621596.0,
      "step": 2520
    },
    {
      "entropy": 5.969037485122681,
      "epoch": 0.28359633851856014,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004996672992810108,
      "loss": 5.8243,
      "mean_token_accuracy": 0.16356842517852782,
      "num_tokens": 5633047.0,
      "step": 2525
    },
    {
      "entropy": 5.798973751068115,
      "epoch": 0.28415791542651764,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004996651131668587,
      "loss": 5.8104,
      "mean_token_accuracy": 0.16191127747297288,
      "num_tokens": 5644970.0,
      "step": 2530
    },
    {
      "entropy": 5.87383804321289,
      "epoch": 0.2847194923344752,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004996629198992823,
      "loss": 5.9332,
      "mean_token_accuracy": 0.15747836232185364,
      "num_tokens": 5656631.0,
      "step": 2535
    },
    {
      "entropy": 5.888602876663208,
      "epoch": 0.28528106924243274,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004996607194783513,
      "loss": 5.8852,
      "mean_token_accuracy": 0.1580764904618263,
      "num_tokens": 5668808.0,
      "step": 2540
    },
    {
      "entropy": 5.818170499801636,
      "epoch": 0.2858426461503903,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004996585119041359,
      "loss": 5.827,
      "mean_token_accuracy": 0.1570512518286705,
      "num_tokens": 5679493.0,
      "step": 2545
    },
    {
      "entropy": 5.916708993911743,
      "epoch": 0.28640422305834784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004996562971767063,
      "loss": 5.8515,
      "mean_token_accuracy": 0.16539256125688553,
      "num_tokens": 5689879.0,
      "step": 2550
    },
    {
      "entropy": 5.871814250946045,
      "epoch": 0.2869657999663054,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499654075296133,
      "loss": 5.9404,
      "mean_token_accuracy": 0.16452383995056152,
      "num_tokens": 5699934.0,
      "step": 2555
    },
    {
      "entropy": 5.8691713333129885,
      "epoch": 0.28752737687426294,
      "grad_norm": 1.03125,
      "learning_rate": 0.000499651846262487,
      "loss": 5.8705,
      "mean_token_accuracy": 0.166011680662632,
      "num_tokens": 5710171.0,
      "step": 2560
    },
    {
      "entropy": 5.884806299209595,
      "epoch": 0.2880889537822205,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004996496100758389,
      "loss": 5.8656,
      "mean_token_accuracy": 0.15255345776677132,
      "num_tokens": 5721190.0,
      "step": 2565
    },
    {
      "entropy": 6.02195897102356,
      "epoch": 0.28865053069017804,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004996473667362604,
      "loss": 5.9701,
      "mean_token_accuracy": 0.15485327020287515,
      "num_tokens": 5733706.0,
      "step": 2570
    },
    {
      "entropy": 5.805244255065918,
      "epoch": 0.28921210759813554,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996451162438224,
      "loss": 5.9271,
      "mean_token_accuracy": 0.15993860512971877,
      "num_tokens": 5745712.0,
      "step": 2575
    },
    {
      "entropy": 5.781348705291748,
      "epoch": 0.2897736845060931,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004996428585985967,
      "loss": 5.74,
      "mean_token_accuracy": 0.17483293563127517,
      "num_tokens": 5757915.0,
      "step": 2580
    },
    {
      "entropy": 5.892542028427124,
      "epoch": 0.29033526141405064,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004996405938006553,
      "loss": 5.8859,
      "mean_token_accuracy": 0.16429722160100937,
      "num_tokens": 5768788.0,
      "step": 2585
    },
    {
      "entropy": 5.890336275100708,
      "epoch": 0.2908968383220082,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004996383218500703,
      "loss": 5.8128,
      "mean_token_accuracy": 0.16249943375587464,
      "num_tokens": 5779806.0,
      "step": 2590
    },
    {
      "entropy": 5.772500848770141,
      "epoch": 0.29145841522996574,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499636042746914,
      "loss": 5.769,
      "mean_token_accuracy": 0.16073892414569854,
      "num_tokens": 5790177.0,
      "step": 2595
    },
    {
      "entropy": 5.839260911941528,
      "epoch": 0.2920199921379233,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000499633756491259,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16765276491641998,
      "num_tokens": 5800251.0,
      "step": 2600
    },
    {
      "entropy": 5.935009479522705,
      "epoch": 0.29258156904588084,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996314630831781,
      "loss": 5.8968,
      "mean_token_accuracy": 0.16182940155267717,
      "num_tokens": 5812210.0,
      "step": 2605
    },
    {
      "entropy": 5.87156400680542,
      "epoch": 0.2931431459538384,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004996291625227443,
      "loss": 5.8769,
      "mean_token_accuracy": 0.16384392827749253,
      "num_tokens": 5823280.0,
      "step": 2610
    },
    {
      "entropy": 5.972989130020141,
      "epoch": 0.29370472286179594,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004996268548100307,
      "loss": 5.951,
      "mean_token_accuracy": 0.1541149839758873,
      "num_tokens": 5835076.0,
      "step": 2615
    },
    {
      "entropy": 5.791325569152832,
      "epoch": 0.2942662997697535,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000499624539945111,
      "loss": 5.7486,
      "mean_token_accuracy": 0.16788308024406434,
      "num_tokens": 5845376.0,
      "step": 2620
    },
    {
      "entropy": 5.746144151687622,
      "epoch": 0.294827876677711,
      "grad_norm": 0.875,
      "learning_rate": 0.0004996222179280589,
      "loss": 5.7775,
      "mean_token_accuracy": 0.16059013456106186,
      "num_tokens": 5857124.0,
      "step": 2625
    },
    {
      "entropy": 5.954985284805298,
      "epoch": 0.29538945358566854,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996198887589481,
      "loss": 5.8689,
      "mean_token_accuracy": 0.15687937289476395,
      "num_tokens": 5868681.0,
      "step": 2630
    },
    {
      "entropy": 5.871660566329956,
      "epoch": 0.2959510304936261,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996175524378531,
      "loss": 5.8737,
      "mean_token_accuracy": 0.16142140179872513,
      "num_tokens": 5880532.0,
      "step": 2635
    },
    {
      "entropy": 5.9063619613647464,
      "epoch": 0.29651260740158364,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499615208964848,
      "loss": 5.8278,
      "mean_token_accuracy": 0.1647784188389778,
      "num_tokens": 5891582.0,
      "step": 2640
    },
    {
      "entropy": 5.830147123336792,
      "epoch": 0.2970741843095412,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996128583400077,
      "loss": 5.8292,
      "mean_token_accuracy": 0.15877834409475328,
      "num_tokens": 5903543.0,
      "step": 2645
    },
    {
      "entropy": 5.811954116821289,
      "epoch": 0.29763576121749874,
      "grad_norm": 1.0,
      "learning_rate": 0.0004996105005634066,
      "loss": 5.8006,
      "mean_token_accuracy": 0.16823345720767974,
      "num_tokens": 5913140.0,
      "step": 2650
    },
    {
      "entropy": 5.814200782775879,
      "epoch": 0.2981973381254563,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004996081356351202,
      "loss": 5.7642,
      "mean_token_accuracy": 0.16816403865814208,
      "num_tokens": 5924010.0,
      "step": 2655
    },
    {
      "entropy": 5.857581901550293,
      "epoch": 0.29875891503341384,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004996057635552237,
      "loss": 5.9688,
      "mean_token_accuracy": 0.1579844906926155,
      "num_tokens": 5936246.0,
      "step": 2660
    },
    {
      "entropy": 5.925976657867432,
      "epoch": 0.2993204919413714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004996033843237925,
      "loss": 5.859,
      "mean_token_accuracy": 0.1524143025279045,
      "num_tokens": 5947594.0,
      "step": 2665
    },
    {
      "entropy": 5.971056890487671,
      "epoch": 0.29988206884932894,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004996009979409025,
      "loss": 5.7942,
      "mean_token_accuracy": 0.15486442297697067,
      "num_tokens": 5957660.0,
      "step": 2670
    },
    {
      "entropy": 5.631026124954223,
      "epoch": 0.30044364575728644,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995986044066295,
      "loss": 5.6228,
      "mean_token_accuracy": 0.17383519262075425,
      "num_tokens": 5968881.0,
      "step": 2675
    },
    {
      "entropy": 5.839225244522095,
      "epoch": 0.301005222665244,
      "grad_norm": 0.96875,
      "learning_rate": 0.00049959620372105,
      "loss": 5.7885,
      "mean_token_accuracy": 0.16763374209403992,
      "num_tokens": 5980032.0,
      "step": 2680
    },
    {
      "entropy": 5.758896398544311,
      "epoch": 0.30156679957320154,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004995937958842401,
      "loss": 5.8011,
      "mean_token_accuracy": 0.1621689423918724,
      "num_tokens": 5991781.0,
      "step": 2685
    },
    {
      "entropy": 5.819126558303833,
      "epoch": 0.3021283764811591,
      "grad_norm": 1.0,
      "learning_rate": 0.0004995913808962767,
      "loss": 5.8394,
      "mean_token_accuracy": 0.16058890372514725,
      "num_tokens": 6002303.0,
      "step": 2690
    },
    {
      "entropy": 5.966246271133423,
      "epoch": 0.30268995338911664,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995889587572366,
      "loss": 5.917,
      "mean_token_accuracy": 0.16286829859018326,
      "num_tokens": 6011523.0,
      "step": 2695
    },
    {
      "entropy": 5.872964239120483,
      "epoch": 0.3032515302970742,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499586529467197,
      "loss": 5.7196,
      "mean_token_accuracy": 0.16652795672416687,
      "num_tokens": 6022169.0,
      "step": 2700
    },
    {
      "entropy": 5.7077781677246096,
      "epoch": 0.30381310720503174,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995840930262351,
      "loss": 5.79,
      "mean_token_accuracy": 0.16523944810032845,
      "num_tokens": 6033393.0,
      "step": 2705
    },
    {
      "entropy": 5.892653083801269,
      "epoch": 0.3043746841129893,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004995816494344287,
      "loss": 5.8429,
      "mean_token_accuracy": 0.15828568935394288,
      "num_tokens": 6045365.0,
      "step": 2710
    },
    {
      "entropy": 5.85001220703125,
      "epoch": 0.30493626102094684,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995791986918554,
      "loss": 5.868,
      "mean_token_accuracy": 0.16434365063905715,
      "num_tokens": 6057482.0,
      "step": 2715
    },
    {
      "entropy": 5.8889617919921875,
      "epoch": 0.30549783792890434,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004995767407985933,
      "loss": 5.782,
      "mean_token_accuracy": 0.16390661150217056,
      "num_tokens": 6068194.0,
      "step": 2720
    },
    {
      "entropy": 5.786661911010742,
      "epoch": 0.3060594148368619,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995742757547207,
      "loss": 5.8248,
      "mean_token_accuracy": 0.1641546294093132,
      "num_tokens": 6080540.0,
      "step": 2725
    },
    {
      "entropy": 5.798826599121094,
      "epoch": 0.30662099174481944,
      "grad_norm": 0.90625,
      "learning_rate": 0.000499571803560316,
      "loss": 5.8653,
      "mean_token_accuracy": 0.16517645716667176,
      "num_tokens": 6092464.0,
      "step": 2730
    },
    {
      "entropy": 5.9122978210449215,
      "epoch": 0.307182568652777,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004995693242154581,
      "loss": 5.8066,
      "mean_token_accuracy": 0.1669014349579811,
      "num_tokens": 6103242.0,
      "step": 2735
    },
    {
      "entropy": 5.813560438156128,
      "epoch": 0.30774414556073454,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004995668377202258,
      "loss": 5.8616,
      "mean_token_accuracy": 0.16158075183629989,
      "num_tokens": 6116383.0,
      "step": 2740
    },
    {
      "entropy": 5.794858694076538,
      "epoch": 0.3083057224686921,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004995643440746982,
      "loss": 5.6608,
      "mean_token_accuracy": 0.17261848896741866,
      "num_tokens": 6126883.0,
      "step": 2745
    },
    {
      "entropy": 5.796203565597534,
      "epoch": 0.30886729937664964,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004995618432789548,
      "loss": 5.8404,
      "mean_token_accuracy": 0.15857291966676712,
      "num_tokens": 6139613.0,
      "step": 2750
    },
    {
      "entropy": 5.854716825485229,
      "epoch": 0.3094288762846072,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004995593353330753,
      "loss": 5.9133,
      "mean_token_accuracy": 0.16079335808753967,
      "num_tokens": 6151950.0,
      "step": 2755
    },
    {
      "entropy": 5.837777042388916,
      "epoch": 0.30999045319256474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004995568202371393,
      "loss": 5.7127,
      "mean_token_accuracy": 0.16686528772115708,
      "num_tokens": 6161563.0,
      "step": 2760
    },
    {
      "entropy": 5.712786054611206,
      "epoch": 0.3105520301005223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995542979912272,
      "loss": 5.7281,
      "mean_token_accuracy": 0.17184682339429855,
      "num_tokens": 6170966.0,
      "step": 2765
    },
    {
      "entropy": 5.817671155929565,
      "epoch": 0.3111136070084798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004995517685954192,
      "loss": 5.798,
      "mean_token_accuracy": 0.16531543880701066,
      "num_tokens": 6182563.0,
      "step": 2770
    },
    {
      "entropy": 5.8397852897644045,
      "epoch": 0.31167518391643734,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004995492320497958,
      "loss": 5.8981,
      "mean_token_accuracy": 0.1579633668065071,
      "num_tokens": 6194495.0,
      "step": 2775
    },
    {
      "entropy": 5.866177606582641,
      "epoch": 0.3122367608243949,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995466883544376,
      "loss": 5.8005,
      "mean_token_accuracy": 0.1617473378777504,
      "num_tokens": 6206594.0,
      "step": 2780
    },
    {
      "entropy": 5.862904596328735,
      "epoch": 0.31279833773235244,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004995441375094258,
      "loss": 5.8319,
      "mean_token_accuracy": 0.15888234823942185,
      "num_tokens": 6218371.0,
      "step": 2785
    },
    {
      "entropy": 5.8452497005462645,
      "epoch": 0.31335991464031,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004995415795148417,
      "loss": 5.7583,
      "mean_token_accuracy": 0.16749636232852935,
      "num_tokens": 6228723.0,
      "step": 2790
    },
    {
      "entropy": 5.830468559265137,
      "epoch": 0.31392149154826754,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004995390143707665,
      "loss": 5.9309,
      "mean_token_accuracy": 0.16207406669855118,
      "num_tokens": 6240535.0,
      "step": 2795
    },
    {
      "entropy": 5.817456340789795,
      "epoch": 0.3144830684562251,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004995364420772821,
      "loss": 5.6419,
      "mean_token_accuracy": 0.17303893119096755,
      "num_tokens": 6250643.0,
      "step": 2800
    },
    {
      "entropy": 5.758491373062133,
      "epoch": 0.31504464536418264,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004995338626344702,
      "loss": 5.7857,
      "mean_token_accuracy": 0.16550625413656234,
      "num_tokens": 6259569.0,
      "step": 2805
    },
    {
      "entropy": 5.837896394729614,
      "epoch": 0.3156062222721402,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499531276042413,
      "loss": 5.9132,
      "mean_token_accuracy": 0.1697901740670204,
      "num_tokens": 6271108.0,
      "step": 2810
    },
    {
      "entropy": 5.893266916275024,
      "epoch": 0.31616779918009774,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004995286823011929,
      "loss": 5.7597,
      "mean_token_accuracy": 0.15869502425193788,
      "num_tokens": 6282352.0,
      "step": 2815
    },
    {
      "entropy": 5.788046741485596,
      "epoch": 0.31672937608805524,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004995260814108926,
      "loss": 5.8766,
      "mean_token_accuracy": 0.16228645592927932,
      "num_tokens": 6294946.0,
      "step": 2820
    },
    {
      "entropy": 5.816434526443482,
      "epoch": 0.3172909529960128,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004995234733715947,
      "loss": 5.6936,
      "mean_token_accuracy": 0.16825819313526152,
      "num_tokens": 6305552.0,
      "step": 2825
    },
    {
      "entropy": 5.727196073532104,
      "epoch": 0.31785252990397034,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004995208581833824,
      "loss": 5.7691,
      "mean_token_accuracy": 0.1745617136359215,
      "num_tokens": 6315815.0,
      "step": 2830
    },
    {
      "entropy": 5.8493071556091305,
      "epoch": 0.3184141068119279,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004995182358463388,
      "loss": 5.7964,
      "mean_token_accuracy": 0.1629154123365879,
      "num_tokens": 6327384.0,
      "step": 2835
    },
    {
      "entropy": 5.843144941329956,
      "epoch": 0.31897568371988544,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004995156063605475,
      "loss": 5.7971,
      "mean_token_accuracy": 0.1567796841263771,
      "num_tokens": 6338083.0,
      "step": 2840
    },
    {
      "entropy": 5.861480331420898,
      "epoch": 0.319537260627843,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004995129697260922,
      "loss": 5.787,
      "mean_token_accuracy": 0.16809219717979432,
      "num_tokens": 6349058.0,
      "step": 2845
    },
    {
      "entropy": 5.789098072052002,
      "epoch": 0.32009883753580054,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004995103259430569,
      "loss": 5.7927,
      "mean_token_accuracy": 0.16856566220521926,
      "num_tokens": 6359655.0,
      "step": 2850
    },
    {
      "entropy": 5.902016878128052,
      "epoch": 0.3206604144437581,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004995076750115258,
      "loss": 5.9143,
      "mean_token_accuracy": 0.16268535405397416,
      "num_tokens": 6372088.0,
      "step": 2855
    },
    {
      "entropy": 5.852352237701416,
      "epoch": 0.32122199135171564,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995050169315832,
      "loss": 5.8272,
      "mean_token_accuracy": 0.16033977046608924,
      "num_tokens": 6382946.0,
      "step": 2860
    },
    {
      "entropy": 5.833240985870361,
      "epoch": 0.32178356825967314,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004995023517033138,
      "loss": 5.831,
      "mean_token_accuracy": 0.1602622725069523,
      "num_tokens": 6392591.0,
      "step": 2865
    },
    {
      "entropy": 5.773062515258789,
      "epoch": 0.3223451451676307,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994996793268024,
      "loss": 5.7582,
      "mean_token_accuracy": 0.16714053303003312,
      "num_tokens": 6403159.0,
      "step": 2870
    },
    {
      "entropy": 5.8594804286956785,
      "epoch": 0.32290672207558824,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994969998021342,
      "loss": 5.7763,
      "mean_token_accuracy": 0.16392116546630858,
      "num_tokens": 6414232.0,
      "step": 2875
    },
    {
      "entropy": 5.80814905166626,
      "epoch": 0.3234682989835458,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004994943131293944,
      "loss": 5.75,
      "mean_token_accuracy": 0.17552235275506972,
      "num_tokens": 6425762.0,
      "step": 2880
    },
    {
      "entropy": 5.751522493362427,
      "epoch": 0.32402987589150334,
      "grad_norm": 1.0,
      "learning_rate": 0.0004994916193086687,
      "loss": 5.8078,
      "mean_token_accuracy": 0.1594655692577362,
      "num_tokens": 6439077.0,
      "step": 2885
    },
    {
      "entropy": 5.91164870262146,
      "epoch": 0.3245914527994609,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004994889183400428,
      "loss": 5.8659,
      "mean_token_accuracy": 0.1664183884859085,
      "num_tokens": 6450425.0,
      "step": 2890
    },
    {
      "entropy": 5.817657470703125,
      "epoch": 0.32515302970741844,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994862102236025,
      "loss": 5.8467,
      "mean_token_accuracy": 0.16085008457303046,
      "num_tokens": 6461347.0,
      "step": 2895
    },
    {
      "entropy": 5.867591190338135,
      "epoch": 0.325714606615376,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994834949594342,
      "loss": 5.8231,
      "mean_token_accuracy": 0.16414998173713685,
      "num_tokens": 6472428.0,
      "step": 2900
    },
    {
      "entropy": 5.853135013580323,
      "epoch": 0.32627618352333354,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004994807725476246,
      "loss": 5.7869,
      "mean_token_accuracy": 0.1745881587266922,
      "num_tokens": 6483393.0,
      "step": 2905
    },
    {
      "entropy": 5.785771131515503,
      "epoch": 0.3268377604312911,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004994780429882599,
      "loss": 5.7757,
      "mean_token_accuracy": 0.1680740788578987,
      "num_tokens": 6493520.0,
      "step": 2910
    },
    {
      "entropy": 5.752371883392334,
      "epoch": 0.3273993373392486,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004994753062814274,
      "loss": 5.6963,
      "mean_token_accuracy": 0.17319443970918655,
      "num_tokens": 6505282.0,
      "step": 2915
    },
    {
      "entropy": 5.91207103729248,
      "epoch": 0.32796091424720614,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004994725624272139,
      "loss": 5.8841,
      "mean_token_accuracy": 0.16525036245584487,
      "num_tokens": 6516808.0,
      "step": 2920
    },
    {
      "entropy": 5.796984434127808,
      "epoch": 0.3285224911551637,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994698114257071,
      "loss": 5.8424,
      "mean_token_accuracy": 0.16388848423957825,
      "num_tokens": 6529184.0,
      "step": 2925
    },
    {
      "entropy": 5.754847717285156,
      "epoch": 0.32908406806312124,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004994670532769944,
      "loss": 5.6797,
      "mean_token_accuracy": 0.17214834541082383,
      "num_tokens": 6539909.0,
      "step": 2930
    },
    {
      "entropy": 5.7354607582092285,
      "epoch": 0.3296456449710788,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004994642879811637,
      "loss": 5.728,
      "mean_token_accuracy": 0.17159772366285325,
      "num_tokens": 6550870.0,
      "step": 2935
    },
    {
      "entropy": 5.773732614517212,
      "epoch": 0.33020722187903634,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499461515538303,
      "loss": 5.7816,
      "mean_token_accuracy": 0.16842659711837768,
      "num_tokens": 6561784.0,
      "step": 2940
    },
    {
      "entropy": 5.744688701629639,
      "epoch": 0.3307687987869939,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994587359485006,
      "loss": 5.7386,
      "mean_token_accuracy": 0.16806551367044448,
      "num_tokens": 6573422.0,
      "step": 2945
    },
    {
      "entropy": 5.877232646942138,
      "epoch": 0.33133037569495144,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000499455949211845,
      "loss": 5.8162,
      "mean_token_accuracy": 0.16845542192459106,
      "num_tokens": 6584565.0,
      "step": 2950
    },
    {
      "entropy": 5.798536968231201,
      "epoch": 0.331891952602909,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994531553284248,
      "loss": 5.7309,
      "mean_token_accuracy": 0.1662512257695198,
      "num_tokens": 6594238.0,
      "step": 2955
    },
    {
      "entropy": 5.757019329071045,
      "epoch": 0.3324535295108665,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994503542983293,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16396647095680236,
      "num_tokens": 6606309.0,
      "step": 2960
    },
    {
      "entropy": 5.9318952560424805,
      "epoch": 0.33301510641882404,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994475461216472,
      "loss": 5.883,
      "mean_token_accuracy": 0.15782105028629304,
      "num_tokens": 6617916.0,
      "step": 2965
    },
    {
      "entropy": 5.816576623916626,
      "epoch": 0.3335766833267816,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004994447307984684,
      "loss": 5.7429,
      "mean_token_accuracy": 0.16411727666854858,
      "num_tokens": 6627951.0,
      "step": 2970
    },
    {
      "entropy": 5.786376428604126,
      "epoch": 0.33413826023473914,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004994419083288823,
      "loss": 5.8053,
      "mean_token_accuracy": 0.16367775201797485,
      "num_tokens": 6639285.0,
      "step": 2975
    },
    {
      "entropy": 5.791878652572632,
      "epoch": 0.3346998371426967,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004994390787129787,
      "loss": 5.6649,
      "mean_token_accuracy": 0.17057127356529236,
      "num_tokens": 6649973.0,
      "step": 2980
    },
    {
      "entropy": 5.695503759384155,
      "epoch": 0.33526141405065424,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004994362419508478,
      "loss": 5.6906,
      "mean_token_accuracy": 0.17208612114191055,
      "num_tokens": 6660625.0,
      "step": 2985
    },
    {
      "entropy": 5.78005313873291,
      "epoch": 0.3358229909586118,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00049943339804258,
      "loss": 5.7787,
      "mean_token_accuracy": 0.16272861510515213,
      "num_tokens": 6673191.0,
      "step": 2990
    },
    {
      "entropy": 5.845469617843628,
      "epoch": 0.33638456786656934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994305469882657,
      "loss": 5.7557,
      "mean_token_accuracy": 0.16925826966762542,
      "num_tokens": 6683993.0,
      "step": 2995
    },
    {
      "entropy": 5.727444458007812,
      "epoch": 0.3369461447745269,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004994276887879957,
      "loss": 5.8048,
      "mean_token_accuracy": 0.16965395361185073,
      "num_tokens": 6694881.0,
      "step": 3000
    },
    {
      "epoch": 0.3369461447745269,
      "eval_entropy": 5.65701306957789,
      "eval_loss": 5.820116996765137,
      "eval_mean_token_accuracy": 0.1707352563135397,
      "eval_num_tokens": 6694881.0,
      "eval_runtime": 23.8089,
      "eval_samples_per_second": 1302.578,
      "eval_steps_per_second": 162.838,
      "step": 3000
    },
    {
      "entropy": 5.8793449878692625,
      "epoch": 0.33750772168248444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004994248234418611,
      "loss": 5.8003,
      "mean_token_accuracy": 0.16803352981805803,
      "num_tokens": 6706648.0,
      "step": 3005
    },
    {
      "entropy": 5.798100757598877,
      "epoch": 0.33806929859044194,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499421950949953,
      "loss": 5.8101,
      "mean_token_accuracy": 0.16942514926195146,
      "num_tokens": 6716858.0,
      "step": 3010
    },
    {
      "entropy": 5.745474720001221,
      "epoch": 0.3386308754983995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000499419071312363,
      "loss": 5.7629,
      "mean_token_accuracy": 0.16839589476585387,
      "num_tokens": 6726558.0,
      "step": 3015
    },
    {
      "entropy": 5.8356531143188475,
      "epoch": 0.33919245240635704,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004994161845291828,
      "loss": 5.7607,
      "mean_token_accuracy": 0.16312356740236283,
      "num_tokens": 6737819.0,
      "step": 3020
    },
    {
      "entropy": 5.769262981414795,
      "epoch": 0.3397540293143146,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004994132906005041,
      "loss": 5.7006,
      "mean_token_accuracy": 0.1680557385087013,
      "num_tokens": 6748593.0,
      "step": 3025
    },
    {
      "entropy": 5.703664064407349,
      "epoch": 0.34031560622227214,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994103895264193,
      "loss": 5.8088,
      "mean_token_accuracy": 0.16019922792911528,
      "num_tokens": 6760631.0,
      "step": 3030
    },
    {
      "entropy": 5.88086051940918,
      "epoch": 0.3408771831302297,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004994074813070207,
      "loss": 5.7241,
      "mean_token_accuracy": 0.16422946602106095,
      "num_tokens": 6772318.0,
      "step": 3035
    },
    {
      "entropy": 5.729892158508301,
      "epoch": 0.34143876003818724,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004994045659424007,
      "loss": 5.7539,
      "mean_token_accuracy": 0.17037225216627122,
      "num_tokens": 6782684.0,
      "step": 3040
    },
    {
      "entropy": 5.785285615921021,
      "epoch": 0.3420003369461448,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004994016434326524,
      "loss": 5.6441,
      "mean_token_accuracy": 0.17063571661710739,
      "num_tokens": 6793895.0,
      "step": 3045
    },
    {
      "entropy": 5.656803035736084,
      "epoch": 0.34256191385410234,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004993987137778687,
      "loss": 5.7016,
      "mean_token_accuracy": 0.16678110808134078,
      "num_tokens": 6804794.0,
      "step": 3050
    },
    {
      "entropy": 5.8625846862792965,
      "epoch": 0.3431234907620599,
      "grad_norm": 0.9375,
      "learning_rate": 0.000499395776978143,
      "loss": 5.8287,
      "mean_token_accuracy": 0.1602445900440216,
      "num_tokens": 6815976.0,
      "step": 3055
    },
    {
      "entropy": 5.857016849517822,
      "epoch": 0.3436850676700174,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004993928330335686,
      "loss": 5.7989,
      "mean_token_accuracy": 0.16048213839530945,
      "num_tokens": 6828027.0,
      "step": 3060
    },
    {
      "entropy": 5.7888294696807865,
      "epoch": 0.34424664457797494,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993898819442395,
      "loss": 5.7864,
      "mean_token_accuracy": 0.15957127138972282,
      "num_tokens": 6839195.0,
      "step": 3065
    },
    {
      "entropy": 5.656088542938233,
      "epoch": 0.3448082214859325,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004993869237102494,
      "loss": 5.6577,
      "mean_token_accuracy": 0.16125886440277098,
      "num_tokens": 6851731.0,
      "step": 3070
    },
    {
      "entropy": 5.826152276992798,
      "epoch": 0.34536979839389004,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004993839583316927,
      "loss": 5.6951,
      "mean_token_accuracy": 0.17295657396316527,
      "num_tokens": 6862916.0,
      "step": 3075
    },
    {
      "entropy": 5.86891393661499,
      "epoch": 0.3459313753018476,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004993809858086638,
      "loss": 5.8666,
      "mean_token_accuracy": 0.15901725590229035,
      "num_tokens": 6876003.0,
      "step": 3080
    },
    {
      "entropy": 5.779653787612915,
      "epoch": 0.34649295220980514,
      "grad_norm": 1.0,
      "learning_rate": 0.0004993780061412571,
      "loss": 5.8486,
      "mean_token_accuracy": 0.16634225398302077,
      "num_tokens": 6886040.0,
      "step": 3085
    },
    {
      "entropy": 5.829333305358887,
      "epoch": 0.3470545291177627,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004993750193295678,
      "loss": 5.7544,
      "mean_token_accuracy": 0.1724129244685173,
      "num_tokens": 6896758.0,
      "step": 3090
    },
    {
      "entropy": 5.782223081588745,
      "epoch": 0.34761610602572024,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004993720253736909,
      "loss": 5.8064,
      "mean_token_accuracy": 0.16331667602062225,
      "num_tokens": 6907287.0,
      "step": 3095
    },
    {
      "entropy": 5.780876302719117,
      "epoch": 0.3481776829336778,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004993690242737216,
      "loss": 5.8374,
      "mean_token_accuracy": 0.1603430137038231,
      "num_tokens": 6918992.0,
      "step": 3100
    },
    {
      "entropy": 5.937829971313477,
      "epoch": 0.3487392598416353,
      "grad_norm": 0.84765625,
      "learning_rate": 0.0004993660160297555,
      "loss": 5.8355,
      "mean_token_accuracy": 0.16085187941789628,
      "num_tokens": 6931493.0,
      "step": 3105
    },
    {
      "entropy": 5.688720989227295,
      "epoch": 0.34930083674959284,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004993630006418885,
      "loss": 5.6717,
      "mean_token_accuracy": 0.1681598588824272,
      "num_tokens": 6941807.0,
      "step": 3110
    },
    {
      "entropy": 5.8561256408691404,
      "epoch": 0.3498624136575504,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993599781102166,
      "loss": 5.8643,
      "mean_token_accuracy": 0.16247122436761857,
      "num_tokens": 6953436.0,
      "step": 3115
    },
    {
      "entropy": 6.006970691680908,
      "epoch": 0.35042399056550794,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993569484348359,
      "loss": 5.9725,
      "mean_token_accuracy": 0.15798764154314995,
      "num_tokens": 6965247.0,
      "step": 3120
    },
    {
      "entropy": 5.787929534912109,
      "epoch": 0.3509855674734655,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004993539116158429,
      "loss": 5.7537,
      "mean_token_accuracy": 0.17097637355327605,
      "num_tokens": 6976605.0,
      "step": 3125
    },
    {
      "entropy": 5.757473373413086,
      "epoch": 0.35154714438142304,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993508676533346,
      "loss": 5.7184,
      "mean_token_accuracy": 0.17815156131982804,
      "num_tokens": 6986205.0,
      "step": 3130
    },
    {
      "entropy": 5.6938660621643065,
      "epoch": 0.3521087212893806,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993478165474074,
      "loss": 5.6817,
      "mean_token_accuracy": 0.1697208508849144,
      "num_tokens": 6997699.0,
      "step": 3135
    },
    {
      "entropy": 5.692649507522583,
      "epoch": 0.35267029819733814,
      "grad_norm": 0.875,
      "learning_rate": 0.0004993447582981588,
      "loss": 5.704,
      "mean_token_accuracy": 0.16726200580596923,
      "num_tokens": 7008968.0,
      "step": 3140
    },
    {
      "entropy": 5.863374423980713,
      "epoch": 0.3532318751052957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000499341692905686,
      "loss": 5.7273,
      "mean_token_accuracy": 0.16293082088232042,
      "num_tokens": 7020633.0,
      "step": 3145
    },
    {
      "entropy": 5.705703496932983,
      "epoch": 0.35379345201325324,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004993386203700868,
      "loss": 5.6435,
      "mean_token_accuracy": 0.17389080673456192,
      "num_tokens": 7031187.0,
      "step": 3150
    },
    {
      "entropy": 5.754469776153565,
      "epoch": 0.35435502892121074,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004993355406914589,
      "loss": 5.884,
      "mean_token_accuracy": 0.16147221475839615,
      "num_tokens": 7042662.0,
      "step": 3155
    },
    {
      "entropy": 5.962271881103516,
      "epoch": 0.3549166058291683,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004993324538699003,
      "loss": 5.8627,
      "mean_token_accuracy": 0.16751048862934112,
      "num_tokens": 7054297.0,
      "step": 3160
    },
    {
      "entropy": 5.819624948501587,
      "epoch": 0.35547818273712584,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004993293599055094,
      "loss": 5.7195,
      "mean_token_accuracy": 0.16985224634408952,
      "num_tokens": 7064961.0,
      "step": 3165
    },
    {
      "entropy": 5.817587852478027,
      "epoch": 0.3560397596450834,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993262587983848,
      "loss": 5.8613,
      "mean_token_accuracy": 0.16336070150136947,
      "num_tokens": 7076857.0,
      "step": 3170
    },
    {
      "entropy": 5.782158851623535,
      "epoch": 0.35660133655304094,
      "grad_norm": 1.046875,
      "learning_rate": 0.000499323150548625,
      "loss": 5.7306,
      "mean_token_accuracy": 0.1675140678882599,
      "num_tokens": 7088926.0,
      "step": 3175
    },
    {
      "entropy": 5.681119823455811,
      "epoch": 0.3571629134609985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004993200351563291,
      "loss": 5.6683,
      "mean_token_accuracy": 0.16968711614608764,
      "num_tokens": 7099478.0,
      "step": 3180
    },
    {
      "entropy": 5.785163640975952,
      "epoch": 0.35772449036895604,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004993169126215962,
      "loss": 5.7,
      "mean_token_accuracy": 0.17341631054878234,
      "num_tokens": 7111776.0,
      "step": 3185
    },
    {
      "entropy": 5.885181474685669,
      "epoch": 0.3582860672769136,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004993137829445259,
      "loss": 5.8006,
      "mean_token_accuracy": 0.16299712061882018,
      "num_tokens": 7123084.0,
      "step": 3190
    },
    {
      "entropy": 5.737665462493896,
      "epoch": 0.35884764418487114,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004993106461252177,
      "loss": 5.7486,
      "mean_token_accuracy": 0.16912526786327362,
      "num_tokens": 7134973.0,
      "step": 3195
    },
    {
      "entropy": 5.746781492233277,
      "epoch": 0.3594092210928287,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004993075021637716,
      "loss": 5.7815,
      "mean_token_accuracy": 0.16657314747571944,
      "num_tokens": 7147227.0,
      "step": 3200
    },
    {
      "entropy": 5.822588920593262,
      "epoch": 0.3599707980007862,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004993043510602875,
      "loss": 5.7609,
      "mean_token_accuracy": 0.16496573388576508,
      "num_tokens": 7158201.0,
      "step": 3205
    },
    {
      "entropy": 5.741505098342896,
      "epoch": 0.36053237490874374,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000499301192814866,
      "loss": 5.6649,
      "mean_token_accuracy": 0.17205592393875122,
      "num_tokens": 7168909.0,
      "step": 3210
    },
    {
      "entropy": 5.8722270965576175,
      "epoch": 0.3610939518167013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992980274276074,
      "loss": 5.838,
      "mean_token_accuracy": 0.16101012676954268,
      "num_tokens": 7179877.0,
      "step": 3215
    },
    {
      "entropy": 5.715766572952271,
      "epoch": 0.36165552872465884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992948548986128,
      "loss": 5.7432,
      "mean_token_accuracy": 0.16797278076410294,
      "num_tokens": 7190823.0,
      "step": 3220
    },
    {
      "entropy": 5.754311180114746,
      "epoch": 0.3622171056326164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992916752279829,
      "loss": 5.6703,
      "mean_token_accuracy": 0.17095502763986586,
      "num_tokens": 7201700.0,
      "step": 3225
    },
    {
      "entropy": 5.784860563278198,
      "epoch": 0.36277868254057394,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499288488415819,
      "loss": 5.7191,
      "mean_token_accuracy": 0.17327634394168853,
      "num_tokens": 7213618.0,
      "step": 3230
    },
    {
      "entropy": 5.902436637878418,
      "epoch": 0.3633402594485315,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992852944622229,
      "loss": 5.8952,
      "mean_token_accuracy": 0.1662729024887085,
      "num_tokens": 7226207.0,
      "step": 3235
    },
    {
      "entropy": 5.705463886260986,
      "epoch": 0.36390183635648904,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004992820933672958,
      "loss": 5.7614,
      "mean_token_accuracy": 0.17185945808887482,
      "num_tokens": 7237545.0,
      "step": 3240
    },
    {
      "entropy": 5.740161943435669,
      "epoch": 0.3644634132644466,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004992788851311399,
      "loss": 5.7228,
      "mean_token_accuracy": 0.1698360711336136,
      "num_tokens": 7247799.0,
      "step": 3245
    },
    {
      "entropy": 5.767627573013305,
      "epoch": 0.3650249901724041,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992756697538574,
      "loss": 5.6923,
      "mean_token_accuracy": 0.16587659120559692,
      "num_tokens": 7260229.0,
      "step": 3250
    },
    {
      "entropy": 5.788318347930908,
      "epoch": 0.36558656708036164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004992724472355505,
      "loss": 5.8105,
      "mean_token_accuracy": 0.16308242380619048,
      "num_tokens": 7271457.0,
      "step": 3255
    },
    {
      "entropy": 5.810742330551148,
      "epoch": 0.3661481439883192,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992692175763219,
      "loss": 5.7089,
      "mean_token_accuracy": 0.16574329882860184,
      "num_tokens": 7282124.0,
      "step": 3260
    },
    {
      "entropy": 5.7773487091064455,
      "epoch": 0.36670972089627674,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992659807762745,
      "loss": 5.8173,
      "mean_token_accuracy": 0.1679944857954979,
      "num_tokens": 7292981.0,
      "step": 3265
    },
    {
      "entropy": 6.030627107620239,
      "epoch": 0.3672712978042343,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004992627368355112,
      "loss": 5.9661,
      "mean_token_accuracy": 0.16412724256515504,
      "num_tokens": 7304591.0,
      "step": 3270
    },
    {
      "entropy": 5.795241117477417,
      "epoch": 0.36783287471219184,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004992594857541355,
      "loss": 5.7114,
      "mean_token_accuracy": 0.16717362105846406,
      "num_tokens": 7315918.0,
      "step": 3275
    },
    {
      "entropy": 5.755579566955566,
      "epoch": 0.3683944516201494,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992562275322506,
      "loss": 5.8109,
      "mean_token_accuracy": 0.16396376341581345,
      "num_tokens": 7327725.0,
      "step": 3280
    },
    {
      "entropy": 5.895589828491211,
      "epoch": 0.36895602852810694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004992529621699606,
      "loss": 5.8149,
      "mean_token_accuracy": 0.16023087799549102,
      "num_tokens": 7339787.0,
      "step": 3285
    },
    {
      "entropy": 5.8424718379974365,
      "epoch": 0.3695176054360645,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004992496896673692,
      "loss": 5.8047,
      "mean_token_accuracy": 0.1638472482562065,
      "num_tokens": 7351284.0,
      "step": 3290
    },
    {
      "entropy": 5.665236949920654,
      "epoch": 0.37007918234402204,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004992464100245809,
      "loss": 5.6204,
      "mean_token_accuracy": 0.1754314810037613,
      "num_tokens": 7362105.0,
      "step": 3295
    },
    {
      "entropy": 5.731363296508789,
      "epoch": 0.37064075925197953,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004992431232416997,
      "loss": 5.7246,
      "mean_token_accuracy": 0.16994064152240754,
      "num_tokens": 7375204.0,
      "step": 3300
    },
    {
      "entropy": 5.95510196685791,
      "epoch": 0.3712023361599371,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004992398293188307,
      "loss": 5.8863,
      "mean_token_accuracy": 0.15913688391447067,
      "num_tokens": 7385853.0,
      "step": 3305
    },
    {
      "entropy": 5.741273689270019,
      "epoch": 0.37176391306789464,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004992365282560784,
      "loss": 5.7394,
      "mean_token_accuracy": 0.1607675224542618,
      "num_tokens": 7397459.0,
      "step": 3310
    },
    {
      "entropy": 5.77800145149231,
      "epoch": 0.3723254899758522,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004992332200535483,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1667885437607765,
      "num_tokens": 7408996.0,
      "step": 3315
    },
    {
      "entropy": 5.831015634536743,
      "epoch": 0.37288706688380974,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992299047113452,
      "loss": 5.8217,
      "mean_token_accuracy": 0.155822491645813,
      "num_tokens": 7419938.0,
      "step": 3320
    },
    {
      "entropy": 5.803605318069458,
      "epoch": 0.3734486437917673,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004992265822295752,
      "loss": 5.8296,
      "mean_token_accuracy": 0.16146015003323555,
      "num_tokens": 7430181.0,
      "step": 3325
    },
    {
      "entropy": 5.8900665760040285,
      "epoch": 0.37401022069972484,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004992232526083438,
      "loss": 5.7822,
      "mean_token_accuracy": 0.16336511075496674,
      "num_tokens": 7440774.0,
      "step": 3330
    },
    {
      "entropy": 5.755934000015259,
      "epoch": 0.3745717976076824,
      "grad_norm": 0.8984375,
      "learning_rate": 0.000499219915847757,
      "loss": 5.7053,
      "mean_token_accuracy": 0.16726386696100234,
      "num_tokens": 7451744.0,
      "step": 3335
    },
    {
      "entropy": 5.6385719776153564,
      "epoch": 0.37513337451563994,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004992165719479211,
      "loss": 5.737,
      "mean_token_accuracy": 0.16925416588783265,
      "num_tokens": 7463731.0,
      "step": 3340
    },
    {
      "entropy": 5.88433427810669,
      "epoch": 0.37569495142359743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004992132209089427,
      "loss": 5.7373,
      "mean_token_accuracy": 0.16732349544763564,
      "num_tokens": 7474392.0,
      "step": 3345
    },
    {
      "entropy": 5.681453609466553,
      "epoch": 0.376256528331555,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004992098627309283,
      "loss": 5.6643,
      "mean_token_accuracy": 0.1815314844250679,
      "num_tokens": 7484860.0,
      "step": 3350
    },
    {
      "entropy": 5.689245319366455,
      "epoch": 0.37681810523951254,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000499206497413985,
      "loss": 5.7166,
      "mean_token_accuracy": 0.17443938702344894,
      "num_tokens": 7495877.0,
      "step": 3355
    },
    {
      "entropy": 5.856110763549805,
      "epoch": 0.3773796821474701,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004992031249582197,
      "loss": 5.7726,
      "mean_token_accuracy": 0.17041053920984267,
      "num_tokens": 7506879.0,
      "step": 3360
    },
    {
      "entropy": 5.711837959289551,
      "epoch": 0.37794125905542764,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004991997453637401,
      "loss": 5.613,
      "mean_token_accuracy": 0.1775125339627266,
      "num_tokens": 7517405.0,
      "step": 3365
    },
    {
      "entropy": 5.734726905822754,
      "epoch": 0.3785028359633852,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991963586306536,
      "loss": 5.7018,
      "mean_token_accuracy": 0.1640342578291893,
      "num_tokens": 7528832.0,
      "step": 3370
    },
    {
      "entropy": 5.72209587097168,
      "epoch": 0.37906441287134274,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991929647590682,
      "loss": 5.6703,
      "mean_token_accuracy": 0.16398927569389343,
      "num_tokens": 7541048.0,
      "step": 3375
    },
    {
      "entropy": 5.757838535308838,
      "epoch": 0.3796259897793003,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004991895637490917,
      "loss": 5.7037,
      "mean_token_accuracy": 0.1725490927696228,
      "num_tokens": 7551849.0,
      "step": 3380
    },
    {
      "entropy": 5.652497720718384,
      "epoch": 0.38018756668725784,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004991861556008326,
      "loss": 5.7193,
      "mean_token_accuracy": 0.1761649042367935,
      "num_tokens": 7563148.0,
      "step": 3385
    },
    {
      "entropy": 5.884858274459839,
      "epoch": 0.3807491435952154,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004991827403143992,
      "loss": 5.8977,
      "mean_token_accuracy": 0.16204432994127274,
      "num_tokens": 7575119.0,
      "step": 3390
    },
    {
      "entropy": 5.852141046524048,
      "epoch": 0.3813107205031729,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004991793178899006,
      "loss": 5.7401,
      "mean_token_accuracy": 0.1687777116894722,
      "num_tokens": 7585899.0,
      "step": 3395
    },
    {
      "entropy": 5.784982538223266,
      "epoch": 0.38187229741113043,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991758883274455,
      "loss": 5.6428,
      "mean_token_accuracy": 0.1756828472018242,
      "num_tokens": 7598938.0,
      "step": 3400
    },
    {
      "entropy": 5.583678531646728,
      "epoch": 0.382433874319088,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004991724516271432,
      "loss": 5.7322,
      "mean_token_accuracy": 0.16203658878803254,
      "num_tokens": 7610087.0,
      "step": 3405
    },
    {
      "entropy": 5.985393333435058,
      "epoch": 0.38299545122704554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991690077891032,
      "loss": 5.8578,
      "mean_token_accuracy": 0.16392939388751984,
      "num_tokens": 7622186.0,
      "step": 3410
    },
    {
      "entropy": 5.803608083724976,
      "epoch": 0.3835570281350031,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991655568134348,
      "loss": 5.7918,
      "mean_token_accuracy": 0.16341013163328172,
      "num_tokens": 7633501.0,
      "step": 3415
    },
    {
      "entropy": 5.736133146286011,
      "epoch": 0.38411860504296064,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004991620987002483,
      "loss": 5.7361,
      "mean_token_accuracy": 0.16523168087005616,
      "num_tokens": 7643656.0,
      "step": 3420
    },
    {
      "entropy": 5.838760137557983,
      "epoch": 0.3846801819509182,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004991586334496536,
      "loss": 5.7407,
      "mean_token_accuracy": 0.16602644175291062,
      "num_tokens": 7654565.0,
      "step": 3425
    },
    {
      "entropy": 5.770573663711548,
      "epoch": 0.38524175885887574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004991551610617611,
      "loss": 5.6557,
      "mean_token_accuracy": 0.17656644880771638,
      "num_tokens": 7664725.0,
      "step": 3430
    },
    {
      "entropy": 5.6511670589447025,
      "epoch": 0.3858033357668333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991516815366812,
      "loss": 5.743,
      "mean_token_accuracy": 0.17064667344093323,
      "num_tokens": 7674772.0,
      "step": 3435
    },
    {
      "entropy": 5.858159780502319,
      "epoch": 0.38636491267479084,
      "grad_norm": 1.015625,
      "learning_rate": 0.000499148194874525,
      "loss": 5.7884,
      "mean_token_accuracy": 0.16303544789552687,
      "num_tokens": 7686488.0,
      "step": 3440
    },
    {
      "entropy": 5.850274276733399,
      "epoch": 0.38692648958274833,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004991447010754032,
      "loss": 5.7792,
      "mean_token_accuracy": 0.16665342897176744,
      "num_tokens": 7698842.0,
      "step": 3445
    },
    {
      "entropy": 5.664803981781006,
      "epoch": 0.3874880664907059,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004991412001394273,
      "loss": 5.5643,
      "mean_token_accuracy": 0.17148852497339248,
      "num_tokens": 7709255.0,
      "step": 3450
    },
    {
      "entropy": 5.558999633789062,
      "epoch": 0.38804964339866344,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004991376920667086,
      "loss": 5.6258,
      "mean_token_accuracy": 0.17108083218336106,
      "num_tokens": 7720182.0,
      "step": 3455
    },
    {
      "entropy": 5.7312616348266605,
      "epoch": 0.388611220306621,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004991341768573589,
      "loss": 5.6459,
      "mean_token_accuracy": 0.1802293837070465,
      "num_tokens": 7730944.0,
      "step": 3460
    },
    {
      "entropy": 5.7747574806213375,
      "epoch": 0.38917279721457854,
      "grad_norm": 1.03125,
      "learning_rate": 0.00049913065451149,
      "loss": 5.7193,
      "mean_token_accuracy": 0.17277462929487228,
      "num_tokens": 7741686.0,
      "step": 3465
    },
    {
      "entropy": 5.773808717727661,
      "epoch": 0.3897343741225361,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004991271250292141,
      "loss": 5.7203,
      "mean_token_accuracy": 0.17095136046409606,
      "num_tokens": 7752245.0,
      "step": 3470
    },
    {
      "entropy": 5.655320549011231,
      "epoch": 0.39029595103049364,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004991235884106437,
      "loss": 5.687,
      "mean_token_accuracy": 0.16102302372455596,
      "num_tokens": 7762622.0,
      "step": 3475
    },
    {
      "entropy": 5.8449396133422855,
      "epoch": 0.3908575279384512,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004991200446558913,
      "loss": 5.8209,
      "mean_token_accuracy": 0.16779100745916367,
      "num_tokens": 7773550.0,
      "step": 3480
    },
    {
      "entropy": 5.72516508102417,
      "epoch": 0.39141910484640874,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004991164937650698,
      "loss": 5.7013,
      "mean_token_accuracy": 0.16410722583532333,
      "num_tokens": 7784988.0,
      "step": 3485
    },
    {
      "entropy": 5.848622608184814,
      "epoch": 0.39198068175436623,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004991129357382922,
      "loss": 5.7906,
      "mean_token_accuracy": 0.16952809393405915,
      "num_tokens": 7797706.0,
      "step": 3490
    },
    {
      "entropy": 5.740201711654663,
      "epoch": 0.3925422586623238,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004991093705756718,
      "loss": 5.6706,
      "mean_token_accuracy": 0.17071947157382966,
      "num_tokens": 7808960.0,
      "step": 3495
    },
    {
      "entropy": 5.65643424987793,
      "epoch": 0.39310383557028133,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004991057982773221,
      "loss": 5.637,
      "mean_token_accuracy": 0.17911625653505325,
      "num_tokens": 7819271.0,
      "step": 3500
    },
    {
      "entropy": 5.754959487915039,
      "epoch": 0.3936654124782389,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004991022188433569,
      "loss": 5.7535,
      "mean_token_accuracy": 0.17093886882066728,
      "num_tokens": 7832257.0,
      "step": 3505
    },
    {
      "entropy": 5.84992356300354,
      "epoch": 0.39422698938619644,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049909863227389,
      "loss": 5.7813,
      "mean_token_accuracy": 0.16372494399547577,
      "num_tokens": 7842684.0,
      "step": 3510
    },
    {
      "entropy": 5.767809677124023,
      "epoch": 0.394788566294154,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004990950385690358,
      "loss": 5.7847,
      "mean_token_accuracy": 0.1649763450026512,
      "num_tokens": 7855031.0,
      "step": 3515
    },
    {
      "entropy": 5.677387332916259,
      "epoch": 0.39535014320211154,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004990914377289088,
      "loss": 5.5928,
      "mean_token_accuracy": 0.16859776824712752,
      "num_tokens": 7865291.0,
      "step": 3520
    },
    {
      "entropy": 5.653917551040649,
      "epoch": 0.3959117201100691,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990878297536233,
      "loss": 5.5564,
      "mean_token_accuracy": 0.1708521917462349,
      "num_tokens": 7876401.0,
      "step": 3525
    },
    {
      "entropy": 5.661824321746826,
      "epoch": 0.39647329701802664,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004990842146432946,
      "loss": 5.6044,
      "mean_token_accuracy": 0.1747349366545677,
      "num_tokens": 7887787.0,
      "step": 3530
    },
    {
      "entropy": 5.622909927368164,
      "epoch": 0.3970348739259842,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004990805923980375,
      "loss": 5.5941,
      "mean_token_accuracy": 0.17232101112604142,
      "num_tokens": 7898550.0,
      "step": 3535
    },
    {
      "entropy": 5.704342174530029,
      "epoch": 0.3975964508339417,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004990769630179673,
      "loss": 5.7573,
      "mean_token_accuracy": 0.16802939623594285,
      "num_tokens": 7909254.0,
      "step": 3540
    },
    {
      "entropy": 5.914259004592895,
      "epoch": 0.39815802774189923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004990733265032,
      "loss": 5.7816,
      "mean_token_accuracy": 0.17458967864513397,
      "num_tokens": 7920164.0,
      "step": 3545
    },
    {
      "entropy": 5.654784154891968,
      "epoch": 0.3987196046498568,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004990696828538508,
      "loss": 5.7188,
      "mean_token_accuracy": 0.1621084302663803,
      "num_tokens": 7931390.0,
      "step": 3550
    },
    {
      "entropy": 5.802833080291748,
      "epoch": 0.39928118155781434,
      "grad_norm": 0.921875,
      "learning_rate": 0.000499066032070036,
      "loss": 5.7305,
      "mean_token_accuracy": 0.16413189023733138,
      "num_tokens": 7943765.0,
      "step": 3555
    },
    {
      "entropy": 5.657956171035766,
      "epoch": 0.3998427584657719,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990623741518719,
      "loss": 5.5689,
      "mean_token_accuracy": 0.17428297698497772,
      "num_tokens": 7954621.0,
      "step": 3560
    },
    {
      "entropy": 5.679213142395019,
      "epoch": 0.40040433537372944,
      "grad_norm": 0.984375,
      "learning_rate": 0.000499058709099475,
      "loss": 5.777,
      "mean_token_accuracy": 0.16626680940389632,
      "num_tokens": 7966063.0,
      "step": 3565
    },
    {
      "entropy": 5.870148324966431,
      "epoch": 0.400965912281687,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990550369129618,
      "loss": 5.7008,
      "mean_token_accuracy": 0.17148687392473222,
      "num_tokens": 7977391.0,
      "step": 3570
    },
    {
      "entropy": 5.764255523681641,
      "epoch": 0.40152748918964454,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004990513575924494,
      "loss": 5.8189,
      "mean_token_accuracy": 0.16532659083604812,
      "num_tokens": 7988786.0,
      "step": 3575
    },
    {
      "entropy": 5.754783296585083,
      "epoch": 0.4020890660976021,
      "grad_norm": 0.859375,
      "learning_rate": 0.0004990476711380548,
      "loss": 5.5956,
      "mean_token_accuracy": 0.16935015022754668,
      "num_tokens": 8000848.0,
      "step": 3580
    },
    {
      "entropy": 5.7061192989349365,
      "epoch": 0.40265064300555964,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004990439775498954,
      "loss": 5.7624,
      "mean_token_accuracy": 0.16678465008735657,
      "num_tokens": 8012621.0,
      "step": 3585
    },
    {
      "entropy": 5.934963750839233,
      "epoch": 0.40321221991351713,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000499040276828089,
      "loss": 5.8086,
      "mean_token_accuracy": 0.16449498683214187,
      "num_tokens": 8023599.0,
      "step": 3590
    },
    {
      "entropy": 5.575307416915893,
      "epoch": 0.4037737968214747,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004990365689727533,
      "loss": 5.6186,
      "mean_token_accuracy": 0.17472917288541795,
      "num_tokens": 8037151.0,
      "step": 3595
    },
    {
      "entropy": 5.750340843200684,
      "epoch": 0.40433537372943223,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004990328539840062,
      "loss": 5.82,
      "mean_token_accuracy": 0.16475343853235244,
      "num_tokens": 8048606.0,
      "step": 3600
    },
    {
      "entropy": 5.665721797943116,
      "epoch": 0.4048969506373898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004990291318619663,
      "loss": 5.5989,
      "mean_token_accuracy": 0.18224734216928482,
      "num_tokens": 8058430.0,
      "step": 3605
    },
    {
      "entropy": 5.650776767730713,
      "epoch": 0.40545852754534734,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990254026067518,
      "loss": 5.6083,
      "mean_token_accuracy": 0.17226047515869142,
      "num_tokens": 8069554.0,
      "step": 3610
    },
    {
      "entropy": 5.819836139678955,
      "epoch": 0.4060201044533049,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004990216662184817,
      "loss": 5.7724,
      "mean_token_accuracy": 0.16451630741357803,
      "num_tokens": 8082131.0,
      "step": 3615
    },
    {
      "entropy": 5.724734497070313,
      "epoch": 0.40658168136126244,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990179226972749,
      "loss": 5.6073,
      "mean_token_accuracy": 0.1736402168869972,
      "num_tokens": 8093454.0,
      "step": 3620
    },
    {
      "entropy": 5.808397531509399,
      "epoch": 0.40714325826922,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004990141720432504,
      "loss": 5.8744,
      "mean_token_accuracy": 0.15492106080055237,
      "num_tokens": 8105502.0,
      "step": 3625
    },
    {
      "entropy": 5.692808961868286,
      "epoch": 0.40770483517717754,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004990104142565279,
      "loss": 5.7399,
      "mean_token_accuracy": 0.16389643847942353,
      "num_tokens": 8116524.0,
      "step": 3630
    },
    {
      "entropy": 5.854559993743896,
      "epoch": 0.40826641208513503,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000499006649337227,
      "loss": 5.7235,
      "mean_token_accuracy": 0.1744697853922844,
      "num_tokens": 8126949.0,
      "step": 3635
    },
    {
      "entropy": 5.699834728240967,
      "epoch": 0.4088279889930926,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004990028772854674,
      "loss": 5.6694,
      "mean_token_accuracy": 0.17251116931438445,
      "num_tokens": 8137346.0,
      "step": 3640
    },
    {
      "entropy": 5.6351542472839355,
      "epoch": 0.40938956590105013,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004989990981013693,
      "loss": 5.7066,
      "mean_token_accuracy": 0.16331776827573777,
      "num_tokens": 8148499.0,
      "step": 3645
    },
    {
      "entropy": 5.808590126037598,
      "epoch": 0.4099511428090077,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000498995311785053,
      "loss": 5.746,
      "mean_token_accuracy": 0.17033252716064454,
      "num_tokens": 8159645.0,
      "step": 3650
    },
    {
      "entropy": 5.726710510253906,
      "epoch": 0.41051271971696524,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004989915183366392,
      "loss": 5.6098,
      "mean_token_accuracy": 0.17256862074136733,
      "num_tokens": 8171033.0,
      "step": 3655
    },
    {
      "entropy": 5.6207366466522215,
      "epoch": 0.4110742966249228,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989877177562485,
      "loss": 5.6777,
      "mean_token_accuracy": 0.16755093336105348,
      "num_tokens": 8181582.0,
      "step": 3660
    },
    {
      "entropy": 5.655508375167846,
      "epoch": 0.41163587353288034,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498983910044002,
      "loss": 5.59,
      "mean_token_accuracy": 0.17049358934164047,
      "num_tokens": 8192432.0,
      "step": 3665
    },
    {
      "entropy": 5.824843883514404,
      "epoch": 0.4121974504408379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004989800952000209,
      "loss": 5.7591,
      "mean_token_accuracy": 0.1676687553524971,
      "num_tokens": 8203995.0,
      "step": 3670
    },
    {
      "entropy": 5.652281475067139,
      "epoch": 0.41275902734879544,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004989762732244268,
      "loss": 5.6125,
      "mean_token_accuracy": 0.1731998696923256,
      "num_tokens": 8214323.0,
      "step": 3675
    },
    {
      "entropy": 5.620546817779541,
      "epoch": 0.413320604256753,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989724441173413,
      "loss": 5.6724,
      "mean_token_accuracy": 0.16555332243442536,
      "num_tokens": 8226285.0,
      "step": 3680
    },
    {
      "entropy": 5.8974133968353275,
      "epoch": 0.4138821811647105,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989686078788862,
      "loss": 5.8429,
      "mean_token_accuracy": 0.16402119845151902,
      "num_tokens": 8237961.0,
      "step": 3685
    },
    {
      "entropy": 5.689703941345215,
      "epoch": 0.41444375807266803,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004989647645091839,
      "loss": 5.6038,
      "mean_token_accuracy": 0.17371499240398408,
      "num_tokens": 8248463.0,
      "step": 3690
    },
    {
      "entropy": 5.722857522964477,
      "epoch": 0.4150053349806256,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004989609140083565,
      "loss": 5.7313,
      "mean_token_accuracy": 0.16978170573711396,
      "num_tokens": 8260342.0,
      "step": 3695
    },
    {
      "entropy": 5.804967403411865,
      "epoch": 0.41556691188858313,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989570563765269,
      "loss": 5.8475,
      "mean_token_accuracy": 0.16731488555669785,
      "num_tokens": 8272055.0,
      "step": 3700
    },
    {
      "entropy": 5.939024829864502,
      "epoch": 0.4161284887965407,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004989531916138176,
      "loss": 5.7842,
      "mean_token_accuracy": 0.1630214974284172,
      "num_tokens": 8282129.0,
      "step": 3705
    },
    {
      "entropy": 5.787438821792603,
      "epoch": 0.41669006570449824,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989493197203519,
      "loss": 5.7883,
      "mean_token_accuracy": 0.16770507395267487,
      "num_tokens": 8294663.0,
      "step": 3710
    },
    {
      "entropy": 5.706148481369018,
      "epoch": 0.4172516426124558,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000498945440696253,
      "loss": 5.5959,
      "mean_token_accuracy": 0.17472037374973298,
      "num_tokens": 8305882.0,
      "step": 3715
    },
    {
      "entropy": 5.691284847259522,
      "epoch": 0.41781321952041334,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989415545416444,
      "loss": 5.685,
      "mean_token_accuracy": 0.17402007728815078,
      "num_tokens": 8316604.0,
      "step": 3720
    },
    {
      "entropy": 5.780489110946656,
      "epoch": 0.4183747964283709,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004989376612566498,
      "loss": 5.7818,
      "mean_token_accuracy": 0.16784609109163284,
      "num_tokens": 8328126.0,
      "step": 3725
    },
    {
      "entropy": 5.749281978607177,
      "epoch": 0.4189363733363284,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004989337608413933,
      "loss": 5.6953,
      "mean_token_accuracy": 0.1748053789138794,
      "num_tokens": 8339050.0,
      "step": 3730
    },
    {
      "entropy": 5.6462561130523685,
      "epoch": 0.41949795024428593,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004989298532959989,
      "loss": 5.5442,
      "mean_token_accuracy": 0.17576915472745896,
      "num_tokens": 8350216.0,
      "step": 3735
    },
    {
      "entropy": 5.702535390853882,
      "epoch": 0.4200595271522435,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004989259386205912,
      "loss": 5.6931,
      "mean_token_accuracy": 0.17227850407361983,
      "num_tokens": 8362311.0,
      "step": 3740
    },
    {
      "entropy": 5.790135145187378,
      "epoch": 0.42062110406020103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004989220168152948,
      "loss": 5.8693,
      "mean_token_accuracy": 0.1704411879181862,
      "num_tokens": 8373751.0,
      "step": 3745
    },
    {
      "entropy": 5.7706399917602536,
      "epoch": 0.4211826809681586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004989180878802346,
      "loss": 5.7031,
      "mean_token_accuracy": 0.1687019571661949,
      "num_tokens": 8384932.0,
      "step": 3750
    },
    {
      "entropy": 5.755870008468628,
      "epoch": 0.42174425787611614,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004989141518155354,
      "loss": 5.656,
      "mean_token_accuracy": 0.17309170961380005,
      "num_tokens": 8394973.0,
      "step": 3755
    },
    {
      "entropy": 5.730078220367432,
      "epoch": 0.4223058347840737,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004989102086213229,
      "loss": 5.7361,
      "mean_token_accuracy": 0.17052859514951707,
      "num_tokens": 8406590.0,
      "step": 3760
    },
    {
      "entropy": 5.696905422210693,
      "epoch": 0.42286741169203124,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004989062582977225,
      "loss": 5.6582,
      "mean_token_accuracy": 0.16957752257585526,
      "num_tokens": 8417127.0,
      "step": 3765
    },
    {
      "entropy": 5.767214012145996,
      "epoch": 0.4234289885999888,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004989023008448599,
      "loss": 5.7486,
      "mean_token_accuracy": 0.17173793464899062,
      "num_tokens": 8429442.0,
      "step": 3770
    },
    {
      "entropy": 5.71106481552124,
      "epoch": 0.42399056550794634,
      "grad_norm": 1.0,
      "learning_rate": 0.0004988983362628614,
      "loss": 5.6708,
      "mean_token_accuracy": 0.17283091247081755,
      "num_tokens": 8439375.0,
      "step": 3775
    },
    {
      "entropy": 5.6981981754302975,
      "epoch": 0.42455214241590383,
      "grad_norm": 0.890625,
      "learning_rate": 0.0004988943645518528,
      "loss": 5.6916,
      "mean_token_accuracy": 0.16624435186386108,
      "num_tokens": 8449817.0,
      "step": 3780
    },
    {
      "entropy": 5.696098661422729,
      "epoch": 0.4251137193238614,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988903857119611,
      "loss": 5.584,
      "mean_token_accuracy": 0.17602667212486267,
      "num_tokens": 8460096.0,
      "step": 3785
    },
    {
      "entropy": 5.727881622314453,
      "epoch": 0.42567529623181893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004988863997433125,
      "loss": 5.6931,
      "mean_token_accuracy": 0.16759534031152726,
      "num_tokens": 8470241.0,
      "step": 3790
    },
    {
      "entropy": 5.79317078590393,
      "epoch": 0.4262368731397765,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004988824066460341,
      "loss": 5.7181,
      "mean_token_accuracy": 0.16424463540315629,
      "num_tokens": 8481438.0,
      "step": 3795
    },
    {
      "entropy": 5.723605823516846,
      "epoch": 0.42679845004773403,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000498878406420253,
      "loss": 5.7641,
      "mean_token_accuracy": 0.16844544112682341,
      "num_tokens": 8492588.0,
      "step": 3800
    },
    {
      "entropy": 5.701861715316772,
      "epoch": 0.4273600269556916,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988743990660967,
      "loss": 5.7,
      "mean_token_accuracy": 0.1730359673500061,
      "num_tokens": 8502298.0,
      "step": 3805
    },
    {
      "entropy": 5.684565734863281,
      "epoch": 0.42792160386364914,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988703845836927,
      "loss": 5.5553,
      "mean_token_accuracy": 0.17092193514108658,
      "num_tokens": 8513507.0,
      "step": 3810
    },
    {
      "entropy": 5.68870587348938,
      "epoch": 0.4284831807716067,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004988663629731687,
      "loss": 5.6483,
      "mean_token_accuracy": 0.17624177038669586,
      "num_tokens": 8523888.0,
      "step": 3815
    },
    {
      "entropy": 5.653003692626953,
      "epoch": 0.42904475767956424,
      "grad_norm": 1.09375,
      "learning_rate": 0.000498862334234653,
      "loss": 5.5915,
      "mean_token_accuracy": 0.17567707151174544,
      "num_tokens": 8534032.0,
      "step": 3820
    },
    {
      "entropy": 5.649330806732178,
      "epoch": 0.4296063345875218,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004988582983682737,
      "loss": 5.7247,
      "mean_token_accuracy": 0.17112672924995423,
      "num_tokens": 8545544.0,
      "step": 3825
    },
    {
      "entropy": 5.793539524078369,
      "epoch": 0.4301679114954793,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004988542553741594,
      "loss": 5.7013,
      "mean_token_accuracy": 0.17370189726352692,
      "num_tokens": 8555428.0,
      "step": 3830
    },
    {
      "entropy": 5.6171880722045895,
      "epoch": 0.43072948840343683,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004988502052524388,
      "loss": 5.5979,
      "mean_token_accuracy": 0.1724550411105156,
      "num_tokens": 8566456.0,
      "step": 3835
    },
    {
      "entropy": 5.743177461624145,
      "epoch": 0.4312910653113944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004988461480032408,
      "loss": 5.5936,
      "mean_token_accuracy": 0.17514830976724624,
      "num_tokens": 8578125.0,
      "step": 3840
    },
    {
      "entropy": 5.716649436950684,
      "epoch": 0.43185264221935193,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004988420836266946,
      "loss": 5.6655,
      "mean_token_accuracy": 0.16592355221509933,
      "num_tokens": 8589530.0,
      "step": 3845
    },
    {
      "entropy": 5.725794553756714,
      "epoch": 0.4324142191273095,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988380121229298,
      "loss": 5.7672,
      "mean_token_accuracy": 0.17028849571943283,
      "num_tokens": 8601655.0,
      "step": 3850
    },
    {
      "entropy": 5.713348484039306,
      "epoch": 0.43297579603526704,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004988339334920757,
      "loss": 5.6767,
      "mean_token_accuracy": 0.1705267533659935,
      "num_tokens": 8613355.0,
      "step": 3855
    },
    {
      "entropy": 5.734662485122681,
      "epoch": 0.4335373729432246,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004988298477342625,
      "loss": 5.6305,
      "mean_token_accuracy": 0.17315073311328888,
      "num_tokens": 8624154.0,
      "step": 3860
    },
    {
      "entropy": 5.699270629882813,
      "epoch": 0.43409894985118214,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00049882575484962,
      "loss": 5.695,
      "mean_token_accuracy": 0.16900694072246553,
      "num_tokens": 8633888.0,
      "step": 3865
    },
    {
      "entropy": 5.7551063060760494,
      "epoch": 0.4346605267591397,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004988216548382787,
      "loss": 5.6884,
      "mean_token_accuracy": 0.1692011445760727,
      "num_tokens": 8644965.0,
      "step": 3870
    },
    {
      "entropy": 5.682348155975342,
      "epoch": 0.4352221036670972,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004988175477003691,
      "loss": 5.627,
      "mean_token_accuracy": 0.1764288291335106,
      "num_tokens": 8655483.0,
      "step": 3875
    },
    {
      "entropy": 5.7363025665283205,
      "epoch": 0.43578368057505473,
      "grad_norm": 0.984375,
      "learning_rate": 0.000498813433436022,
      "loss": 5.5988,
      "mean_token_accuracy": 0.17191628813743592,
      "num_tokens": 8666443.0,
      "step": 3880
    },
    {
      "entropy": 5.703504657745361,
      "epoch": 0.4363452574830123,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004988093120453684,
      "loss": 5.67,
      "mean_token_accuracy": 0.17085333168506622,
      "num_tokens": 8677198.0,
      "step": 3885
    },
    {
      "entropy": 5.684973621368409,
      "epoch": 0.43690683439096983,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004988051835285394,
      "loss": 5.7306,
      "mean_token_accuracy": 0.16749130338430404,
      "num_tokens": 8688303.0,
      "step": 3890
    },
    {
      "entropy": 5.781972980499267,
      "epoch": 0.4374684112989274,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004988010478856667,
      "loss": 5.6741,
      "mean_token_accuracy": 0.1697043225169182,
      "num_tokens": 8700326.0,
      "step": 3895
    },
    {
      "entropy": 5.625845098495484,
      "epoch": 0.43802998820688493,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004987969051168816,
      "loss": 5.6329,
      "mean_token_accuracy": 0.17153273224830629,
      "num_tokens": 8711101.0,
      "step": 3900
    },
    {
      "entropy": 5.743406009674072,
      "epoch": 0.4385915651148425,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004987927552223165,
      "loss": 5.7649,
      "mean_token_accuracy": 0.16504329591989517,
      "num_tokens": 8721751.0,
      "step": 3905
    },
    {
      "entropy": 5.804507160186768,
      "epoch": 0.43915314202280004,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004987885982021032,
      "loss": 5.8565,
      "mean_token_accuracy": 0.1615806683897972,
      "num_tokens": 8733550.0,
      "step": 3910
    },
    {
      "entropy": 5.763341474533081,
      "epoch": 0.4397147189307576,
      "grad_norm": 1.0,
      "learning_rate": 0.0004987844340563741,
      "loss": 5.651,
      "mean_token_accuracy": 0.17049803584814072,
      "num_tokens": 8744218.0,
      "step": 3915
    },
    {
      "entropy": 5.672728681564331,
      "epoch": 0.44027629583871514,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004987802627852619,
      "loss": 5.6748,
      "mean_token_accuracy": 0.17184038907289506,
      "num_tokens": 8756531.0,
      "step": 3920
    },
    {
      "entropy": 5.8487523078918455,
      "epoch": 0.44083787274667263,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987760843888992,
      "loss": 5.7415,
      "mean_token_accuracy": 0.16975418776273726,
      "num_tokens": 8768114.0,
      "step": 3925
    },
    {
      "entropy": 5.679783916473388,
      "epoch": 0.4413994496546302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004987718988674193,
      "loss": 5.7219,
      "mean_token_accuracy": 0.16620885878801345,
      "num_tokens": 8780010.0,
      "step": 3930
    },
    {
      "entropy": 5.7933605194091795,
      "epoch": 0.44196102656258773,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987677062209552,
      "loss": 5.6814,
      "mean_token_accuracy": 0.16475942283868789,
      "num_tokens": 8792183.0,
      "step": 3935
    },
    {
      "entropy": 5.781735372543335,
      "epoch": 0.4425226034705453,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004987635064496407,
      "loss": 5.7837,
      "mean_token_accuracy": 0.16257910430431366,
      "num_tokens": 8804513.0,
      "step": 3940
    },
    {
      "entropy": 5.6498900890350345,
      "epoch": 0.44308418037850283,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004987592995536093,
      "loss": 5.6765,
      "mean_token_accuracy": 0.1743319720029831,
      "num_tokens": 8814918.0,
      "step": 3945
    },
    {
      "entropy": 5.6842875480651855,
      "epoch": 0.4436457572864604,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004987550855329951,
      "loss": 5.6045,
      "mean_token_accuracy": 0.17486772388219834,
      "num_tokens": 8826176.0,
      "step": 3950
    },
    {
      "entropy": 5.625200366973877,
      "epoch": 0.44420733419441794,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004987508643879322,
      "loss": 5.6519,
      "mean_token_accuracy": 0.17125591188669204,
      "num_tokens": 8837242.0,
      "step": 3955
    },
    {
      "entropy": 5.761265802383423,
      "epoch": 0.4447689111023755,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004987466361185549,
      "loss": 5.6224,
      "mean_token_accuracy": 0.17221447229385375,
      "num_tokens": 8848972.0,
      "step": 3960
    },
    {
      "entropy": 5.615032911300659,
      "epoch": 0.44533048801033304,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0004987424007249979,
      "loss": 5.5954,
      "mean_token_accuracy": 0.1674984112381935,
      "num_tokens": 8861108.0,
      "step": 3965
    },
    {
      "entropy": 5.71315631866455,
      "epoch": 0.4458920649182906,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004987381582073962,
      "loss": 5.6948,
      "mean_token_accuracy": 0.1702861726284027,
      "num_tokens": 8872161.0,
      "step": 3970
    },
    {
      "entropy": 5.908317613601684,
      "epoch": 0.4464536418262481,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004987339085658848,
      "loss": 5.7831,
      "mean_token_accuracy": 0.167988520860672,
      "num_tokens": 8882497.0,
      "step": 3975
    },
    {
      "entropy": 5.535442876815796,
      "epoch": 0.44701521873420563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004987296518005988,
      "loss": 5.4955,
      "mean_token_accuracy": 0.17948410362005235,
      "num_tokens": 8893155.0,
      "step": 3980
    },
    {
      "entropy": 5.623105573654175,
      "epoch": 0.4475767956421632,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498725387911674,
      "loss": 5.588,
      "mean_token_accuracy": 0.1777444913983345,
      "num_tokens": 8904478.0,
      "step": 3985
    },
    {
      "entropy": 5.647171211242676,
      "epoch": 0.44813837255012073,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004987211168992461,
      "loss": 5.6971,
      "mean_token_accuracy": 0.16709236651659012,
      "num_tokens": 8915470.0,
      "step": 3990
    },
    {
      "entropy": 5.748912000656128,
      "epoch": 0.4486999494580783,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000498716838763451,
      "loss": 5.6248,
      "mean_token_accuracy": 0.17297569513320923,
      "num_tokens": 8925644.0,
      "step": 3995
    },
    {
      "entropy": 5.69286994934082,
      "epoch": 0.44926152636603583,
      "grad_norm": 1.0,
      "learning_rate": 0.000498712553504425,
      "loss": 5.6803,
      "mean_token_accuracy": 0.1693950116634369,
      "num_tokens": 8937000.0,
      "step": 4000
    },
    {
      "entropy": 5.702338647842407,
      "epoch": 0.4498231032739934,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004987082611223046,
      "loss": 5.652,
      "mean_token_accuracy": 0.16663608700037003,
      "num_tokens": 8948258.0,
      "step": 4005
    },
    {
      "entropy": 5.7069127559661865,
      "epoch": 0.45038468018195094,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004987039616172265,
      "loss": 5.6658,
      "mean_token_accuracy": 0.1694944828748703,
      "num_tokens": 8960107.0,
      "step": 4010
    },
    {
      "entropy": 5.695326137542724,
      "epoch": 0.4509462570899085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986996549893273,
      "loss": 5.6466,
      "mean_token_accuracy": 0.17219445556402208,
      "num_tokens": 8969612.0,
      "step": 4015
    },
    {
      "entropy": 5.7281805038452145,
      "epoch": 0.451507833997866,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004986953412387445,
      "loss": 5.7671,
      "mean_token_accuracy": 0.16545552909374237,
      "num_tokens": 8980970.0,
      "step": 4020
    },
    {
      "entropy": 5.767594480514527,
      "epoch": 0.45206941090582353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004986910203656151,
      "loss": 5.6849,
      "mean_token_accuracy": 0.17032318115234374,
      "num_tokens": 8992218.0,
      "step": 4025
    },
    {
      "entropy": 5.634401750564575,
      "epoch": 0.4526309878137811,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986866923700769,
      "loss": 5.6238,
      "mean_token_accuracy": 0.17393509894609452,
      "num_tokens": 9003737.0,
      "step": 4030
    },
    {
      "entropy": 5.749180650711059,
      "epoch": 0.45319256472173863,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004986823572522677,
      "loss": 5.6819,
      "mean_token_accuracy": 0.1711251273751259,
      "num_tokens": 9015427.0,
      "step": 4035
    },
    {
      "entropy": 5.680558538436889,
      "epoch": 0.4537541416296962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004986780150123254,
      "loss": 5.6964,
      "mean_token_accuracy": 0.16373877823352814,
      "num_tokens": 9025003.0,
      "step": 4040
    },
    {
      "entropy": 5.71775975227356,
      "epoch": 0.45431571853765373,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004986736656503885,
      "loss": 5.6494,
      "mean_token_accuracy": 0.17180948406457902,
      "num_tokens": 9036074.0,
      "step": 4045
    },
    {
      "entropy": 5.863864326477051,
      "epoch": 0.4548772954456113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004986693091665952,
      "loss": 5.6649,
      "mean_token_accuracy": 0.16726166009902954,
      "num_tokens": 9046878.0,
      "step": 4050
    },
    {
      "entropy": 5.714562797546387,
      "epoch": 0.45543887235356884,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986649455610845,
      "loss": 5.71,
      "mean_token_accuracy": 0.17038390934467315,
      "num_tokens": 9058366.0,
      "step": 4055
    },
    {
      "entropy": 5.740604114532471,
      "epoch": 0.4560004492615264,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000498660574833995,
      "loss": 5.7392,
      "mean_token_accuracy": 0.16881180554628372,
      "num_tokens": 9069542.0,
      "step": 4060
    },
    {
      "entropy": 5.799572229385376,
      "epoch": 0.45656202616948394,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004986561969854661,
      "loss": 5.8954,
      "mean_token_accuracy": 0.16551897376775743,
      "num_tokens": 9081219.0,
      "step": 4065
    },
    {
      "entropy": 5.775657701492309,
      "epoch": 0.45712360307744143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004986518120156373,
      "loss": 5.6023,
      "mean_token_accuracy": 0.17761090099811555,
      "num_tokens": 9092101.0,
      "step": 4070
    },
    {
      "entropy": 5.7041154384613035,
      "epoch": 0.457685179985399,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000498647419924648,
      "loss": 5.6463,
      "mean_token_accuracy": 0.16810258626937866,
      "num_tokens": 9101900.0,
      "step": 4075
    },
    {
      "entropy": 5.778073501586914,
      "epoch": 0.45824675689335653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498643020712638,
      "loss": 5.7838,
      "mean_token_accuracy": 0.16465375125408171,
      "num_tokens": 9114369.0,
      "step": 4080
    },
    {
      "entropy": 5.724344539642334,
      "epoch": 0.4588083338013141,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004986386143797475,
      "loss": 5.5867,
      "mean_token_accuracy": 0.1736169695854187,
      "num_tokens": 9126548.0,
      "step": 4085
    },
    {
      "entropy": 5.690601825714111,
      "epoch": 0.45936991070927163,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004986342009261169,
      "loss": 5.7894,
      "mean_token_accuracy": 0.17296182215213776,
      "num_tokens": 9138070.0,
      "step": 4090
    },
    {
      "entropy": 5.75274248123169,
      "epoch": 0.4599314876172292,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004986297803518865,
      "loss": 5.6849,
      "mean_token_accuracy": 0.16803380250930786,
      "num_tokens": 9149489.0,
      "step": 4095
    },
    {
      "entropy": 5.7161500453948975,
      "epoch": 0.46049306452518673,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004986253526571971,
      "loss": 5.6729,
      "mean_token_accuracy": 0.17992850244045258,
      "num_tokens": 9160363.0,
      "step": 4100
    },
    {
      "entropy": 5.758535861968994,
      "epoch": 0.4610546414331443,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986209178421899,
      "loss": 5.7325,
      "mean_token_accuracy": 0.16581229269504547,
      "num_tokens": 9171619.0,
      "step": 4105
    },
    {
      "entropy": 5.65956220626831,
      "epoch": 0.46161621834110184,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498616475907006,
      "loss": 5.5758,
      "mean_token_accuracy": 0.17373363375663758,
      "num_tokens": 9182828.0,
      "step": 4110
    },
    {
      "entropy": 5.785628318786621,
      "epoch": 0.46217779524905933,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004986120268517865,
      "loss": 5.7284,
      "mean_token_accuracy": 0.17375926226377486,
      "num_tokens": 9195249.0,
      "step": 4115
    },
    {
      "entropy": 5.650385236740112,
      "epoch": 0.4627393721570169,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004986075706766736,
      "loss": 5.5489,
      "mean_token_accuracy": 0.17288807034492493,
      "num_tokens": 9205770.0,
      "step": 4120
    },
    {
      "entropy": 5.618177700042724,
      "epoch": 0.46330094906497443,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004986031073818088,
      "loss": 5.565,
      "mean_token_accuracy": 0.17578472793102265,
      "num_tokens": 9216367.0,
      "step": 4125
    },
    {
      "entropy": 5.655687379837036,
      "epoch": 0.463862525972932,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004985986369673343,
      "loss": 5.6633,
      "mean_token_accuracy": 0.16507088094949723,
      "num_tokens": 9228524.0,
      "step": 4130
    },
    {
      "entropy": 5.6996095180511475,
      "epoch": 0.46442410288088953,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985941594333926,
      "loss": 5.6356,
      "mean_token_accuracy": 0.1752336338162422,
      "num_tokens": 9238907.0,
      "step": 4135
    },
    {
      "entropy": 5.714912223815918,
      "epoch": 0.4649856797888471,
      "grad_norm": 1.0,
      "learning_rate": 0.000498589674780126,
      "loss": 5.77,
      "mean_token_accuracy": 0.17423523664474488,
      "num_tokens": 9250491.0,
      "step": 4140
    },
    {
      "entropy": 5.73002142906189,
      "epoch": 0.46554725669680463,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985851830076775,
      "loss": 5.5621,
      "mean_token_accuracy": 0.1795693039894104,
      "num_tokens": 9260984.0,
      "step": 4145
    },
    {
      "entropy": 5.695757102966309,
      "epoch": 0.4661088336047622,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004985806841161901,
      "loss": 5.7172,
      "mean_token_accuracy": 0.17257772982120514,
      "num_tokens": 9272416.0,
      "step": 4150
    },
    {
      "entropy": 5.755163145065308,
      "epoch": 0.46667041051271974,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000498576178105807,
      "loss": 5.6989,
      "mean_token_accuracy": 0.17022405862808226,
      "num_tokens": 9284513.0,
      "step": 4155
    },
    {
      "entropy": 5.807848739624023,
      "epoch": 0.4672319874206773,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004985716649766717,
      "loss": 5.6046,
      "mean_token_accuracy": 0.18052294999361038,
      "num_tokens": 9293422.0,
      "step": 4160
    },
    {
      "entropy": 5.560371351242066,
      "epoch": 0.4677935643286348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004985671447289277,
      "loss": 5.5873,
      "mean_token_accuracy": 0.1724488064646721,
      "num_tokens": 9304735.0,
      "step": 4165
    },
    {
      "entropy": 5.6328857421875,
      "epoch": 0.46835514123659233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004985626173627194,
      "loss": 5.7043,
      "mean_token_accuracy": 0.16736284494400025,
      "num_tokens": 9315659.0,
      "step": 4170
    },
    {
      "entropy": 5.726731061935425,
      "epoch": 0.4689167181445499,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004985580828781904,
      "loss": 5.7151,
      "mean_token_accuracy": 0.1724039524793625,
      "num_tokens": 9327106.0,
      "step": 4175
    },
    {
      "entropy": 5.876207685470581,
      "epoch": 0.46947829505250743,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985535412754855,
      "loss": 5.7555,
      "mean_token_accuracy": 0.1699582353234291,
      "num_tokens": 9338762.0,
      "step": 4180
    },
    {
      "entropy": 5.685064792633057,
      "epoch": 0.470039871960465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000498548992554749,
      "loss": 5.6221,
      "mean_token_accuracy": 0.17380139529705046,
      "num_tokens": 9349495.0,
      "step": 4185
    },
    {
      "entropy": 5.800124883651733,
      "epoch": 0.47060144886842253,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004985444367161259,
      "loss": 5.7205,
      "mean_token_accuracy": 0.17481711953878404,
      "num_tokens": 9360830.0,
      "step": 4190
    },
    {
      "entropy": 5.617068433761597,
      "epoch": 0.4711630257763801,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985398737597614,
      "loss": 5.6033,
      "mean_token_accuracy": 0.173471137881279,
      "num_tokens": 9371668.0,
      "step": 4195
    },
    {
      "entropy": 5.7108642101287845,
      "epoch": 0.47172460268433763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004985353036858004,
      "loss": 5.728,
      "mean_token_accuracy": 0.1676153853535652,
      "num_tokens": 9383437.0,
      "step": 4200
    },
    {
      "entropy": 5.6791785717010494,
      "epoch": 0.4722861795922952,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985307264943887,
      "loss": 5.5581,
      "mean_token_accuracy": 0.17802588790655136,
      "num_tokens": 9394597.0,
      "step": 4205
    },
    {
      "entropy": 5.551296997070312,
      "epoch": 0.47284775650025274,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004985261421856721,
      "loss": 5.5116,
      "mean_token_accuracy": 0.17718030661344528,
      "num_tokens": 9404322.0,
      "step": 4210
    },
    {
      "entropy": 5.776150417327881,
      "epoch": 0.47340933340821023,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004985215507597963,
      "loss": 5.8525,
      "mean_token_accuracy": 0.1577995389699936,
      "num_tokens": 9416740.0,
      "step": 4215
    },
    {
      "entropy": 5.773025894165039,
      "epoch": 0.4739709103161678,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004985169522169076,
      "loss": 5.6717,
      "mean_token_accuracy": 0.17395939379930497,
      "num_tokens": 9428617.0,
      "step": 4220
    },
    {
      "entropy": 5.618877124786377,
      "epoch": 0.47453248722412533,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004985123465571526,
      "loss": 5.6091,
      "mean_token_accuracy": 0.17624282240867614,
      "num_tokens": 9439686.0,
      "step": 4225
    },
    {
      "entropy": 5.7461926460266115,
      "epoch": 0.4750940641320829,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004985077337806777,
      "loss": 5.606,
      "mean_token_accuracy": 0.17703299671411515,
      "num_tokens": 9450378.0,
      "step": 4230
    },
    {
      "entropy": 5.698893260955811,
      "epoch": 0.47565564104004043,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004985031138876299,
      "loss": 5.6278,
      "mean_token_accuracy": 0.1687405675649643,
      "num_tokens": 9461659.0,
      "step": 4235
    },
    {
      "entropy": 5.665384578704834,
      "epoch": 0.476217217947998,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004984984868781563,
      "loss": 5.6683,
      "mean_token_accuracy": 0.17358199805021285,
      "num_tokens": 9472528.0,
      "step": 4240
    },
    {
      "entropy": 5.721430969238281,
      "epoch": 0.47677879485595553,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984938527524041,
      "loss": 5.595,
      "mean_token_accuracy": 0.18256423324346543,
      "num_tokens": 9483388.0,
      "step": 4245
    },
    {
      "entropy": 5.504148483276367,
      "epoch": 0.4773403717639131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000498489211510521,
      "loss": 5.4575,
      "mean_token_accuracy": 0.18474066406488418,
      "num_tokens": 9495178.0,
      "step": 4250
    },
    {
      "entropy": 5.655213975906372,
      "epoch": 0.47790194867187064,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004984845631526547,
      "loss": 5.704,
      "mean_token_accuracy": 0.16891445368528366,
      "num_tokens": 9506226.0,
      "step": 4255
    },
    {
      "entropy": 5.75639328956604,
      "epoch": 0.47846352557982813,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0004984799076789534,
      "loss": 5.5916,
      "mean_token_accuracy": 0.1722211003303528,
      "num_tokens": 9517671.0,
      "step": 4260
    },
    {
      "entropy": 5.593520927429199,
      "epoch": 0.4790251024877857,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984752450895649,
      "loss": 5.6747,
      "mean_token_accuracy": 0.17680443227291107,
      "num_tokens": 9527637.0,
      "step": 4265
    },
    {
      "entropy": 5.814142894744873,
      "epoch": 0.47958667939574323,
      "grad_norm": 1.015625,
      "learning_rate": 0.000498470575384638,
      "loss": 5.7664,
      "mean_token_accuracy": 0.1659485802054405,
      "num_tokens": 9539686.0,
      "step": 4270
    },
    {
      "entropy": 5.7889586925506595,
      "epoch": 0.4801482563037008,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004984658985643214,
      "loss": 5.7242,
      "mean_token_accuracy": 0.17376924455165862,
      "num_tokens": 9550775.0,
      "step": 4275
    },
    {
      "entropy": 5.684207963943481,
      "epoch": 0.48070983321165833,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004984612146287638,
      "loss": 5.629,
      "mean_token_accuracy": 0.17294028997421265,
      "num_tokens": 9561760.0,
      "step": 4280
    },
    {
      "entropy": 5.680255651473999,
      "epoch": 0.4812714101196159,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004984565235781147,
      "loss": 5.6651,
      "mean_token_accuracy": 0.16704404801130296,
      "num_tokens": 9573899.0,
      "step": 4285
    },
    {
      "entropy": 5.705337476730347,
      "epoch": 0.48183298702757343,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498451825412523,
      "loss": 5.6078,
      "mean_token_accuracy": 0.1714568927884102,
      "num_tokens": 9584336.0,
      "step": 4290
    },
    {
      "entropy": 5.807095241546631,
      "epoch": 0.482394563935531,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004984471201321386,
      "loss": 5.8142,
      "mean_token_accuracy": 0.16588468104600906,
      "num_tokens": 9596614.0,
      "step": 4295
    },
    {
      "entropy": 5.73848524093628,
      "epoch": 0.48295614084348854,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004984424077371112,
      "loss": 5.6509,
      "mean_token_accuracy": 0.17467584908008577,
      "num_tokens": 9606796.0,
      "step": 4300
    },
    {
      "entropy": 5.669296789169311,
      "epoch": 0.4835177177514461,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984376882275909,
      "loss": 5.6237,
      "mean_token_accuracy": 0.17255420088768006,
      "num_tokens": 9617705.0,
      "step": 4305
    },
    {
      "entropy": 5.63599853515625,
      "epoch": 0.4840792946594036,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004984329616037279,
      "loss": 5.6547,
      "mean_token_accuracy": 0.1768753245472908,
      "num_tokens": 9628098.0,
      "step": 4310
    },
    {
      "entropy": 5.729220342636109,
      "epoch": 0.48464087156736113,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004984282278656727,
      "loss": 5.6179,
      "mean_token_accuracy": 0.17815225571393967,
      "num_tokens": 9637697.0,
      "step": 4315
    },
    {
      "entropy": 5.583554744720459,
      "epoch": 0.4852024484753187,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004984234870135762,
      "loss": 5.495,
      "mean_token_accuracy": 0.18339364230632782,
      "num_tokens": 9648221.0,
      "step": 4320
    },
    {
      "entropy": 5.797052574157715,
      "epoch": 0.48576402538327623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004984187390475891,
      "loss": 5.6917,
      "mean_token_accuracy": 0.166446553170681,
      "num_tokens": 9660182.0,
      "step": 4325
    },
    {
      "entropy": 5.671581077575683,
      "epoch": 0.4863256022912338,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004984139839678628,
      "loss": 5.6953,
      "mean_token_accuracy": 0.17090761065483093,
      "num_tokens": 9672121.0,
      "step": 4330
    },
    {
      "entropy": 5.74064621925354,
      "epoch": 0.48688717919919133,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004984092217745486,
      "loss": 5.5858,
      "mean_token_accuracy": 0.17702260613441467,
      "num_tokens": 9682651.0,
      "step": 4335
    },
    {
      "entropy": 5.702525949478149,
      "epoch": 0.4874487561071489,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004984044524677981,
      "loss": 5.751,
      "mean_token_accuracy": 0.16179810613393783,
      "num_tokens": 9693989.0,
      "step": 4340
    },
    {
      "entropy": 5.620242309570313,
      "epoch": 0.48801033301510643,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004983996760477632,
      "loss": 5.596,
      "mean_token_accuracy": 0.17746072858572007,
      "num_tokens": 9703802.0,
      "step": 4345
    },
    {
      "entropy": 5.643984460830689,
      "epoch": 0.488571909923064,
      "grad_norm": 1.109375,
      "learning_rate": 0.000498394892514596,
      "loss": 5.543,
      "mean_token_accuracy": 0.1815675228834152,
      "num_tokens": 9714287.0,
      "step": 4350
    },
    {
      "entropy": 5.644256639480591,
      "epoch": 0.48913348683102154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983901018684486,
      "loss": 5.5823,
      "mean_token_accuracy": 0.17668364942073822,
      "num_tokens": 9724819.0,
      "step": 4355
    },
    {
      "entropy": 5.718612432479858,
      "epoch": 0.48969506373897903,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004983853041094739,
      "loss": 5.7221,
      "mean_token_accuracy": 0.165558297932148,
      "num_tokens": 9736326.0,
      "step": 4360
    },
    {
      "entropy": 5.774798202514648,
      "epoch": 0.4902566406469366,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004983804992378243,
      "loss": 5.7507,
      "mean_token_accuracy": 0.17167295068502425,
      "num_tokens": 9747883.0,
      "step": 4365
    },
    {
      "entropy": 5.689925479888916,
      "epoch": 0.49081821755489413,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983756872536532,
      "loss": 5.601,
      "mean_token_accuracy": 0.17454955279827117,
      "num_tokens": 9758191.0,
      "step": 4370
    },
    {
      "entropy": 5.730838441848755,
      "epoch": 0.4913797944628517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983708681571134,
      "loss": 5.7159,
      "mean_token_accuracy": 0.1638861894607544,
      "num_tokens": 9768852.0,
      "step": 4375
    },
    {
      "entropy": 5.745866346359253,
      "epoch": 0.49194137137080923,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983660419483586,
      "loss": 5.6589,
      "mean_token_accuracy": 0.1688293218612671,
      "num_tokens": 9779033.0,
      "step": 4380
    },
    {
      "entropy": 5.605787324905395,
      "epoch": 0.4925029482787668,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004983612086275424,
      "loss": 5.5672,
      "mean_token_accuracy": 0.1765485405921936,
      "num_tokens": 9790777.0,
      "step": 4385
    },
    {
      "entropy": 5.888398504257202,
      "epoch": 0.49306452518672433,
      "grad_norm": 1.0,
      "learning_rate": 0.0004983563681948187,
      "loss": 5.843,
      "mean_token_accuracy": 0.16131045520305634,
      "num_tokens": 9802305.0,
      "step": 4390
    },
    {
      "entropy": 5.710119247436523,
      "epoch": 0.4936261020946819,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004983515206503416,
      "loss": 5.6445,
      "mean_token_accuracy": 0.16851651966571807,
      "num_tokens": 9813590.0,
      "step": 4395
    },
    {
      "entropy": 5.668114519119262,
      "epoch": 0.49418767900263944,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983466659942655,
      "loss": 5.722,
      "mean_token_accuracy": 0.16949656307697297,
      "num_tokens": 9824080.0,
      "step": 4400
    },
    {
      "entropy": 5.713350105285644,
      "epoch": 0.49474925591059693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004983418042267449,
      "loss": 5.5935,
      "mean_token_accuracy": 0.16558514684438705,
      "num_tokens": 9835441.0,
      "step": 4405
    },
    {
      "entropy": 5.660351276397705,
      "epoch": 0.4953108328185545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004983369353479345,
      "loss": 5.5933,
      "mean_token_accuracy": 0.17601150274276733,
      "num_tokens": 9846750.0,
      "step": 4410
    },
    {
      "entropy": 5.720973062515259,
      "epoch": 0.49587240972651203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004983320593579895,
      "loss": 5.7379,
      "mean_token_accuracy": 0.16703337281942368,
      "num_tokens": 9857987.0,
      "step": 4415
    },
    {
      "entropy": 5.721000099182129,
      "epoch": 0.4964339866344696,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004983271762570653,
      "loss": 5.6852,
      "mean_token_accuracy": 0.17118700444698334,
      "num_tokens": 9869736.0,
      "step": 4420
    },
    {
      "entropy": 5.729393053054809,
      "epoch": 0.49699556354242713,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000498322286045317,
      "loss": 5.6645,
      "mean_token_accuracy": 0.16667413115501403,
      "num_tokens": 9881610.0,
      "step": 4425
    },
    {
      "entropy": 5.718217992782593,
      "epoch": 0.4975571404503847,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004983173887229006,
      "loss": 5.6515,
      "mean_token_accuracy": 0.1720265492796898,
      "num_tokens": 9893203.0,
      "step": 4430
    },
    {
      "entropy": 5.659297847747803,
      "epoch": 0.49811871735834223,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004983124842899719,
      "loss": 5.5568,
      "mean_token_accuracy": 0.17038003504276275,
      "num_tokens": 9904392.0,
      "step": 4435
    },
    {
      "entropy": 5.501509332656861,
      "epoch": 0.4986802942662998,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004983075727466871,
      "loss": 5.5744,
      "mean_token_accuracy": 0.17519020140171052,
      "num_tokens": 9916184.0,
      "step": 4440
    },
    {
      "entropy": 5.780448913574219,
      "epoch": 0.49924187117425733,
      "grad_norm": 0.87890625,
      "learning_rate": 0.0004983026540932026,
      "loss": 5.6722,
      "mean_token_accuracy": 0.17609916627407074,
      "num_tokens": 9927516.0,
      "step": 4445
    },
    {
      "entropy": 5.6783655166625975,
      "epoch": 0.4998034480822149,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982977283296749,
      "loss": 5.694,
      "mean_token_accuracy": 0.1671729415655136,
      "num_tokens": 9939182.0,
      "step": 4450
    },
    {
      "entropy": 5.767272520065307,
      "epoch": 0.5003650249901724,
      "grad_norm": 1.046875,
      "learning_rate": 0.000498292795456261,
      "loss": 5.7116,
      "mean_token_accuracy": 0.16822021901607515,
      "num_tokens": 9950382.0,
      "step": 4455
    },
    {
      "entropy": 5.745260000228882,
      "epoch": 0.50092660189813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004982878554731178,
      "loss": 5.7076,
      "mean_token_accuracy": 0.17256211042404174,
      "num_tokens": 9961053.0,
      "step": 4460
    },
    {
      "entropy": 5.66381425857544,
      "epoch": 0.5014881788060875,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004982829083804028,
      "loss": 5.623,
      "mean_token_accuracy": 0.17677563279867173,
      "num_tokens": 9971313.0,
      "step": 4465
    },
    {
      "entropy": 5.666288185119629,
      "epoch": 0.5020497557140451,
      "grad_norm": 1.0,
      "learning_rate": 0.0004982779541782733,
      "loss": 5.6324,
      "mean_token_accuracy": 0.18253414034843446,
      "num_tokens": 9981257.0,
      "step": 4470
    },
    {
      "entropy": 5.647140789031982,
      "epoch": 0.5026113326220026,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004982729928668872,
      "loss": 5.5542,
      "mean_token_accuracy": 0.1735476166009903,
      "num_tokens": 9991860.0,
      "step": 4475
    },
    {
      "entropy": 5.646668100357056,
      "epoch": 0.5031729095299601,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982680244464024,
      "loss": 5.6658,
      "mean_token_accuracy": 0.17404285818338394,
      "num_tokens": 10002320.0,
      "step": 4480
    },
    {
      "entropy": 5.753510284423828,
      "epoch": 0.5037344864379176,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982630489169771,
      "loss": 5.7404,
      "mean_token_accuracy": 0.1676941141486168,
      "num_tokens": 10013187.0,
      "step": 4485
    },
    {
      "entropy": 5.6779985427856445,
      "epoch": 0.5042960633458752,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004982580662787697,
      "loss": 5.4805,
      "mean_token_accuracy": 0.18031218498945237,
      "num_tokens": 10024268.0,
      "step": 4490
    },
    {
      "entropy": 5.717615222930908,
      "epoch": 0.5048576402538327,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004982530765319389,
      "loss": 5.7388,
      "mean_token_accuracy": 0.16889102533459663,
      "num_tokens": 10034738.0,
      "step": 4495
    },
    {
      "entropy": 5.7839428901672365,
      "epoch": 0.5054192171617903,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004982480796766436,
      "loss": 5.691,
      "mean_token_accuracy": 0.16897645890712737,
      "num_tokens": 10045666.0,
      "step": 4500
    },
    {
      "entropy": 5.626480913162231,
      "epoch": 0.5059807940697478,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004982430757130429,
      "loss": 5.5775,
      "mean_token_accuracy": 0.1755047470331192,
      "num_tokens": 10057084.0,
      "step": 4505
    },
    {
      "entropy": 5.602354907989502,
      "epoch": 0.5065423709777054,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004982380646412959,
      "loss": 5.5455,
      "mean_token_accuracy": 0.17009968608617781,
      "num_tokens": 10066754.0,
      "step": 4510
    },
    {
      "entropy": 5.617758274078369,
      "epoch": 0.5071039478856629,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004982330464615625,
      "loss": 5.5546,
      "mean_token_accuracy": 0.17863859385251998,
      "num_tokens": 10078139.0,
      "step": 4515
    },
    {
      "entropy": 5.678718709945679,
      "epoch": 0.5076655247936205,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004982280211740023,
      "loss": 5.6238,
      "mean_token_accuracy": 0.1695124551653862,
      "num_tokens": 10089652.0,
      "step": 4520
    },
    {
      "entropy": 5.48136339187622,
      "epoch": 0.508227101701578,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004982229887787753,
      "loss": 5.463,
      "mean_token_accuracy": 0.17854669988155364,
      "num_tokens": 10100642.0,
      "step": 4525
    },
    {
      "entropy": 5.7619740009307865,
      "epoch": 0.5087886786095356,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004982179492760418,
      "loss": 5.6819,
      "mean_token_accuracy": 0.1769241452217102,
      "num_tokens": 10112036.0,
      "step": 4530
    },
    {
      "entropy": 5.61248927116394,
      "epoch": 0.5093502555174931,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004982129026659621,
      "loss": 5.5483,
      "mean_token_accuracy": 0.16963760256767274,
      "num_tokens": 10121653.0,
      "step": 4535
    },
    {
      "entropy": 5.642135858535767,
      "epoch": 0.5099118324254507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004982078489486972,
      "loss": 5.6458,
      "mean_token_accuracy": 0.1726953536272049,
      "num_tokens": 10132271.0,
      "step": 4540
    },
    {
      "entropy": 5.611723756790161,
      "epoch": 0.5104734093334082,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004982027881244076,
      "loss": 5.5524,
      "mean_token_accuracy": 0.17125996798276902,
      "num_tokens": 10143411.0,
      "step": 4545
    },
    {
      "entropy": 5.780136585235596,
      "epoch": 0.5110349862413658,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981977201932549,
      "loss": 5.6472,
      "mean_token_accuracy": 0.17798845767974852,
      "num_tokens": 10154338.0,
      "step": 4550
    },
    {
      "entropy": 5.703117322921753,
      "epoch": 0.5115965631493233,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004981926451554001,
      "loss": 5.6857,
      "mean_token_accuracy": 0.17487033307552338,
      "num_tokens": 10164810.0,
      "step": 4555
    },
    {
      "entropy": 5.698769855499267,
      "epoch": 0.5121581400572809,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004981875630110048,
      "loss": 5.586,
      "mean_token_accuracy": 0.17893914729356766,
      "num_tokens": 10176323.0,
      "step": 4560
    },
    {
      "entropy": 5.648852825164795,
      "epoch": 0.5127197169652384,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981824737602311,
      "loss": 5.6298,
      "mean_token_accuracy": 0.17639541029930114,
      "num_tokens": 10187110.0,
      "step": 4565
    },
    {
      "entropy": 5.599327182769775,
      "epoch": 0.513281293873196,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981773774032409,
      "loss": 5.5906,
      "mean_token_accuracy": 0.17272181212902069,
      "num_tokens": 10199099.0,
      "step": 4570
    },
    {
      "entropy": 5.647946834564209,
      "epoch": 0.5138428707811534,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981722739401962,
      "loss": 5.5767,
      "mean_token_accuracy": 0.17903112918138503,
      "num_tokens": 10209726.0,
      "step": 4575
    },
    {
      "entropy": 5.722096872329712,
      "epoch": 0.514404447689111,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049816716337126,
      "loss": 5.6819,
      "mean_token_accuracy": 0.17278391122817993,
      "num_tokens": 10220858.0,
      "step": 4580
    },
    {
      "entropy": 5.582174587249756,
      "epoch": 0.5149660245970685,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004981620456965946,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18093932121992112,
      "num_tokens": 10232250.0,
      "step": 4585
    },
    {
      "entropy": 5.81311411857605,
      "epoch": 0.5155276015050261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004981569209163633,
      "loss": 5.7022,
      "mean_token_accuracy": 0.16420883685350418,
      "num_tokens": 10244584.0,
      "step": 4590
    },
    {
      "entropy": 5.683686208724976,
      "epoch": 0.5160891784129836,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981517890307288,
      "loss": 5.5779,
      "mean_token_accuracy": 0.1777512863278389,
      "num_tokens": 10255926.0,
      "step": 4595
    },
    {
      "entropy": 5.6573210716247555,
      "epoch": 0.5166507553209412,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000498146650039855,
      "loss": 5.65,
      "mean_token_accuracy": 0.17132266908884047,
      "num_tokens": 10266691.0,
      "step": 4600
    },
    {
      "entropy": 5.7006834030151365,
      "epoch": 0.5172123322288987,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004981415039439054,
      "loss": 5.714,
      "mean_token_accuracy": 0.16987471729516984,
      "num_tokens": 10277744.0,
      "step": 4605
    },
    {
      "entropy": 5.6352331161499025,
      "epoch": 0.5177739091368563,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004981363507430436,
      "loss": 5.5693,
      "mean_token_accuracy": 0.17686766833066941,
      "num_tokens": 10288792.0,
      "step": 4610
    },
    {
      "entropy": 5.709258222579956,
      "epoch": 0.5183354860448138,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004981311904374339,
      "loss": 5.6174,
      "mean_token_accuracy": 0.17828530371189116,
      "num_tokens": 10299569.0,
      "step": 4615
    },
    {
      "entropy": 5.6432445526123045,
      "epoch": 0.5188970629527714,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004981260230272405,
      "loss": 5.6014,
      "mean_token_accuracy": 0.1772746756672859,
      "num_tokens": 10309742.0,
      "step": 4620
    },
    {
      "entropy": 5.705870723724365,
      "epoch": 0.5194586398607289,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004981208485126281,
      "loss": 5.6349,
      "mean_token_accuracy": 0.16997847259044646,
      "num_tokens": 10320603.0,
      "step": 4625
    },
    {
      "entropy": 5.649820137023926,
      "epoch": 0.5200202167686865,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004981156668937613,
      "loss": 5.5889,
      "mean_token_accuracy": 0.17648871541023253,
      "num_tokens": 10331498.0,
      "step": 4630
    },
    {
      "entropy": 5.689421367645264,
      "epoch": 0.520581793676644,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004981104781708051,
      "loss": 5.6418,
      "mean_token_accuracy": 0.17633580565452575,
      "num_tokens": 10342002.0,
      "step": 4635
    },
    {
      "entropy": 5.650698804855347,
      "epoch": 0.5211433705846016,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004981052823439247,
      "loss": 5.5762,
      "mean_token_accuracy": 0.1742894858121872,
      "num_tokens": 10351797.0,
      "step": 4640
    },
    {
      "entropy": 5.602195072174072,
      "epoch": 0.5217049474925591,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004981000794132858,
      "loss": 5.5256,
      "mean_token_accuracy": 0.17473063617944717,
      "num_tokens": 10362538.0,
      "step": 4645
    },
    {
      "entropy": 5.668619585037232,
      "epoch": 0.5222665244005167,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980948693790536,
      "loss": 5.5433,
      "mean_token_accuracy": 0.1789860725402832,
      "num_tokens": 10373365.0,
      "step": 4650
    },
    {
      "entropy": 5.540914106369018,
      "epoch": 0.5228281013084742,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980896522413943,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17348830550909042,
      "num_tokens": 10385087.0,
      "step": 4655
    },
    {
      "entropy": 5.6270355701446535,
      "epoch": 0.5233896782164318,
      "grad_norm": 1.0,
      "learning_rate": 0.0004980844280004739,
      "loss": 5.6637,
      "mean_token_accuracy": 0.1717521905899048,
      "num_tokens": 10396014.0,
      "step": 4660
    },
    {
      "entropy": 5.677709627151489,
      "epoch": 0.5239512551243893,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004980791966564589,
      "loss": 5.5886,
      "mean_token_accuracy": 0.17231246829032898,
      "num_tokens": 10407844.0,
      "step": 4665
    },
    {
      "entropy": 5.631974363327027,
      "epoch": 0.5245128320323468,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004980739582095158,
      "loss": 5.6409,
      "mean_token_accuracy": 0.17033209204673766,
      "num_tokens": 10419892.0,
      "step": 4670
    },
    {
      "entropy": 5.694100761413575,
      "epoch": 0.5250744089403043,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004980687126598111,
      "loss": 5.6182,
      "mean_token_accuracy": 0.16709839850664138,
      "num_tokens": 10430686.0,
      "step": 4675
    },
    {
      "entropy": 5.6673839569091795,
      "epoch": 0.5256359858482619,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004980634600075123,
      "loss": 5.6297,
      "mean_token_accuracy": 0.16909531205892564,
      "num_tokens": 10443785.0,
      "step": 4680
    },
    {
      "entropy": 5.568236303329468,
      "epoch": 0.5261975627562194,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004980582002527863,
      "loss": 5.5188,
      "mean_token_accuracy": 0.18584645986557008,
      "num_tokens": 10454889.0,
      "step": 4685
    },
    {
      "entropy": 5.65562105178833,
      "epoch": 0.526759139664177,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004980529333958008,
      "loss": 5.6833,
      "mean_token_accuracy": 0.17190586179494857,
      "num_tokens": 10466525.0,
      "step": 4690
    },
    {
      "entropy": 5.736246919631958,
      "epoch": 0.5273207165721345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004980476594367233,
      "loss": 5.6479,
      "mean_token_accuracy": 0.17422852665185928,
      "num_tokens": 10476329.0,
      "step": 4695
    },
    {
      "entropy": 5.713373851776123,
      "epoch": 0.5278822934800921,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004980423783757219,
      "loss": 5.6941,
      "mean_token_accuracy": 0.1705111175775528,
      "num_tokens": 10487178.0,
      "step": 4700
    },
    {
      "entropy": 5.673933410644532,
      "epoch": 0.5284438703880496,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004980370902129646,
      "loss": 5.639,
      "mean_token_accuracy": 0.16202449351549147,
      "num_tokens": 10499293.0,
      "step": 4705
    },
    {
      "entropy": 5.801592874526977,
      "epoch": 0.5290054472960072,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049803179494862,
      "loss": 5.8375,
      "mean_token_accuracy": 0.16088610589504243,
      "num_tokens": 10511762.0,
      "step": 4710
    },
    {
      "entropy": 5.705195379257202,
      "epoch": 0.5295670242039647,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004980264925828563,
      "loss": 5.5433,
      "mean_token_accuracy": 0.17806048691272736,
      "num_tokens": 10521725.0,
      "step": 4715
    },
    {
      "entropy": 5.575414800643921,
      "epoch": 0.5301286011119223,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004980211831158427,
      "loss": 5.5123,
      "mean_token_accuracy": 0.17416888177394868,
      "num_tokens": 10531764.0,
      "step": 4720
    },
    {
      "entropy": 5.686986017227173,
      "epoch": 0.5306901780198798,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004980158665477481,
      "loss": 5.5968,
      "mean_token_accuracy": 0.1773478016257286,
      "num_tokens": 10542127.0,
      "step": 4725
    },
    {
      "entropy": 5.767588186264038,
      "epoch": 0.5312517549278374,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004980105428787419,
      "loss": 5.6746,
      "mean_token_accuracy": 0.168606673181057,
      "num_tokens": 10552580.0,
      "step": 4730
    },
    {
      "entropy": 5.71980037689209,
      "epoch": 0.5318133318357949,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004980052121089934,
      "loss": 5.6111,
      "mean_token_accuracy": 0.17273032516241074,
      "num_tokens": 10562961.0,
      "step": 4735
    },
    {
      "entropy": 5.67135534286499,
      "epoch": 0.5323749087437525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004979998742386725,
      "loss": 5.6059,
      "mean_token_accuracy": 0.17421879917383193,
      "num_tokens": 10573210.0,
      "step": 4740
    },
    {
      "entropy": 5.5456846237182615,
      "epoch": 0.53293648565171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979945292679492,
      "loss": 5.5834,
      "mean_token_accuracy": 0.17576988488435746,
      "num_tokens": 10583878.0,
      "step": 4745
    },
    {
      "entropy": 5.75994610786438,
      "epoch": 0.5334980625596676,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979891771969935,
      "loss": 5.6496,
      "mean_token_accuracy": 0.17160634100437164,
      "num_tokens": 10593420.0,
      "step": 4750
    },
    {
      "entropy": 5.647671222686768,
      "epoch": 0.5340596394676251,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497983818025976,
      "loss": 5.6557,
      "mean_token_accuracy": 0.17253002822399138,
      "num_tokens": 10604213.0,
      "step": 4755
    },
    {
      "entropy": 5.70485782623291,
      "epoch": 0.5346212163755827,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979784517550671,
      "loss": 5.6377,
      "mean_token_accuracy": 0.16934602707624435,
      "num_tokens": 10615502.0,
      "step": 4760
    },
    {
      "entropy": 5.713788461685181,
      "epoch": 0.5351827932835401,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979730783844379,
      "loss": 5.6848,
      "mean_token_accuracy": 0.16481025665998458,
      "num_tokens": 10627589.0,
      "step": 4765
    },
    {
      "entropy": 5.6035338878631595,
      "epoch": 0.5357443701914977,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979676979142595,
      "loss": 5.5418,
      "mean_token_accuracy": 0.17140628397464752,
      "num_tokens": 10638365.0,
      "step": 4770
    },
    {
      "entropy": 5.628456687927246,
      "epoch": 0.5363059470994552,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497962310344703,
      "loss": 5.6213,
      "mean_token_accuracy": 0.17850914150476455,
      "num_tokens": 10650664.0,
      "step": 4775
    },
    {
      "entropy": 5.745699739456176,
      "epoch": 0.5368675240074128,
      "grad_norm": 0.984375,
      "learning_rate": 0.00049795691567594,
      "loss": 5.6795,
      "mean_token_accuracy": 0.1768557459115982,
      "num_tokens": 10662418.0,
      "step": 4780
    },
    {
      "entropy": 5.677951622009277,
      "epoch": 0.5374291009153703,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004979515139081426,
      "loss": 5.5651,
      "mean_token_accuracy": 0.17295360714197158,
      "num_tokens": 10672793.0,
      "step": 4785
    },
    {
      "entropy": 5.593189001083374,
      "epoch": 0.5379906778233279,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004979461050414822,
      "loss": 5.6038,
      "mean_token_accuracy": 0.180092054605484,
      "num_tokens": 10684082.0,
      "step": 4790
    },
    {
      "entropy": 5.5773176670074465,
      "epoch": 0.5385522547312854,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004979406890761315,
      "loss": 5.5397,
      "mean_token_accuracy": 0.17775683850049973,
      "num_tokens": 10694490.0,
      "step": 4795
    },
    {
      "entropy": 5.614388370513916,
      "epoch": 0.539113831639243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004979352660122629,
      "loss": 5.5572,
      "mean_token_accuracy": 0.17515850961208343,
      "num_tokens": 10704731.0,
      "step": 4800
    },
    {
      "entropy": 5.610679578781128,
      "epoch": 0.5396754085472005,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004979298358500488,
      "loss": 5.5621,
      "mean_token_accuracy": 0.17771035432815552,
      "num_tokens": 10715735.0,
      "step": 4805
    },
    {
      "entropy": 5.630765390396118,
      "epoch": 0.5402369854551581,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004979243985896624,
      "loss": 5.546,
      "mean_token_accuracy": 0.17329697459936141,
      "num_tokens": 10726951.0,
      "step": 4810
    },
    {
      "entropy": 5.581228590011596,
      "epoch": 0.5407985623631156,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004979189542312765,
      "loss": 5.6131,
      "mean_token_accuracy": 0.1692335769534111,
      "num_tokens": 10737609.0,
      "step": 4815
    },
    {
      "entropy": 5.64575686454773,
      "epoch": 0.5413601392710732,
      "grad_norm": 1.0,
      "learning_rate": 0.0004979135027750649,
      "loss": 5.5058,
      "mean_token_accuracy": 0.18206682056188583,
      "num_tokens": 10748957.0,
      "step": 4820
    },
    {
      "entropy": 5.610311841964721,
      "epoch": 0.5419217161790307,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004979080442212007,
      "loss": 5.6864,
      "mean_token_accuracy": 0.1654767796397209,
      "num_tokens": 10759398.0,
      "step": 4825
    },
    {
      "entropy": 5.691197299957276,
      "epoch": 0.5424832930869883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004979025785698581,
      "loss": 5.6504,
      "mean_token_accuracy": 0.17272533029317855,
      "num_tokens": 10770436.0,
      "step": 4830
    },
    {
      "entropy": 5.610402250289917,
      "epoch": 0.5430448699949458,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004978971058212108,
      "loss": 5.5027,
      "mean_token_accuracy": 0.1822532132267952,
      "num_tokens": 10781160.0,
      "step": 4835
    },
    {
      "entropy": 5.696642541885376,
      "epoch": 0.5436064469029034,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004978916259754331,
      "loss": 5.6104,
      "mean_token_accuracy": 0.18096253424882888,
      "num_tokens": 10793170.0,
      "step": 4840
    },
    {
      "entropy": 5.580870532989502,
      "epoch": 0.5441680238108609,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004978861390326998,
      "loss": 5.5641,
      "mean_token_accuracy": 0.17518757581710814,
      "num_tokens": 10803515.0,
      "step": 4845
    },
    {
      "entropy": 5.675894641876221,
      "epoch": 0.5447296007188185,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004978806449931854,
      "loss": 5.608,
      "mean_token_accuracy": 0.1721531130373478,
      "num_tokens": 10814508.0,
      "step": 4850
    },
    {
      "entropy": 5.6596986770629885,
      "epoch": 0.545291177626776,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004978751438570647,
      "loss": 5.7022,
      "mean_token_accuracy": 0.16957924216985704,
      "num_tokens": 10825924.0,
      "step": 4855
    },
    {
      "entropy": 5.720737171173096,
      "epoch": 0.5458527545347336,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004978696356245129,
      "loss": 5.6712,
      "mean_token_accuracy": 0.17078727930784227,
      "num_tokens": 10837854.0,
      "step": 4860
    },
    {
      "entropy": 5.685639333724976,
      "epoch": 0.546414331442691,
      "grad_norm": 1.0,
      "learning_rate": 0.0004978641202957057,
      "loss": 5.5789,
      "mean_token_accuracy": 0.17435066550970077,
      "num_tokens": 10848279.0,
      "step": 4865
    },
    {
      "entropy": 5.567231512069702,
      "epoch": 0.5469759083506486,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004978585978708183,
      "loss": 5.5833,
      "mean_token_accuracy": 0.1691250056028366,
      "num_tokens": 10859615.0,
      "step": 4870
    },
    {
      "entropy": 5.794380474090576,
      "epoch": 0.5475374852586061,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004978530683500267,
      "loss": 5.5611,
      "mean_token_accuracy": 0.17468865364789962,
      "num_tokens": 10871388.0,
      "step": 4875
    },
    {
      "entropy": 5.555207204818726,
      "epoch": 0.5480990621665637,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004978475317335071,
      "loss": 5.615,
      "mean_token_accuracy": 0.1761869341135025,
      "num_tokens": 10882411.0,
      "step": 4880
    },
    {
      "entropy": 5.60682110786438,
      "epoch": 0.5486606390745212,
      "grad_norm": 0.89453125,
      "learning_rate": 0.0004978419880214356,
      "loss": 5.7231,
      "mean_token_accuracy": 0.1712908700108528,
      "num_tokens": 10894116.0,
      "step": 4885
    },
    {
      "entropy": 5.646119260787964,
      "epoch": 0.5492222159824788,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004978364372139887,
      "loss": 5.5101,
      "mean_token_accuracy": 0.18172912746667863,
      "num_tokens": 10904514.0,
      "step": 4890
    },
    {
      "entropy": 5.563148355484008,
      "epoch": 0.5497837928904363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004978308793113434,
      "loss": 5.5456,
      "mean_token_accuracy": 0.17463553845882415,
      "num_tokens": 10915602.0,
      "step": 4895
    },
    {
      "entropy": 5.59888596534729,
      "epoch": 0.5503453697983939,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978253143136762,
      "loss": 5.5249,
      "mean_token_accuracy": 0.17883053570985794,
      "num_tokens": 10926172.0,
      "step": 4900
    },
    {
      "entropy": 5.746354436874389,
      "epoch": 0.5509069467063514,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004978197422211647,
      "loss": 5.664,
      "mean_token_accuracy": 0.1679267555475235,
      "num_tokens": 10938435.0,
      "step": 4905
    },
    {
      "entropy": 5.593893718719483,
      "epoch": 0.551468523614309,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004978141630339863,
      "loss": 5.5407,
      "mean_token_accuracy": 0.1758596658706665,
      "num_tokens": 10949118.0,
      "step": 4910
    },
    {
      "entropy": 5.604520559310913,
      "epoch": 0.5520301005222665,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004978085767523184,
      "loss": 5.5908,
      "mean_token_accuracy": 0.17219796776771545,
      "num_tokens": 10960120.0,
      "step": 4915
    },
    {
      "entropy": 5.648363637924194,
      "epoch": 0.5525916774302241,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497802983376339,
      "loss": 5.5598,
      "mean_token_accuracy": 0.17874993234872819,
      "num_tokens": 10971210.0,
      "step": 4920
    },
    {
      "entropy": 5.6162371158599855,
      "epoch": 0.5531532543381816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004977973829062263,
      "loss": 5.5465,
      "mean_token_accuracy": 0.18135998547077178,
      "num_tokens": 10983446.0,
      "step": 4925
    },
    {
      "entropy": 5.704362583160401,
      "epoch": 0.5537148312461392,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004977917753421585,
      "loss": 5.6708,
      "mean_token_accuracy": 0.16886882930994035,
      "num_tokens": 10994776.0,
      "step": 4930
    },
    {
      "entropy": 5.6581809520721436,
      "epoch": 0.5542764081540967,
      "grad_norm": 0.984375,
      "learning_rate": 0.000497786160684314,
      "loss": 5.5743,
      "mean_token_accuracy": 0.17270086109638214,
      "num_tokens": 11005280.0,
      "step": 4935
    },
    {
      "entropy": 5.579220867156982,
      "epoch": 0.5548379850620543,
      "grad_norm": 1.0,
      "learning_rate": 0.0004977805389328719,
      "loss": 5.5723,
      "mean_token_accuracy": 0.1701466292142868,
      "num_tokens": 11017509.0,
      "step": 4940
    },
    {
      "entropy": 5.687378692626953,
      "epoch": 0.5553995619700118,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497774910088011,
      "loss": 5.5985,
      "mean_token_accuracy": 0.17401109635829926,
      "num_tokens": 11029087.0,
      "step": 4945
    },
    {
      "entropy": 5.673885774612427,
      "epoch": 0.5559611388779694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004977692741499106,
      "loss": 5.6295,
      "mean_token_accuracy": 0.170565165579319,
      "num_tokens": 11041144.0,
      "step": 4950
    },
    {
      "entropy": 5.788410806655884,
      "epoch": 0.5565227157859269,
      "grad_norm": 1.015625,
      "learning_rate": 0.00049776363111875,
      "loss": 5.7193,
      "mean_token_accuracy": 0.16906611621379852,
      "num_tokens": 11054050.0,
      "step": 4955
    },
    {
      "entropy": 5.593288993835449,
      "epoch": 0.5570842926938844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004977579809947091,
      "loss": 5.5303,
      "mean_token_accuracy": 0.18044005632400512,
      "num_tokens": 11064257.0,
      "step": 4960
    },
    {
      "entropy": 5.662447118759156,
      "epoch": 0.5576458696018419,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977523237779676,
      "loss": 5.651,
      "mean_token_accuracy": 0.17158914357423782,
      "num_tokens": 11075568.0,
      "step": 4965
    },
    {
      "entropy": 5.703676795959472,
      "epoch": 0.5582074465097995,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004977466594687057,
      "loss": 5.6041,
      "mean_token_accuracy": 0.1758514553308487,
      "num_tokens": 11087078.0,
      "step": 4970
    },
    {
      "entropy": 5.784855556488037,
      "epoch": 0.558769023417757,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004977409880671039,
      "loss": 5.7419,
      "mean_token_accuracy": 0.16748689413070678,
      "num_tokens": 11100655.0,
      "step": 4975
    },
    {
      "entropy": 5.736449909210205,
      "epoch": 0.5593306003257146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977353095733427,
      "loss": 5.6459,
      "mean_token_accuracy": 0.17205908745527268,
      "num_tokens": 11110675.0,
      "step": 4980
    },
    {
      "entropy": 5.663420724868774,
      "epoch": 0.5598921772336721,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004977296239876026,
      "loss": 5.6393,
      "mean_token_accuracy": 0.1684981942176819,
      "num_tokens": 11122337.0,
      "step": 4985
    },
    {
      "entropy": 5.765861749649048,
      "epoch": 0.5604537541416297,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497723931310065,
      "loss": 5.5544,
      "mean_token_accuracy": 0.17807243168354034,
      "num_tokens": 11132672.0,
      "step": 4990
    },
    {
      "entropy": 5.574818468093872,
      "epoch": 0.5610153310495872,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004977182315409111,
      "loss": 5.5773,
      "mean_token_accuracy": 0.18216426223516463,
      "num_tokens": 11143740.0,
      "step": 4995
    },
    {
      "entropy": 5.579388809204102,
      "epoch": 0.5615769079575448,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004977125246803224,
      "loss": 5.5413,
      "mean_token_accuracy": 0.17574964314699174,
      "num_tokens": 11153722.0,
      "step": 5000
    },
    {
      "entropy": 5.735503721237182,
      "epoch": 0.5621384848655023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977068107284803,
      "loss": 5.6446,
      "mean_token_accuracy": 0.1708996534347534,
      "num_tokens": 11165042.0,
      "step": 5005
    },
    {
      "entropy": 5.624398469924927,
      "epoch": 0.5627000617734599,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004977010896855672,
      "loss": 5.5252,
      "mean_token_accuracy": 0.18079792857170104,
      "num_tokens": 11176834.0,
      "step": 5010
    },
    {
      "entropy": 5.622362422943115,
      "epoch": 0.5632616386814174,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004976953615517649,
      "loss": 5.5907,
      "mean_token_accuracy": 0.177802574634552,
      "num_tokens": 11188442.0,
      "step": 5015
    },
    {
      "entropy": 5.6082813262939455,
      "epoch": 0.563823215589375,
      "grad_norm": 1.015625,
      "learning_rate": 0.000497689626327256,
      "loss": 5.5235,
      "mean_token_accuracy": 0.18011727482080458,
      "num_tokens": 11197855.0,
      "step": 5020
    },
    {
      "entropy": 5.672953367233276,
      "epoch": 0.5643847924973325,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000497683884012223,
      "loss": 5.6007,
      "mean_token_accuracy": 0.17265776991844178,
      "num_tokens": 11209300.0,
      "step": 5025
    },
    {
      "entropy": 5.63689956665039,
      "epoch": 0.5649463694052901,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004976781346068487,
      "loss": 5.6224,
      "mean_token_accuracy": 0.17184539884328842,
      "num_tokens": 11222175.0,
      "step": 5030
    },
    {
      "entropy": 5.737363958358765,
      "epoch": 0.5655079463132476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004976723781113163,
      "loss": 5.609,
      "mean_token_accuracy": 0.17474291622638702,
      "num_tokens": 11233277.0,
      "step": 5035
    },
    {
      "entropy": 5.693420314788819,
      "epoch": 0.5660695232212052,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004976666145258089,
      "loss": 5.6556,
      "mean_token_accuracy": 0.17394862473011016,
      "num_tokens": 11245241.0,
      "step": 5040
    },
    {
      "entropy": 5.718694019317627,
      "epoch": 0.5666311001291627,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976608438505103,
      "loss": 5.6608,
      "mean_token_accuracy": 0.17288564145565033,
      "num_tokens": 11257359.0,
      "step": 5045
    },
    {
      "entropy": 5.603536891937256,
      "epoch": 0.5671926770371203,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976550660856041,
      "loss": 5.5753,
      "mean_token_accuracy": 0.1767792820930481,
      "num_tokens": 11270651.0,
      "step": 5050
    },
    {
      "entropy": 5.605350923538208,
      "epoch": 0.5677542539450777,
      "grad_norm": 1.0625,
      "learning_rate": 0.000497649281231274,
      "loss": 5.564,
      "mean_token_accuracy": 0.17950876504182817,
      "num_tokens": 11280164.0,
      "step": 5055
    },
    {
      "entropy": 5.632503318786621,
      "epoch": 0.5683158308530353,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004976434892877046,
      "loss": 5.515,
      "mean_token_accuracy": 0.17546210587024688,
      "num_tokens": 11290156.0,
      "step": 5060
    },
    {
      "entropy": 5.623936700820923,
      "epoch": 0.5688774077609928,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004976376902550802,
      "loss": 5.5543,
      "mean_token_accuracy": 0.17809658050537108,
      "num_tokens": 11301929.0,
      "step": 5065
    },
    {
      "entropy": 5.55612154006958,
      "epoch": 0.5694389846689504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004976318841335854,
      "loss": 5.4894,
      "mean_token_accuracy": 0.17908723801374435,
      "num_tokens": 11312805.0,
      "step": 5070
    },
    {
      "entropy": 5.473297786712647,
      "epoch": 0.5700005615769079,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004976260709234049,
      "loss": 5.4005,
      "mean_token_accuracy": 0.1841706857085228,
      "num_tokens": 11322316.0,
      "step": 5075
    },
    {
      "entropy": 5.590039253234863,
      "epoch": 0.5705621384848655,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000497620250624724,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17745646834373474,
      "num_tokens": 11332994.0,
      "step": 5080
    },
    {
      "entropy": 5.715634965896607,
      "epoch": 0.571123715392823,
      "grad_norm": 0.921875,
      "learning_rate": 0.000497614423237728,
      "loss": 5.5777,
      "mean_token_accuracy": 0.1673261821269989,
      "num_tokens": 11344355.0,
      "step": 5085
    },
    {
      "entropy": 5.604832792282105,
      "epoch": 0.5716852923007806,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004976085887626026,
      "loss": 5.5996,
      "mean_token_accuracy": 0.16973475962877274,
      "num_tokens": 11356275.0,
      "step": 5090
    },
    {
      "entropy": 5.658981513977051,
      "epoch": 0.5722468692087381,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0004976027471995331,
      "loss": 5.5899,
      "mean_token_accuracy": 0.1737036183476448,
      "num_tokens": 11368624.0,
      "step": 5095
    },
    {
      "entropy": 5.7313109874725345,
      "epoch": 0.5728084461166957,
      "grad_norm": 0.96875,
      "learning_rate": 0.000497596898548706,
      "loss": 5.6984,
      "mean_token_accuracy": 0.170048588514328,
      "num_tokens": 11380520.0,
      "step": 5100
    },
    {
      "entropy": 5.563168954849243,
      "epoch": 0.5733700230246532,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975910428103072,
      "loss": 5.5951,
      "mean_token_accuracy": 0.1739296168088913,
      "num_tokens": 11392293.0,
      "step": 5105
    },
    {
      "entropy": 5.713708162307739,
      "epoch": 0.5739315999326108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004975851799845233,
      "loss": 5.6549,
      "mean_token_accuracy": 0.17305538058280945,
      "num_tokens": 11403866.0,
      "step": 5110
    },
    {
      "entropy": 5.654724073410034,
      "epoch": 0.5744931768405683,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000497579310071541,
      "loss": 5.5005,
      "mean_token_accuracy": 0.18463232219219208,
      "num_tokens": 11415334.0,
      "step": 5115
    },
    {
      "entropy": 5.587343692779541,
      "epoch": 0.5750547537485259,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004975734330715471,
      "loss": 5.6231,
      "mean_token_accuracy": 0.1778287947177887,
      "num_tokens": 11426840.0,
      "step": 5120
    },
    {
      "entropy": 5.804072618484497,
      "epoch": 0.5756163306564834,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004975675489847287,
      "loss": 5.7792,
      "mean_token_accuracy": 0.16639406532049178,
      "num_tokens": 11438497.0,
      "step": 5125
    },
    {
      "entropy": 5.651816749572754,
      "epoch": 0.576177907564441,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004975616578112734,
      "loss": 5.5173,
      "mean_token_accuracy": 0.1801862582564354,
      "num_tokens": 11448705.0,
      "step": 5130
    },
    {
      "entropy": 5.7441857814788815,
      "epoch": 0.5767394844723985,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004975557595513684,
      "loss": 5.7661,
      "mean_token_accuracy": 0.1652762323617935,
      "num_tokens": 11460610.0,
      "step": 5135
    },
    {
      "entropy": 5.656605529785156,
      "epoch": 0.5773010613803561,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004975498542052018,
      "loss": 5.5253,
      "mean_token_accuracy": 0.17764451652765273,
      "num_tokens": 11471170.0,
      "step": 5140
    },
    {
      "entropy": 5.592125511169433,
      "epoch": 0.5778626382883136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004975439417729615,
      "loss": 5.5162,
      "mean_token_accuracy": 0.1843281075358391,
      "num_tokens": 11482090.0,
      "step": 5145
    },
    {
      "entropy": 5.574777841567993,
      "epoch": 0.5784242151962711,
      "grad_norm": 1.0,
      "learning_rate": 0.0004975380222548358,
      "loss": 5.6152,
      "mean_token_accuracy": 0.17407711446285248,
      "num_tokens": 11493650.0,
      "step": 5150
    },
    {
      "entropy": 5.682715988159179,
      "epoch": 0.5789857921042286,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004975320956510131,
      "loss": 5.5923,
      "mean_token_accuracy": 0.16880524009466172,
      "num_tokens": 11505903.0,
      "step": 5155
    },
    {
      "entropy": 5.754818487167358,
      "epoch": 0.5795473690121862,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004975261619616822,
      "loss": 5.6363,
      "mean_token_accuracy": 0.1682542994618416,
      "num_tokens": 11517194.0,
      "step": 5160
    },
    {
      "entropy": 5.6626605033874515,
      "epoch": 0.5801089459201437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004975202211870321,
      "loss": 5.6421,
      "mean_token_accuracy": 0.1754619747400284,
      "num_tokens": 11528683.0,
      "step": 5165
    },
    {
      "entropy": 5.732306480407715,
      "epoch": 0.5806705228281013,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004975142733272518,
      "loss": 5.6659,
      "mean_token_accuracy": 0.17479248046875,
      "num_tokens": 11541940.0,
      "step": 5170
    },
    {
      "entropy": 5.617989826202392,
      "epoch": 0.5812320997360588,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004975083183825306,
      "loss": 5.5512,
      "mean_token_accuracy": 0.17255982160568237,
      "num_tokens": 11553045.0,
      "step": 5175
    },
    {
      "entropy": 5.498443698883056,
      "epoch": 0.5817936766440164,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004975023563530584,
      "loss": 5.5262,
      "mean_token_accuracy": 0.17844814211130142,
      "num_tokens": 11564401.0,
      "step": 5180
    },
    {
      "entropy": 5.637786388397217,
      "epoch": 0.5823552535519739,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004974963872390249,
      "loss": 5.5623,
      "mean_token_accuracy": 0.1751536950469017,
      "num_tokens": 11576424.0,
      "step": 5185
    },
    {
      "entropy": 5.643254089355469,
      "epoch": 0.5829168304599315,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00049749041104062,
      "loss": 5.5665,
      "mean_token_accuracy": 0.1733489826321602,
      "num_tokens": 11587853.0,
      "step": 5190
    },
    {
      "entropy": 5.64462890625,
      "epoch": 0.583478407367889,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004974844277580343,
      "loss": 5.559,
      "mean_token_accuracy": 0.17800140827894212,
      "num_tokens": 11597938.0,
      "step": 5195
    },
    {
      "entropy": 5.780268383026123,
      "epoch": 0.5840399842758466,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004974784373914579,
      "loss": 5.7311,
      "mean_token_accuracy": 0.17299314886331557,
      "num_tokens": 11610077.0,
      "step": 5200
    },
    {
      "entropy": 5.617901372909546,
      "epoch": 0.5846015611838041,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974724399410818,
      "loss": 5.6,
      "mean_token_accuracy": 0.1762971892952919,
      "num_tokens": 11622079.0,
      "step": 5205
    },
    {
      "entropy": 5.568739986419677,
      "epoch": 0.5851631380917617,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004974664354070971,
      "loss": 5.5204,
      "mean_token_accuracy": 0.18156790882349014,
      "num_tokens": 11633047.0,
      "step": 5210
    },
    {
      "entropy": 5.6070304870605465,
      "epoch": 0.5857247149997192,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004974604237896947,
      "loss": 5.4684,
      "mean_token_accuracy": 0.17941075563430786,
      "num_tokens": 11643858.0,
      "step": 5215
    },
    {
      "entropy": 5.5261218547821045,
      "epoch": 0.5862862919076768,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004974544050890662,
      "loss": 5.5059,
      "mean_token_accuracy": 0.18191392421722413,
      "num_tokens": 11654550.0,
      "step": 5220
    },
    {
      "entropy": 5.566579532623291,
      "epoch": 0.5868478688156343,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000497448379305403,
      "loss": 5.537,
      "mean_token_accuracy": 0.17683645486831664,
      "num_tokens": 11665787.0,
      "step": 5225
    },
    {
      "entropy": 5.652027702331543,
      "epoch": 0.5874094457235919,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974423464388973,
      "loss": 5.6438,
      "mean_token_accuracy": 0.17096551954746247,
      "num_tokens": 11677717.0,
      "step": 5230
    },
    {
      "entropy": 5.607941961288452,
      "epoch": 0.5879710226315494,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000497436306489741,
      "loss": 5.5489,
      "mean_token_accuracy": 0.1790213853120804,
      "num_tokens": 11687714.0,
      "step": 5235
    },
    {
      "entropy": 5.5932074069976805,
      "epoch": 0.588532599539507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004974302594581264,
      "loss": 5.4333,
      "mean_token_accuracy": 0.18538083881139755,
      "num_tokens": 11698777.0,
      "step": 5240
    },
    {
      "entropy": 5.540124750137329,
      "epoch": 0.5890941764474645,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004974242053442461,
      "loss": 5.5277,
      "mean_token_accuracy": 0.1804819867014885,
      "num_tokens": 11710922.0,
      "step": 5245
    },
    {
      "entropy": 5.594572448730469,
      "epoch": 0.589655753355422,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004974181441482927,
      "loss": 5.5817,
      "mean_token_accuracy": 0.17460837811231614,
      "num_tokens": 11721450.0,
      "step": 5250
    },
    {
      "entropy": 5.636133337020874,
      "epoch": 0.5902173302633795,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004974120758704595,
      "loss": 5.492,
      "mean_token_accuracy": 0.17582428604364395,
      "num_tokens": 11732458.0,
      "step": 5255
    },
    {
      "entropy": 5.534214925765991,
      "epoch": 0.5907789071713371,
      "grad_norm": 1.0,
      "learning_rate": 0.0004974060005109394,
      "loss": 5.4545,
      "mean_token_accuracy": 0.17990118861198426,
      "num_tokens": 11742649.0,
      "step": 5260
    },
    {
      "entropy": 5.5487494468688965,
      "epoch": 0.5913404840792946,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000497399918069926,
      "loss": 5.4897,
      "mean_token_accuracy": 0.18184779286384584,
      "num_tokens": 11753811.0,
      "step": 5265
    },
    {
      "entropy": 5.645927667617798,
      "epoch": 0.5919020609872522,
      "grad_norm": 1.125,
      "learning_rate": 0.0004973938285476131,
      "loss": 5.5951,
      "mean_token_accuracy": 0.16961152851581573,
      "num_tokens": 11763965.0,
      "step": 5270
    },
    {
      "entropy": 5.691934967041016,
      "epoch": 0.5924636378952097,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973877319441943,
      "loss": 5.5892,
      "mean_token_accuracy": 0.17947103679180146,
      "num_tokens": 11774323.0,
      "step": 5275
    },
    {
      "entropy": 5.6107930660247805,
      "epoch": 0.5930252148031673,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973816282598639,
      "loss": 5.5683,
      "mean_token_accuracy": 0.1753183662891388,
      "num_tokens": 11786157.0,
      "step": 5280
    },
    {
      "entropy": 5.657044553756714,
      "epoch": 0.5935867917111248,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004973755174948162,
      "loss": 5.5511,
      "mean_token_accuracy": 0.18412764966487885,
      "num_tokens": 11797344.0,
      "step": 5285
    },
    {
      "entropy": 5.4586217403411865,
      "epoch": 0.5941483686190824,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004973693996492458,
      "loss": 5.4932,
      "mean_token_accuracy": 0.1788274332880974,
      "num_tokens": 11808149.0,
      "step": 5290
    },
    {
      "entropy": 5.567597675323486,
      "epoch": 0.5947099455270399,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004973632747233475,
      "loss": 5.5008,
      "mean_token_accuracy": 0.1850872039794922,
      "num_tokens": 11820203.0,
      "step": 5295
    },
    {
      "entropy": 5.647253799438476,
      "epoch": 0.5952715224349975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004973571427173163,
      "loss": 5.5847,
      "mean_token_accuracy": 0.18070656359195708,
      "num_tokens": 11830172.0,
      "step": 5300
    },
    {
      "entropy": 5.627700281143189,
      "epoch": 0.595833099342955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004973510036313473,
      "loss": 5.5473,
      "mean_token_accuracy": 0.1839230865240097,
      "num_tokens": 11840203.0,
      "step": 5305
    },
    {
      "entropy": 5.646620893478394,
      "epoch": 0.5963946762509126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004973448574656362,
      "loss": 5.4897,
      "mean_token_accuracy": 0.17596776634454728,
      "num_tokens": 11851002.0,
      "step": 5310
    },
    {
      "entropy": 5.572283935546875,
      "epoch": 0.5969562531588701,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004973387042203786,
      "loss": 5.5072,
      "mean_token_accuracy": 0.17337351143360138,
      "num_tokens": 11861559.0,
      "step": 5315
    },
    {
      "entropy": 5.555094242095947,
      "epoch": 0.5975178300668277,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004973325438957704,
      "loss": 5.5849,
      "mean_token_accuracy": 0.17335902750492097,
      "num_tokens": 11872535.0,
      "step": 5320
    },
    {
      "entropy": 5.697778701782227,
      "epoch": 0.5980794069747852,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004973263764920079,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17950138449668884,
      "num_tokens": 11884308.0,
      "step": 5325
    },
    {
      "entropy": 5.722074222564697,
      "epoch": 0.5986409838827428,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004973202020092871,
      "loss": 5.6691,
      "mean_token_accuracy": 0.17130828499794007,
      "num_tokens": 11896136.0,
      "step": 5330
    },
    {
      "entropy": 5.6341705322265625,
      "epoch": 0.5992025607907003,
      "grad_norm": 1.125,
      "learning_rate": 0.000497314020447805,
      "loss": 5.5029,
      "mean_token_accuracy": 0.1843678131699562,
      "num_tokens": 11906497.0,
      "step": 5335
    },
    {
      "entropy": 5.684994745254516,
      "epoch": 0.5997641376986579,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004973078318077583,
      "loss": 5.6137,
      "mean_token_accuracy": 0.1684497594833374,
      "num_tokens": 11917872.0,
      "step": 5340
    },
    {
      "entropy": 5.670975875854492,
      "epoch": 0.6003257146066153,
      "grad_norm": 1.0625,
      "learning_rate": 0.000497301636089344,
      "loss": 5.6527,
      "mean_token_accuracy": 0.17195726484060286,
      "num_tokens": 11929727.0,
      "step": 5345
    },
    {
      "entropy": 5.599681425094604,
      "epoch": 0.6008872915145729,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972954332927592,
      "loss": 5.4195,
      "mean_token_accuracy": 0.17955358028411866,
      "num_tokens": 11939996.0,
      "step": 5350
    },
    {
      "entropy": 5.577372598648071,
      "epoch": 0.6014488684225304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004972892234182018,
      "loss": 5.5118,
      "mean_token_accuracy": 0.1806427389383316,
      "num_tokens": 11950683.0,
      "step": 5355
    },
    {
      "entropy": 5.557851648330688,
      "epoch": 0.602010445330488,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004972830064658692,
      "loss": 5.5002,
      "mean_token_accuracy": 0.17415184229612352,
      "num_tokens": 11962447.0,
      "step": 5360
    },
    {
      "entropy": 5.675047206878662,
      "epoch": 0.6025720222384455,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004972767824359596,
      "loss": 5.6029,
      "mean_token_accuracy": 0.1744776502251625,
      "num_tokens": 11972740.0,
      "step": 5365
    },
    {
      "entropy": 5.602142953872681,
      "epoch": 0.6031335991464031,
      "grad_norm": 1.0,
      "learning_rate": 0.0004972705513286709,
      "loss": 5.4694,
      "mean_token_accuracy": 0.1808154582977295,
      "num_tokens": 11983292.0,
      "step": 5370
    },
    {
      "entropy": 5.672557640075683,
      "epoch": 0.6036951760543606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972643131442017,
      "loss": 5.6798,
      "mean_token_accuracy": 0.1694831445813179,
      "num_tokens": 11996384.0,
      "step": 5375
    },
    {
      "entropy": 5.71242208480835,
      "epoch": 0.6042567529623182,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972580678827506,
      "loss": 5.6136,
      "mean_token_accuracy": 0.1834661290049553,
      "num_tokens": 12006938.0,
      "step": 5380
    },
    {
      "entropy": 5.487227725982666,
      "epoch": 0.6048183298702757,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972518155445163,
      "loss": 5.433,
      "mean_token_accuracy": 0.17905438095331191,
      "num_tokens": 12018729.0,
      "step": 5385
    },
    {
      "entropy": 5.794947671890259,
      "epoch": 0.6053799067782333,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004972455561296982,
      "loss": 5.791,
      "mean_token_accuracy": 0.16544435620307923,
      "num_tokens": 12029625.0,
      "step": 5390
    },
    {
      "entropy": 5.640093898773193,
      "epoch": 0.6059414836861908,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004972392896384952,
      "loss": 5.4831,
      "mean_token_accuracy": 0.17690706700086595,
      "num_tokens": 12040243.0,
      "step": 5395
    },
    {
      "entropy": 5.640956211090088,
      "epoch": 0.6065030605941484,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004972330160711071,
      "loss": 5.7548,
      "mean_token_accuracy": 0.16721907407045364,
      "num_tokens": 12051961.0,
      "step": 5400
    },
    {
      "entropy": 5.6837477684021,
      "epoch": 0.6070646375021059,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004972267354277336,
      "loss": 5.6222,
      "mean_token_accuracy": 0.17716004103422164,
      "num_tokens": 12063969.0,
      "step": 5405
    },
    {
      "entropy": 5.632869482040405,
      "epoch": 0.6076262144100635,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004972204477085746,
      "loss": 5.512,
      "mean_token_accuracy": 0.17920950055122375,
      "num_tokens": 12075682.0,
      "step": 5410
    },
    {
      "entropy": 5.664663791656494,
      "epoch": 0.608187791318021,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004972141529138303,
      "loss": 5.6542,
      "mean_token_accuracy": 0.17303356230258943,
      "num_tokens": 12086595.0,
      "step": 5415
    },
    {
      "entropy": 5.678081321716308,
      "epoch": 0.6087493682259786,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004972078510437013,
      "loss": 5.5295,
      "mean_token_accuracy": 0.17658078968524932,
      "num_tokens": 12099421.0,
      "step": 5420
    },
    {
      "entropy": 5.688371849060059,
      "epoch": 0.6093109451339361,
      "grad_norm": 1.125,
      "learning_rate": 0.0004972015420983882,
      "loss": 5.6419,
      "mean_token_accuracy": 0.17382928282022475,
      "num_tokens": 12109680.0,
      "step": 5425
    },
    {
      "entropy": 5.5488300800323485,
      "epoch": 0.6098725220418937,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004971952260780915,
      "loss": 5.5322,
      "mean_token_accuracy": 0.1834533169865608,
      "num_tokens": 12121196.0,
      "step": 5430
    },
    {
      "entropy": 5.58459439277649,
      "epoch": 0.6104340989498512,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000497188902983013,
      "loss": 5.4618,
      "mean_token_accuracy": 0.17039896100759505,
      "num_tokens": 12132262.0,
      "step": 5435
    },
    {
      "entropy": 5.680936861038208,
      "epoch": 0.6109956758578087,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004971825728133535,
      "loss": 5.6389,
      "mean_token_accuracy": 0.17605651170015335,
      "num_tokens": 12143747.0,
      "step": 5440
    },
    {
      "entropy": 5.822443008422852,
      "epoch": 0.6115572527657662,
      "grad_norm": 1.0,
      "learning_rate": 0.0004971762355693146,
      "loss": 5.7316,
      "mean_token_accuracy": 0.16901175826787948,
      "num_tokens": 12155887.0,
      "step": 5445
    },
    {
      "entropy": 5.557335662841797,
      "epoch": 0.6121188296737238,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971698912510983,
      "loss": 5.5106,
      "mean_token_accuracy": 0.1743317186832428,
      "num_tokens": 12166718.0,
      "step": 5450
    },
    {
      "entropy": 5.637910413742065,
      "epoch": 0.6126804065816813,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004971635398589066,
      "loss": 5.6718,
      "mean_token_accuracy": 0.17733672857284546,
      "num_tokens": 12178540.0,
      "step": 5455
    },
    {
      "entropy": 5.852599477767944,
      "epoch": 0.6132419834896389,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004971571813929415,
      "loss": 5.7001,
      "mean_token_accuracy": 0.16884040534496308,
      "num_tokens": 12189820.0,
      "step": 5460
    },
    {
      "entropy": 5.640134477615357,
      "epoch": 0.6138035603975964,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004971508158534056,
      "loss": 5.6151,
      "mean_token_accuracy": 0.17462412565946578,
      "num_tokens": 12201679.0,
      "step": 5465
    },
    {
      "entropy": 5.628084373474121,
      "epoch": 0.614365137305554,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004971444432405015,
      "loss": 5.6023,
      "mean_token_accuracy": 0.1724008411169052,
      "num_tokens": 12213793.0,
      "step": 5470
    },
    {
      "entropy": 5.793092823028564,
      "epoch": 0.6149267142135115,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004971380635544323,
      "loss": 5.7036,
      "mean_token_accuracy": 0.1700220748782158,
      "num_tokens": 12224519.0,
      "step": 5475
    },
    {
      "entropy": 5.735996389389038,
      "epoch": 0.6154882911214691,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004971316767954009,
      "loss": 5.6797,
      "mean_token_accuracy": 0.17044796645641327,
      "num_tokens": 12236401.0,
      "step": 5480
    },
    {
      "entropy": 5.675604009628296,
      "epoch": 0.6160498680294266,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004971252829636108,
      "loss": 5.7181,
      "mean_token_accuracy": 0.1707397386431694,
      "num_tokens": 12247670.0,
      "step": 5485
    },
    {
      "entropy": 5.616259431838989,
      "epoch": 0.6166114449373842,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004971188820592656,
      "loss": 5.5748,
      "mean_token_accuracy": 0.17626833766698838,
      "num_tokens": 12258877.0,
      "step": 5490
    },
    {
      "entropy": 5.587227010726929,
      "epoch": 0.6171730218453417,
      "grad_norm": 1.03125,
      "learning_rate": 0.000497112474082569,
      "loss": 5.5507,
      "mean_token_accuracy": 0.17732103168964386,
      "num_tokens": 12268507.0,
      "step": 5495
    },
    {
      "entropy": 5.604141807556152,
      "epoch": 0.6177345987532993,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000497106059033725,
      "loss": 5.5476,
      "mean_token_accuracy": 0.1763266682624817,
      "num_tokens": 12280420.0,
      "step": 5500
    },
    {
      "entropy": 5.58247332572937,
      "epoch": 0.6182961756612568,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970996369129381,
      "loss": 5.5396,
      "mean_token_accuracy": 0.1739584818482399,
      "num_tokens": 12290876.0,
      "step": 5505
    },
    {
      "entropy": 5.618482065200806,
      "epoch": 0.6188577525692144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004970932077204125,
      "loss": 5.4521,
      "mean_token_accuracy": 0.18287576586008072,
      "num_tokens": 12301831.0,
      "step": 5510
    },
    {
      "entropy": 5.5254891872406,
      "epoch": 0.6194193294771719,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000497086771456353,
      "loss": 5.4799,
      "mean_token_accuracy": 0.18308470398187637,
      "num_tokens": 12312678.0,
      "step": 5515
    },
    {
      "entropy": 5.668888568878174,
      "epoch": 0.6199809063851295,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004970803281209646,
      "loss": 5.6543,
      "mean_token_accuracy": 0.1691954791545868,
      "num_tokens": 12322905.0,
      "step": 5520
    },
    {
      "entropy": 5.61128077507019,
      "epoch": 0.620542483293087,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004970738777144524,
      "loss": 5.522,
      "mean_token_accuracy": 0.17409571260213852,
      "num_tokens": 12333091.0,
      "step": 5525
    },
    {
      "entropy": 5.635380601882934,
      "epoch": 0.6211040602010446,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970674202370218,
      "loss": 5.5052,
      "mean_token_accuracy": 0.17812769263982772,
      "num_tokens": 12346310.0,
      "step": 5530
    },
    {
      "entropy": 5.550295639038086,
      "epoch": 0.621665637109002,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004970609556888785,
      "loss": 5.4492,
      "mean_token_accuracy": 0.18000506311655046,
      "num_tokens": 12357138.0,
      "step": 5535
    },
    {
      "entropy": 5.611942100524902,
      "epoch": 0.6222272140169596,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970544840702282,
      "loss": 5.5436,
      "mean_token_accuracy": 0.18242397010326386,
      "num_tokens": 12367382.0,
      "step": 5540
    },
    {
      "entropy": 5.621787214279175,
      "epoch": 0.6227887909249171,
      "grad_norm": 1.0,
      "learning_rate": 0.0004970480053812771,
      "loss": 5.5237,
      "mean_token_accuracy": 0.17412640750408173,
      "num_tokens": 12380357.0,
      "step": 5545
    },
    {
      "entropy": 5.7478477478027346,
      "epoch": 0.6233503678328747,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004970415196222312,
      "loss": 5.7008,
      "mean_token_accuracy": 0.1674142822623253,
      "num_tokens": 12392187.0,
      "step": 5550
    },
    {
      "entropy": 5.610329580307007,
      "epoch": 0.6239119447408322,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004970350267932973,
      "loss": 5.5745,
      "mean_token_accuracy": 0.1778603732585907,
      "num_tokens": 12402924.0,
      "step": 5555
    },
    {
      "entropy": 5.691086149215698,
      "epoch": 0.6244735216487898,
      "grad_norm": 1.046875,
      "learning_rate": 0.000497028526894682,
      "loss": 5.6886,
      "mean_token_accuracy": 0.1716107726097107,
      "num_tokens": 12413554.0,
      "step": 5560
    },
    {
      "entropy": 5.7034632682800295,
      "epoch": 0.6250350985567473,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004970220199265923,
      "loss": 5.5744,
      "mean_token_accuracy": 0.18121662437915803,
      "num_tokens": 12424525.0,
      "step": 5565
    },
    {
      "entropy": 5.559814405441284,
      "epoch": 0.6255966754647049,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004970155058892354,
      "loss": 5.581,
      "mean_token_accuracy": 0.17716656923294066,
      "num_tokens": 12436405.0,
      "step": 5570
    },
    {
      "entropy": 5.686614608764648,
      "epoch": 0.6261582523726624,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004970089847828187,
      "loss": 5.6343,
      "mean_token_accuracy": 0.17088887840509415,
      "num_tokens": 12446909.0,
      "step": 5575
    },
    {
      "entropy": 5.734358930587769,
      "epoch": 0.62671982928062,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004970024566075497,
      "loss": 5.626,
      "mean_token_accuracy": 0.17589510679244996,
      "num_tokens": 12458739.0,
      "step": 5580
    },
    {
      "entropy": 5.599149751663208,
      "epoch": 0.6272814061885775,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969959213636365,
      "loss": 5.5423,
      "mean_token_accuracy": 0.17475462853908538,
      "num_tokens": 12469529.0,
      "step": 5585
    },
    {
      "entropy": 5.692594957351685,
      "epoch": 0.6278429830965351,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004969893790512871,
      "loss": 5.5706,
      "mean_token_accuracy": 0.16985998451709747,
      "num_tokens": 12480067.0,
      "step": 5590
    },
    {
      "entropy": 5.566493225097656,
      "epoch": 0.6284045600044926,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004969828296707097,
      "loss": 5.4704,
      "mean_token_accuracy": 0.18619227260351182,
      "num_tokens": 12491043.0,
      "step": 5595
    },
    {
      "entropy": 5.515522480010986,
      "epoch": 0.6289661369124502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004969762732221129,
      "loss": 5.4015,
      "mean_token_accuracy": 0.1820645272731781,
      "num_tokens": 12501852.0,
      "step": 5600
    },
    {
      "entropy": 5.637913942337036,
      "epoch": 0.6295277138204077,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004969697097057054,
      "loss": 5.685,
      "mean_token_accuracy": 0.17365383878350257,
      "num_tokens": 12512696.0,
      "step": 5605
    },
    {
      "entropy": 5.63007664680481,
      "epoch": 0.6300892907283653,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004969631391216964,
      "loss": 5.5515,
      "mean_token_accuracy": 0.18211620002985002,
      "num_tokens": 12523643.0,
      "step": 5610
    },
    {
      "entropy": 5.633838653564453,
      "epoch": 0.6306508676363228,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969565614702948,
      "loss": 5.5304,
      "mean_token_accuracy": 0.17696837037801744,
      "num_tokens": 12534827.0,
      "step": 5615
    },
    {
      "entropy": 5.47821307182312,
      "epoch": 0.6312124445442804,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969499767517104,
      "loss": 5.4135,
      "mean_token_accuracy": 0.17736863791942598,
      "num_tokens": 12545977.0,
      "step": 5620
    },
    {
      "entropy": 5.4920238018035885,
      "epoch": 0.6317740214522379,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004969433849661525,
      "loss": 5.472,
      "mean_token_accuracy": 0.17436029016971588,
      "num_tokens": 12555890.0,
      "step": 5625
    },
    {
      "entropy": 5.616329145431519,
      "epoch": 0.6323355983601955,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004969367861138313,
      "loss": 5.5401,
      "mean_token_accuracy": 0.18098318427801133,
      "num_tokens": 12566412.0,
      "step": 5630
    },
    {
      "entropy": 5.654417991638184,
      "epoch": 0.6328971752681529,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004969301801949565,
      "loss": 5.5611,
      "mean_token_accuracy": 0.1757127285003662,
      "num_tokens": 12577175.0,
      "step": 5635
    },
    {
      "entropy": 5.554512310028076,
      "epoch": 0.6334587521761105,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000496923567209739,
      "loss": 5.4884,
      "mean_token_accuracy": 0.17531097084283828,
      "num_tokens": 12588760.0,
      "step": 5640
    },
    {
      "entropy": 5.59276933670044,
      "epoch": 0.634020329084068,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004969169471583889,
      "loss": 5.5501,
      "mean_token_accuracy": 0.18024812787771224,
      "num_tokens": 12599899.0,
      "step": 5645
    },
    {
      "entropy": 5.6611004829406735,
      "epoch": 0.6345819059920256,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004969103200411171,
      "loss": 5.5313,
      "mean_token_accuracy": 0.1840333729982376,
      "num_tokens": 12610836.0,
      "step": 5650
    },
    {
      "entropy": 5.5443943500518795,
      "epoch": 0.6351434828999831,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004969036858581346,
      "loss": 5.4877,
      "mean_token_accuracy": 0.18098128587007523,
      "num_tokens": 12621855.0,
      "step": 5655
    },
    {
      "entropy": 5.638803386688233,
      "epoch": 0.6357050598079407,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968970446096527,
      "loss": 5.5795,
      "mean_token_accuracy": 0.17581598162651063,
      "num_tokens": 12632964.0,
      "step": 5660
    },
    {
      "entropy": 5.532229471206665,
      "epoch": 0.6362666367158982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000496890396295883,
      "loss": 5.4269,
      "mean_token_accuracy": 0.18607352674007416,
      "num_tokens": 12643717.0,
      "step": 5665
    },
    {
      "entropy": 5.618521451950073,
      "epoch": 0.6368282136238558,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004968837409170368,
      "loss": 5.5368,
      "mean_token_accuracy": 0.17255922853946687,
      "num_tokens": 12654491.0,
      "step": 5670
    },
    {
      "entropy": 5.58553786277771,
      "epoch": 0.6373897905318133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004968770784733263,
      "loss": 5.5919,
      "mean_token_accuracy": 0.1750367432832718,
      "num_tokens": 12666128.0,
      "step": 5675
    },
    {
      "entropy": 5.705466032028198,
      "epoch": 0.6379513674397709,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004968704089649636,
      "loss": 5.6676,
      "mean_token_accuracy": 0.1743661493062973,
      "num_tokens": 12677957.0,
      "step": 5680
    },
    {
      "entropy": 5.637051963806153,
      "epoch": 0.6385129443477284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004968637323921611,
      "loss": 5.5125,
      "mean_token_accuracy": 0.17991692274808885,
      "num_tokens": 12687964.0,
      "step": 5685
    },
    {
      "entropy": 5.47965292930603,
      "epoch": 0.639074521255686,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004968570487551312,
      "loss": 5.5173,
      "mean_token_accuracy": 0.17588139325380325,
      "num_tokens": 12698805.0,
      "step": 5690
    },
    {
      "entropy": 5.647631978988647,
      "epoch": 0.6396360981636435,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004968503580540869,
      "loss": 5.5205,
      "mean_token_accuracy": 0.17461606860160828,
      "num_tokens": 12709942.0,
      "step": 5695
    },
    {
      "entropy": 5.684104633331299,
      "epoch": 0.6401976750716011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004968436602892411,
      "loss": 5.6107,
      "mean_token_accuracy": 0.1750674232840538,
      "num_tokens": 12721856.0,
      "step": 5700
    },
    {
      "entropy": 5.487252569198608,
      "epoch": 0.6407592519795586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004968369554608071,
      "loss": 5.4359,
      "mean_token_accuracy": 0.17459376603364946,
      "num_tokens": 12732619.0,
      "step": 5705
    },
    {
      "entropy": 5.607162046432495,
      "epoch": 0.6413208288875162,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968302435689985,
      "loss": 5.5262,
      "mean_token_accuracy": 0.18006090074777603,
      "num_tokens": 12743478.0,
      "step": 5710
    },
    {
      "entropy": 5.573984622955322,
      "epoch": 0.6418824057954737,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004968235246140288,
      "loss": 5.4773,
      "mean_token_accuracy": 0.1702056735754013,
      "num_tokens": 12754761.0,
      "step": 5715
    },
    {
      "entropy": 5.586852550506592,
      "epoch": 0.6424439827034313,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004968167985961121,
      "loss": 5.5773,
      "mean_token_accuracy": 0.18108247220516205,
      "num_tokens": 12765803.0,
      "step": 5720
    },
    {
      "entropy": 5.659261798858642,
      "epoch": 0.6430055596113888,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004968100655154625,
      "loss": 5.6119,
      "mean_token_accuracy": 0.1718067392706871,
      "num_tokens": 12777882.0,
      "step": 5725
    },
    {
      "entropy": 5.696913862228394,
      "epoch": 0.6435671365193463,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004968033253722943,
      "loss": 5.6212,
      "mean_token_accuracy": 0.17264451831579208,
      "num_tokens": 12788268.0,
      "step": 5730
    },
    {
      "entropy": 5.590059232711792,
      "epoch": 0.6441287134273038,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004967965781668223,
      "loss": 5.6203,
      "mean_token_accuracy": 0.17548608183860778,
      "num_tokens": 12800506.0,
      "step": 5735
    },
    {
      "entropy": 5.624031591415405,
      "epoch": 0.6446902903352614,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004967898238992613,
      "loss": 5.583,
      "mean_token_accuracy": 0.17776862680912017,
      "num_tokens": 12813040.0,
      "step": 5740
    },
    {
      "entropy": 5.716559171676636,
      "epoch": 0.6452518672432189,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004967830625698262,
      "loss": 5.6083,
      "mean_token_accuracy": 0.17837660908699035,
      "num_tokens": 12825534.0,
      "step": 5745
    },
    {
      "entropy": 5.562139081954956,
      "epoch": 0.6458134441511765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967762941787323,
      "loss": 5.5135,
      "mean_token_accuracy": 0.17737233489751816,
      "num_tokens": 12835882.0,
      "step": 5750
    },
    {
      "entropy": 5.622623634338379,
      "epoch": 0.646375021059134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967695187261952,
      "loss": 5.6844,
      "mean_token_accuracy": 0.16672947406768798,
      "num_tokens": 12847663.0,
      "step": 5755
    },
    {
      "entropy": 5.864485597610473,
      "epoch": 0.6469365979670916,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004967627362124306,
      "loss": 5.814,
      "mean_token_accuracy": 0.16439027190208436,
      "num_tokens": 12860789.0,
      "step": 5760
    },
    {
      "entropy": 5.739031076431274,
      "epoch": 0.6474981748750491,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004967559466376546,
      "loss": 5.5668,
      "mean_token_accuracy": 0.17534830868244172,
      "num_tokens": 12872509.0,
      "step": 5765
    },
    {
      "entropy": 5.527986812591553,
      "epoch": 0.6480597517830067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004967491500020831,
      "loss": 5.5355,
      "mean_token_accuracy": 0.17819956243038176,
      "num_tokens": 12884321.0,
      "step": 5770
    },
    {
      "entropy": 5.460300588607788,
      "epoch": 0.6486213286909642,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004967423463059328,
      "loss": 5.4395,
      "mean_token_accuracy": 0.18405805379152299,
      "num_tokens": 12894559.0,
      "step": 5775
    },
    {
      "entropy": 5.622215032577515,
      "epoch": 0.6491829055989218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004967355355494201,
      "loss": 5.543,
      "mean_token_accuracy": 0.1831359088420868,
      "num_tokens": 12904934.0,
      "step": 5780
    },
    {
      "entropy": 5.748392343521118,
      "epoch": 0.6497444825068793,
      "grad_norm": 1.125,
      "learning_rate": 0.000496728717732762,
      "loss": 5.6551,
      "mean_token_accuracy": 0.16987751573324203,
      "num_tokens": 12916551.0,
      "step": 5785
    },
    {
      "entropy": 5.623045635223389,
      "epoch": 0.6503060594148369,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004967218928561755,
      "loss": 5.5708,
      "mean_token_accuracy": 0.1752079874277115,
      "num_tokens": 12928097.0,
      "step": 5790
    },
    {
      "entropy": 5.5887308597564695,
      "epoch": 0.6508676363227944,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004967150609198779,
      "loss": 5.448,
      "mean_token_accuracy": 0.18454979807138444,
      "num_tokens": 12939161.0,
      "step": 5795
    },
    {
      "entropy": 5.468107891082764,
      "epoch": 0.651429213230752,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004967082219240869,
      "loss": 5.4922,
      "mean_token_accuracy": 0.17835438400506973,
      "num_tokens": 12950862.0,
      "step": 5800
    },
    {
      "entropy": 5.615010929107666,
      "epoch": 0.6519907901387095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004967013758690199,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17528652250766755,
      "num_tokens": 12962333.0,
      "step": 5805
    },
    {
      "entropy": 5.648796701431275,
      "epoch": 0.6525523670466671,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966945227548953,
      "loss": 5.5172,
      "mean_token_accuracy": 0.1784067139029503,
      "num_tokens": 12973136.0,
      "step": 5810
    },
    {
      "entropy": 5.5914441585540775,
      "epoch": 0.6531139439546246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966876625819309,
      "loss": 5.5386,
      "mean_token_accuracy": 0.17675970792770385,
      "num_tokens": 12983602.0,
      "step": 5815
    },
    {
      "entropy": 5.576560640335083,
      "epoch": 0.6536755208625822,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004966807953503456,
      "loss": 5.4594,
      "mean_token_accuracy": 0.18512107878923417,
      "num_tokens": 12994129.0,
      "step": 5820
    },
    {
      "entropy": 5.606117010116577,
      "epoch": 0.6542370977705396,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966739210603578,
      "loss": 5.5323,
      "mean_token_accuracy": 0.17496279627084732,
      "num_tokens": 13005071.0,
      "step": 5825
    },
    {
      "entropy": 5.731236505508423,
      "epoch": 0.6547986746784972,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004966670397121862,
      "loss": 5.6894,
      "mean_token_accuracy": 0.17773077487945557,
      "num_tokens": 13016202.0,
      "step": 5830
    },
    {
      "entropy": 5.744341135025024,
      "epoch": 0.6553602515864547,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004966601513060502,
      "loss": 5.682,
      "mean_token_accuracy": 0.16525742560625076,
      "num_tokens": 13028468.0,
      "step": 5835
    },
    {
      "entropy": 5.667176151275635,
      "epoch": 0.6559218284944123,
      "grad_norm": 1.0625,
      "learning_rate": 0.000496653255842169,
      "loss": 5.5641,
      "mean_token_accuracy": 0.1750602275133133,
      "num_tokens": 13039667.0,
      "step": 5840
    },
    {
      "entropy": 5.584512424468994,
      "epoch": 0.6564834054023698,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966463533207621,
      "loss": 5.6284,
      "mean_token_accuracy": 0.17104295194149016,
      "num_tokens": 13051583.0,
      "step": 5845
    },
    {
      "entropy": 5.6398578643798825,
      "epoch": 0.6570449823103274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004966394437420493,
      "loss": 5.5785,
      "mean_token_accuracy": 0.17410314977169036,
      "num_tokens": 13062401.0,
      "step": 5850
    },
    {
      "entropy": 5.556688928604126,
      "epoch": 0.6576065592182849,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966325271062508,
      "loss": 5.464,
      "mean_token_accuracy": 0.17937254756689072,
      "num_tokens": 13074219.0,
      "step": 5855
    },
    {
      "entropy": 5.6250214099884035,
      "epoch": 0.6581681361262425,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004966256034135866,
      "loss": 5.6485,
      "mean_token_accuracy": 0.17628799974918366,
      "num_tokens": 13086073.0,
      "step": 5860
    },
    {
      "entropy": 5.673035478591919,
      "epoch": 0.6587297130342,
      "grad_norm": 1.0,
      "learning_rate": 0.0004966186726642772,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17301213294267653,
      "num_tokens": 13096340.0,
      "step": 5865
    },
    {
      "entropy": 5.575248861312867,
      "epoch": 0.6592912899421576,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004966117348585434,
      "loss": 5.5359,
      "mean_token_accuracy": 0.17825251668691636,
      "num_tokens": 13107273.0,
      "step": 5870
    },
    {
      "entropy": 5.593775987625122,
      "epoch": 0.6598528668501151,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004966047899966059,
      "loss": 5.5099,
      "mean_token_accuracy": 0.1744076833128929,
      "num_tokens": 13117388.0,
      "step": 5875
    },
    {
      "entropy": 5.6087267875671385,
      "epoch": 0.6604144437580727,
      "grad_norm": 1.078125,
      "learning_rate": 0.000496597838078686,
      "loss": 5.6227,
      "mean_token_accuracy": 0.17380800396203994,
      "num_tokens": 13128720.0,
      "step": 5880
    },
    {
      "entropy": 5.673722219467163,
      "epoch": 0.6609760206660302,
      "grad_norm": 1.078125,
      "learning_rate": 0.000496590879105005,
      "loss": 5.5462,
      "mean_token_accuracy": 0.18040583282709122,
      "num_tokens": 13139405.0,
      "step": 5885
    },
    {
      "entropy": 5.631203317642212,
      "epoch": 0.6615375975739878,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965839130757845,
      "loss": 5.6219,
      "mean_token_accuracy": 0.17713258415460587,
      "num_tokens": 13151408.0,
      "step": 5890
    },
    {
      "entropy": 5.62841911315918,
      "epoch": 0.6620991744819453,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965769399912462,
      "loss": 5.521,
      "mean_token_accuracy": 0.1794424533843994,
      "num_tokens": 13162592.0,
      "step": 5895
    },
    {
      "entropy": 5.70211763381958,
      "epoch": 0.6626607513899029,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004965699598516122,
      "loss": 5.5729,
      "mean_token_accuracy": 0.18016252666711807,
      "num_tokens": 13173040.0,
      "step": 5900
    },
    {
      "entropy": 5.585992050170899,
      "epoch": 0.6632223282978604,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004965629726571047,
      "loss": 5.6437,
      "mean_token_accuracy": 0.174474835395813,
      "num_tokens": 13185475.0,
      "step": 5905
    },
    {
      "entropy": 5.60920934677124,
      "epoch": 0.663783905205818,
      "grad_norm": 1.125,
      "learning_rate": 0.0004965559784079463,
      "loss": 5.5117,
      "mean_token_accuracy": 0.18488989025354385,
      "num_tokens": 13196481.0,
      "step": 5910
    },
    {
      "entropy": 5.572127103805542,
      "epoch": 0.6643454821137755,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004965489771043596,
      "loss": 5.4092,
      "mean_token_accuracy": 0.18228244930505752,
      "num_tokens": 13207947.0,
      "step": 5915
    },
    {
      "entropy": 5.435831546783447,
      "epoch": 0.664907059021733,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004965419687465676,
      "loss": 5.4454,
      "mean_token_accuracy": 0.19303639233112335,
      "num_tokens": 13219392.0,
      "step": 5920
    },
    {
      "entropy": 5.552551126480102,
      "epoch": 0.6654686359296905,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004965349533347934,
      "loss": 5.4831,
      "mean_token_accuracy": 0.18510687202215195,
      "num_tokens": 13231556.0,
      "step": 5925
    },
    {
      "entropy": 5.5612060546875,
      "epoch": 0.6660302128376481,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004965279308692602,
      "loss": 5.4252,
      "mean_token_accuracy": 0.18248932361602782,
      "num_tokens": 13241287.0,
      "step": 5930
    },
    {
      "entropy": 5.518649005889893,
      "epoch": 0.6665917897456056,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004965209013501919,
      "loss": 5.4774,
      "mean_token_accuracy": 0.17345127910375596,
      "num_tokens": 13254120.0,
      "step": 5935
    },
    {
      "entropy": 5.640679836273193,
      "epoch": 0.6671533666535632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004965138647778122,
      "loss": 5.5575,
      "mean_token_accuracy": 0.1721458002924919,
      "num_tokens": 13265745.0,
      "step": 5940
    },
    {
      "entropy": 5.675885009765625,
      "epoch": 0.6677149435615207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000496506821152345,
      "loss": 5.61,
      "mean_token_accuracy": 0.17708085775375365,
      "num_tokens": 13277791.0,
      "step": 5945
    },
    {
      "entropy": 5.5192066669464115,
      "epoch": 0.6682765204694783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004964997704740147,
      "loss": 5.4731,
      "mean_token_accuracy": 0.1778041511774063,
      "num_tokens": 13288078.0,
      "step": 5950
    },
    {
      "entropy": 5.625132369995117,
      "epoch": 0.6688380973774358,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000496492712743046,
      "loss": 5.5164,
      "mean_token_accuracy": 0.18207676112651824,
      "num_tokens": 13300547.0,
      "step": 5955
    },
    {
      "entropy": 5.636836814880371,
      "epoch": 0.6693996742853934,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964856479596632,
      "loss": 5.5938,
      "mean_token_accuracy": 0.17118555605411528,
      "num_tokens": 13310339.0,
      "step": 5960
    },
    {
      "entropy": 5.722489213943481,
      "epoch": 0.6699612511933509,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004964785761240915,
      "loss": 5.5579,
      "mean_token_accuracy": 0.17542766332626342,
      "num_tokens": 13321294.0,
      "step": 5965
    },
    {
      "entropy": 5.636022901535034,
      "epoch": 0.6705228281013085,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964714972365562,
      "loss": 5.5833,
      "mean_token_accuracy": 0.1756971925497055,
      "num_tokens": 13332608.0,
      "step": 5970
    },
    {
      "entropy": 5.593559837341308,
      "epoch": 0.671084405009266,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964644112972824,
      "loss": 5.5239,
      "mean_token_accuracy": 0.18908276557922363,
      "num_tokens": 13344169.0,
      "step": 5975
    },
    {
      "entropy": 5.552310132980347,
      "epoch": 0.6716459819172236,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004964573183064959,
      "loss": 5.4761,
      "mean_token_accuracy": 0.18131648302078246,
      "num_tokens": 13355713.0,
      "step": 5980
    },
    {
      "entropy": 5.644558382034302,
      "epoch": 0.6722075588251811,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964502182644225,
      "loss": 5.5829,
      "mean_token_accuracy": 0.17490034103393554,
      "num_tokens": 13366644.0,
      "step": 5985
    },
    {
      "entropy": 5.577767753601075,
      "epoch": 0.6727691357331387,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964431111712883,
      "loss": 5.5486,
      "mean_token_accuracy": 0.1742811381816864,
      "num_tokens": 13378125.0,
      "step": 5990
    },
    {
      "entropy": 5.63827543258667,
      "epoch": 0.6733307126410962,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004964359970273195,
      "loss": 5.5907,
      "mean_token_accuracy": 0.1733713924884796,
      "num_tokens": 13389681.0,
      "step": 5995
    },
    {
      "entropy": 5.720663785934448,
      "epoch": 0.6738922895490538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004964288758327426,
      "loss": 5.5889,
      "mean_token_accuracy": 0.17525839507579805,
      "num_tokens": 13399814.0,
      "step": 6000
    },
    {
      "epoch": 0.6738922895490538,
      "eval_entropy": 5.508367524554135,
      "eval_loss": 5.577201843261719,
      "eval_mean_token_accuracy": 0.18273891875823503,
      "eval_num_tokens": 13399814.0,
      "eval_runtime": 24.0646,
      "eval_samples_per_second": 1288.741,
      "eval_steps_per_second": 161.108,
      "step": 6000
    },
    {
      "entropy": 5.6744519710540775,
      "epoch": 0.6744538664570113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004964217475877846,
      "loss": 5.6104,
      "mean_token_accuracy": 0.1786242127418518,
      "num_tokens": 13410614.0,
      "step": 6005
    },
    {
      "entropy": 5.625640249252319,
      "epoch": 0.6750154433649689,
      "grad_norm": 1.109375,
      "learning_rate": 0.000496414612292672,
      "loss": 5.6077,
      "mean_token_accuracy": 0.1749860405921936,
      "num_tokens": 13421392.0,
      "step": 6010
    },
    {
      "entropy": 5.677163887023926,
      "epoch": 0.6755770202729264,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004964074699476325,
      "loss": 5.6333,
      "mean_token_accuracy": 0.17317446917295456,
      "num_tokens": 13432004.0,
      "step": 6015
    },
    {
      "entropy": 5.638656091690064,
      "epoch": 0.6761385971808839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004964003205528933,
      "loss": 5.4835,
      "mean_token_accuracy": 0.17764865159988402,
      "num_tokens": 13442450.0,
      "step": 6020
    },
    {
      "entropy": 5.593809747695923,
      "epoch": 0.6767001740888414,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963931641086819,
      "loss": 5.5356,
      "mean_token_accuracy": 0.17503505945205688,
      "num_tokens": 13453848.0,
      "step": 6025
    },
    {
      "entropy": 5.568034076690674,
      "epoch": 0.677261750996799,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004963860006152262,
      "loss": 5.5415,
      "mean_token_accuracy": 0.1778825968503952,
      "num_tokens": 13465119.0,
      "step": 6030
    },
    {
      "entropy": 5.5338887691497805,
      "epoch": 0.6778233279047565,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004963788300727546,
      "loss": 5.517,
      "mean_token_accuracy": 0.17714934796094894,
      "num_tokens": 13475937.0,
      "step": 6035
    },
    {
      "entropy": 5.615345764160156,
      "epoch": 0.6783849048127141,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004963716524814951,
      "loss": 5.5031,
      "mean_token_accuracy": 0.1838056907057762,
      "num_tokens": 13487430.0,
      "step": 6040
    },
    {
      "entropy": 5.621393346786499,
      "epoch": 0.6789464817206716,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004963644678416763,
      "loss": 5.4868,
      "mean_token_accuracy": 0.1825977087020874,
      "num_tokens": 13499631.0,
      "step": 6045
    },
    {
      "entropy": 5.623709058761596,
      "epoch": 0.6795080586286292,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000496357276153527,
      "loss": 5.5812,
      "mean_token_accuracy": 0.18011384755373,
      "num_tokens": 13511753.0,
      "step": 6050
    },
    {
      "entropy": 5.617553615570069,
      "epoch": 0.6800696355365867,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004963500774172763,
      "loss": 5.64,
      "mean_token_accuracy": 0.1679525762796402,
      "num_tokens": 13524228.0,
      "step": 6055
    },
    {
      "entropy": 5.76211724281311,
      "epoch": 0.6806312124445443,
      "grad_norm": 0.984375,
      "learning_rate": 0.000496342871633153,
      "loss": 5.6445,
      "mean_token_accuracy": 0.17650653272867203,
      "num_tokens": 13535700.0,
      "step": 6060
    },
    {
      "entropy": 5.597730350494385,
      "epoch": 0.6811927893525018,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004963356588013871,
      "loss": 5.6393,
      "mean_token_accuracy": 0.1685441717505455,
      "num_tokens": 13547337.0,
      "step": 6065
    },
    {
      "entropy": 5.5392931461334225,
      "epoch": 0.6817543662604594,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004963284389222078,
      "loss": 5.4883,
      "mean_token_accuracy": 0.17627575844526291,
      "num_tokens": 13558350.0,
      "step": 6070
    },
    {
      "entropy": 5.623018074035644,
      "epoch": 0.6823159431684169,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004963212119958453,
      "loss": 5.5296,
      "mean_token_accuracy": 0.1789425104856491,
      "num_tokens": 13569230.0,
      "step": 6075
    },
    {
      "entropy": 5.668340969085693,
      "epoch": 0.6828775200763745,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963139780225295,
      "loss": 5.625,
      "mean_token_accuracy": 0.17555352449417114,
      "num_tokens": 13580502.0,
      "step": 6080
    },
    {
      "entropy": 5.657190752029419,
      "epoch": 0.683439096984332,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004963067370024909,
      "loss": 5.5146,
      "mean_token_accuracy": 0.18026141971349716,
      "num_tokens": 13591524.0,
      "step": 6085
    },
    {
      "entropy": 5.654711818695068,
      "epoch": 0.6840006738922896,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004962994889359598,
      "loss": 5.5477,
      "mean_token_accuracy": 0.1845819652080536,
      "num_tokens": 13602432.0,
      "step": 6090
    },
    {
      "entropy": 5.478832674026489,
      "epoch": 0.6845622508002471,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962922338231673,
      "loss": 5.4797,
      "mean_token_accuracy": 0.18590179234743118,
      "num_tokens": 13613052.0,
      "step": 6095
    },
    {
      "entropy": 5.582803630828858,
      "epoch": 0.6851238277082047,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004962849716643441,
      "loss": 5.5122,
      "mean_token_accuracy": 0.18142480701208114,
      "num_tokens": 13623961.0,
      "step": 6100
    },
    {
      "entropy": 5.534552574157715,
      "epoch": 0.6856854046161622,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004962777024597216,
      "loss": 5.4987,
      "mean_token_accuracy": 0.17624426931142806,
      "num_tokens": 13635285.0,
      "step": 6105
    },
    {
      "entropy": 5.707353353500366,
      "epoch": 0.6862469815241198,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004962704262095314,
      "loss": 5.6293,
      "mean_token_accuracy": 0.1722757786512375,
      "num_tokens": 13647720.0,
      "step": 6110
    },
    {
      "entropy": 5.5553606986999515,
      "epoch": 0.6868085584320772,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004962631429140048,
      "loss": 5.5002,
      "mean_token_accuracy": 0.1757523909211159,
      "num_tokens": 13659726.0,
      "step": 6115
    },
    {
      "entropy": 5.699248933792115,
      "epoch": 0.6873701353400348,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000496255852573374,
      "loss": 5.6857,
      "mean_token_accuracy": 0.1701590061187744,
      "num_tokens": 13670332.0,
      "step": 6120
    },
    {
      "entropy": 5.764449214935302,
      "epoch": 0.6879317122479923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004962485551878711,
      "loss": 5.6238,
      "mean_token_accuracy": 0.16969262212514877,
      "num_tokens": 13682419.0,
      "step": 6125
    },
    {
      "entropy": 5.463296699523926,
      "epoch": 0.6884932891559499,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004962412507577283,
      "loss": 5.3901,
      "mean_token_accuracy": 0.18372554928064347,
      "num_tokens": 13692698.0,
      "step": 6130
    },
    {
      "entropy": 5.524677658081055,
      "epoch": 0.6890548660639074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004962339392831783,
      "loss": 5.5756,
      "mean_token_accuracy": 0.1756581887602806,
      "num_tokens": 13704791.0,
      "step": 6135
    },
    {
      "entropy": 5.6459520816802975,
      "epoch": 0.689616442971865,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000496226620764454,
      "loss": 5.5087,
      "mean_token_accuracy": 0.18012100905179979,
      "num_tokens": 13715860.0,
      "step": 6140
    },
    {
      "entropy": 5.561775255203247,
      "epoch": 0.6901780198798225,
      "grad_norm": 1.171875,
      "learning_rate": 0.000496219295201788,
      "loss": 5.4738,
      "mean_token_accuracy": 0.18186445385217667,
      "num_tokens": 13727029.0,
      "step": 6145
    },
    {
      "entropy": 5.587733030319214,
      "epoch": 0.6907395967877801,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004962119625954141,
      "loss": 5.4703,
      "mean_token_accuracy": 0.18644033074378968,
      "num_tokens": 13737168.0,
      "step": 6150
    },
    {
      "entropy": 5.664116239547729,
      "epoch": 0.6913011736957376,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004962046229455653,
      "loss": 5.6141,
      "mean_token_accuracy": 0.1694582298398018,
      "num_tokens": 13749828.0,
      "step": 6155
    },
    {
      "entropy": 5.516778659820557,
      "epoch": 0.6918627506036952,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961972762524757,
      "loss": 5.4196,
      "mean_token_accuracy": 0.17955852299928665,
      "num_tokens": 13761617.0,
      "step": 6160
    },
    {
      "entropy": 5.585779285430908,
      "epoch": 0.6924243275116527,
      "grad_norm": 1.140625,
      "learning_rate": 0.000496189922516379,
      "loss": 5.4958,
      "mean_token_accuracy": 0.18332138508558274,
      "num_tokens": 13772721.0,
      "step": 6165
    },
    {
      "entropy": 5.533534955978394,
      "epoch": 0.6929859044196103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004961825617375093,
      "loss": 5.4756,
      "mean_token_accuracy": 0.176021970808506,
      "num_tokens": 13783215.0,
      "step": 6170
    },
    {
      "entropy": 5.544955825805664,
      "epoch": 0.6935474813275678,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004961751939161011,
      "loss": 5.441,
      "mean_token_accuracy": 0.18633371144533156,
      "num_tokens": 13793496.0,
      "step": 6175
    },
    {
      "entropy": 5.481276130676269,
      "epoch": 0.6941090582355254,
      "grad_norm": 1.109375,
      "learning_rate": 0.000496167819052389,
      "loss": 5.3649,
      "mean_token_accuracy": 0.18051891773939133,
      "num_tokens": 13802818.0,
      "step": 6180
    },
    {
      "entropy": 5.789962387084961,
      "epoch": 0.6946706351434829,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004961604371466077,
      "loss": 5.6746,
      "mean_token_accuracy": 0.17664305120706558,
      "num_tokens": 13814449.0,
      "step": 6185
    },
    {
      "entropy": 5.5759687423706055,
      "epoch": 0.6952322120514405,
      "grad_norm": 1.125,
      "learning_rate": 0.0004961530481989923,
      "loss": 5.5967,
      "mean_token_accuracy": 0.1766406327486038,
      "num_tokens": 13825369.0,
      "step": 6190
    },
    {
      "entropy": 5.557200717926025,
      "epoch": 0.695793788959398,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004961456522097782,
      "loss": 5.4469,
      "mean_token_accuracy": 0.17568059712648393,
      "num_tokens": 13837992.0,
      "step": 6195
    },
    {
      "entropy": 5.634961271286011,
      "epoch": 0.6963553658673556,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004961382491792006,
      "loss": 5.5396,
      "mean_token_accuracy": 0.17198241651058196,
      "num_tokens": 13849771.0,
      "step": 6200
    },
    {
      "entropy": 5.532082462310791,
      "epoch": 0.6969169427753131,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961308391074953,
      "loss": 5.5302,
      "mean_token_accuracy": 0.17718511074781418,
      "num_tokens": 13861997.0,
      "step": 6205
    },
    {
      "entropy": 5.694039583206177,
      "epoch": 0.6974785196832706,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004961234219948986,
      "loss": 5.6218,
      "mean_token_accuracy": 0.17499276846647263,
      "num_tokens": 13872972.0,
      "step": 6210
    },
    {
      "entropy": 5.499757194519043,
      "epoch": 0.6980400965912281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004961159978416461,
      "loss": 5.4024,
      "mean_token_accuracy": 0.18218694031238555,
      "num_tokens": 13883139.0,
      "step": 6215
    },
    {
      "entropy": 5.5940985679626465,
      "epoch": 0.6986016734991857,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004961085666479746,
      "loss": 5.5481,
      "mean_token_accuracy": 0.1795855402946472,
      "num_tokens": 13894326.0,
      "step": 6220
    },
    {
      "entropy": 5.626336193084716,
      "epoch": 0.6991632504071432,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004961011284141206,
      "loss": 5.5909,
      "mean_token_accuracy": 0.18014867156744002,
      "num_tokens": 13905525.0,
      "step": 6225
    },
    {
      "entropy": 5.6208432674407955,
      "epoch": 0.6997248273151008,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960936831403209,
      "loss": 5.5119,
      "mean_token_accuracy": 0.17700459212064742,
      "num_tokens": 13915973.0,
      "step": 6230
    },
    {
      "entropy": 5.545084619522095,
      "epoch": 0.7002864042230583,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004960862308268126,
      "loss": 5.346,
      "mean_token_accuracy": 0.18674954921007156,
      "num_tokens": 13926380.0,
      "step": 6235
    },
    {
      "entropy": 5.517192697525024,
      "epoch": 0.7008479811310159,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004960787714738329,
      "loss": 5.5212,
      "mean_token_accuracy": 0.1762975573539734,
      "num_tokens": 13938523.0,
      "step": 6240
    },
    {
      "entropy": 5.5097815990448,
      "epoch": 0.7014095580389734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960713050816194,
      "loss": 5.5021,
      "mean_token_accuracy": 0.18251630365848542,
      "num_tokens": 13949343.0,
      "step": 6245
    },
    {
      "entropy": 5.668053913116455,
      "epoch": 0.701971134946931,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960638316504098,
      "loss": 5.624,
      "mean_token_accuracy": 0.1783301293849945,
      "num_tokens": 13960056.0,
      "step": 6250
    },
    {
      "entropy": 5.714975786209107,
      "epoch": 0.7025327118548885,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960563511804421,
      "loss": 5.5877,
      "mean_token_accuracy": 0.17589129954576493,
      "num_tokens": 13971932.0,
      "step": 6255
    },
    {
      "entropy": 5.654637813568115,
      "epoch": 0.7030942887628461,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004960488636719545,
      "loss": 5.5849,
      "mean_token_accuracy": 0.17803931087255478,
      "num_tokens": 13981992.0,
      "step": 6260
    },
    {
      "entropy": 5.609004116058349,
      "epoch": 0.7036558656708036,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960413691251852,
      "loss": 5.5318,
      "mean_token_accuracy": 0.18366885632276536,
      "num_tokens": 13992215.0,
      "step": 6265
    },
    {
      "entropy": 5.595944166183472,
      "epoch": 0.7042174425787612,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004960338675403731,
      "loss": 5.446,
      "mean_token_accuracy": 0.19255190044641496,
      "num_tokens": 14002689.0,
      "step": 6270
    },
    {
      "entropy": 5.576307392120361,
      "epoch": 0.7047790194867187,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004960263589177568,
      "loss": 5.5593,
      "mean_token_accuracy": 0.17833008915185927,
      "num_tokens": 14014951.0,
      "step": 6275
    },
    {
      "entropy": 5.550465822219849,
      "epoch": 0.7053405963946763,
      "grad_norm": 1.0,
      "learning_rate": 0.0004960188432575757,
      "loss": 5.495,
      "mean_token_accuracy": 0.18250441998243333,
      "num_tokens": 14025502.0,
      "step": 6280
    },
    {
      "entropy": 5.7021482467651365,
      "epoch": 0.7059021733026338,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004960113205600688,
      "loss": 5.5438,
      "mean_token_accuracy": 0.17949292212724685,
      "num_tokens": 14036576.0,
      "step": 6285
    },
    {
      "entropy": 5.588206577301025,
      "epoch": 0.7064637502105914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004960037908254757,
      "loss": 5.5664,
      "mean_token_accuracy": 0.1804117351770401,
      "num_tokens": 14046493.0,
      "step": 6290
    },
    {
      "entropy": 5.672357654571533,
      "epoch": 0.7070253271185489,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004959962540540361,
      "loss": 5.6129,
      "mean_token_accuracy": 0.1806743010878563,
      "num_tokens": 14057297.0,
      "step": 6295
    },
    {
      "entropy": 5.5541548252105715,
      "epoch": 0.7075869040265065,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004959887102459903,
      "loss": 5.4067,
      "mean_token_accuracy": 0.18236117959022521,
      "num_tokens": 14068947.0,
      "step": 6300
    },
    {
      "entropy": 5.5343427658081055,
      "epoch": 0.7081484809344639,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004959811594015781,
      "loss": 5.4948,
      "mean_token_accuracy": 0.18413618057966233,
      "num_tokens": 14078979.0,
      "step": 6305
    },
    {
      "entropy": 5.526570081710815,
      "epoch": 0.7087100578424215,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004959736015210401,
      "loss": 5.4343,
      "mean_token_accuracy": 0.1832636833190918,
      "num_tokens": 14089785.0,
      "step": 6310
    },
    {
      "entropy": 5.597959661483765,
      "epoch": 0.709271634750379,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004959660366046169,
      "loss": 5.5243,
      "mean_token_accuracy": 0.17803296446800232,
      "num_tokens": 14099921.0,
      "step": 6315
    },
    {
      "entropy": 5.559597492218018,
      "epoch": 0.7098332116583366,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004959584646525494,
      "loss": 5.4335,
      "mean_token_accuracy": 0.17865606248378754,
      "num_tokens": 14110965.0,
      "step": 6320
    },
    {
      "entropy": 5.593908452987671,
      "epoch": 0.7103947885662941,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004959508856650788,
      "loss": 5.5665,
      "mean_token_accuracy": 0.17921946495771407,
      "num_tokens": 14122409.0,
      "step": 6325
    },
    {
      "entropy": 5.587578153610229,
      "epoch": 0.7109563654742517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004959432996424463,
      "loss": 5.5192,
      "mean_token_accuracy": 0.179829241335392,
      "num_tokens": 14134017.0,
      "step": 6330
    },
    {
      "entropy": 5.620336627960205,
      "epoch": 0.7115179423822092,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004959357065848932,
      "loss": 5.6225,
      "mean_token_accuracy": 0.1757448822259903,
      "num_tokens": 14144843.0,
      "step": 6335
    },
    {
      "entropy": 5.598922157287598,
      "epoch": 0.7120795192901668,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004959281064926618,
      "loss": 5.4881,
      "mean_token_accuracy": 0.18098643571138381,
      "num_tokens": 14155944.0,
      "step": 6340
    },
    {
      "entropy": 5.624665069580078,
      "epoch": 0.7126410961981243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004959204993659936,
      "loss": 5.5493,
      "mean_token_accuracy": 0.18257707506418228,
      "num_tokens": 14167064.0,
      "step": 6345
    },
    {
      "entropy": 5.655816173553466,
      "epoch": 0.7132026731060819,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004959128852051311,
      "loss": 5.5536,
      "mean_token_accuracy": 0.17361179739236832,
      "num_tokens": 14178583.0,
      "step": 6350
    },
    {
      "entropy": 5.550536394119263,
      "epoch": 0.7137642500140394,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004959052640103167,
      "loss": 5.4539,
      "mean_token_accuracy": 0.18530286997556686,
      "num_tokens": 14188752.0,
      "step": 6355
    },
    {
      "entropy": 5.566188669204712,
      "epoch": 0.714325826921997,
      "grad_norm": 1.09375,
      "learning_rate": 0.000495897635781793,
      "loss": 5.4407,
      "mean_token_accuracy": 0.1849163830280304,
      "num_tokens": 14199962.0,
      "step": 6360
    },
    {
      "entropy": 5.567184162139893,
      "epoch": 0.7148874038299545,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958900005198029,
      "loss": 5.4662,
      "mean_token_accuracy": 0.18421811908483504,
      "num_tokens": 14209877.0,
      "step": 6365
    },
    {
      "entropy": 5.521599388122558,
      "epoch": 0.7154489807379121,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004958823582245894,
      "loss": 5.5177,
      "mean_token_accuracy": 0.1838495761156082,
      "num_tokens": 14220672.0,
      "step": 6370
    },
    {
      "entropy": 5.705976486206055,
      "epoch": 0.7160105576458696,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004958747088963961,
      "loss": 5.5032,
      "mean_token_accuracy": 0.182656130194664,
      "num_tokens": 14232348.0,
      "step": 6375
    },
    {
      "entropy": 5.54701943397522,
      "epoch": 0.7165721345538272,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004958670525354663,
      "loss": 5.5191,
      "mean_token_accuracy": 0.17342815399169922,
      "num_tokens": 14243985.0,
      "step": 6380
    },
    {
      "entropy": 5.546517705917358,
      "epoch": 0.7171337114617847,
      "grad_norm": 0.984375,
      "learning_rate": 0.000495859389142044,
      "loss": 5.3952,
      "mean_token_accuracy": 0.18906868547201156,
      "num_tokens": 14253548.0,
      "step": 6385
    },
    {
      "entropy": 5.634414625167847,
      "epoch": 0.7176952883697423,
      "grad_norm": 1.15625,
      "learning_rate": 0.000495851718716373,
      "loss": 5.533,
      "mean_token_accuracy": 0.17814864218235016,
      "num_tokens": 14264757.0,
      "step": 6390
    },
    {
      "entropy": 5.473217868804932,
      "epoch": 0.7182568652776998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004958440412586977,
      "loss": 5.4523,
      "mean_token_accuracy": 0.1940762296319008,
      "num_tokens": 14275946.0,
      "step": 6395
    },
    {
      "entropy": 5.557712507247925,
      "epoch": 0.7188184421856574,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004958363567692624,
      "loss": 5.5314,
      "mean_token_accuracy": 0.17504859268665313,
      "num_tokens": 14287006.0,
      "step": 6400
    },
    {
      "entropy": 5.728769779205322,
      "epoch": 0.7193800190936148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004958286652483118,
      "loss": 5.5642,
      "mean_token_accuracy": 0.17381344437599183,
      "num_tokens": 14299328.0,
      "step": 6405
    },
    {
      "entropy": 5.6041662216186525,
      "epoch": 0.7199415960015724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004958209666960909,
      "loss": 5.5416,
      "mean_token_accuracy": 0.17951284497976303,
      "num_tokens": 14310048.0,
      "step": 6410
    },
    {
      "entropy": 5.541184997558593,
      "epoch": 0.7205031729095299,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004958132611128448,
      "loss": 5.4849,
      "mean_token_accuracy": 0.183475524187088,
      "num_tokens": 14321107.0,
      "step": 6415
    },
    {
      "entropy": 5.559459114074707,
      "epoch": 0.7210647498174875,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004958055484988188,
      "loss": 5.3971,
      "mean_token_accuracy": 0.18156871497631072,
      "num_tokens": 14332561.0,
      "step": 6420
    },
    {
      "entropy": 5.655962753295898,
      "epoch": 0.721626326725445,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004957978288542585,
      "loss": 5.5915,
      "mean_token_accuracy": 0.17599238604307174,
      "num_tokens": 14343238.0,
      "step": 6425
    },
    {
      "entropy": 5.583608913421631,
      "epoch": 0.7221879036334026,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004957901021794098,
      "loss": 5.4916,
      "mean_token_accuracy": 0.18004190325737,
      "num_tokens": 14356067.0,
      "step": 6430
    },
    {
      "entropy": 5.55387544631958,
      "epoch": 0.7227494805413601,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004957823684745184,
      "loss": 5.5079,
      "mean_token_accuracy": 0.1867664486169815,
      "num_tokens": 14367302.0,
      "step": 6435
    },
    {
      "entropy": 5.554821157455445,
      "epoch": 0.7233110574493177,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957746277398309,
      "loss": 5.4277,
      "mean_token_accuracy": 0.18782316148281097,
      "num_tokens": 14377535.0,
      "step": 6440
    },
    {
      "entropy": 5.655693054199219,
      "epoch": 0.7238726343572752,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004957668799755937,
      "loss": 5.5869,
      "mean_token_accuracy": 0.1728835567831993,
      "num_tokens": 14389493.0,
      "step": 6445
    },
    {
      "entropy": 5.57498197555542,
      "epoch": 0.7244342112652328,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004957591251820532,
      "loss": 5.4881,
      "mean_token_accuracy": 0.18310966044664384,
      "num_tokens": 14401564.0,
      "step": 6450
    },
    {
      "entropy": 5.606549978256226,
      "epoch": 0.7249957881731903,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004957513633594567,
      "loss": 5.5442,
      "mean_token_accuracy": 0.182178358733654,
      "num_tokens": 14412123.0,
      "step": 6455
    },
    {
      "entropy": 5.592750215530396,
      "epoch": 0.7255573650811479,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004957435945080512,
      "loss": 5.5364,
      "mean_token_accuracy": 0.18023575097322464,
      "num_tokens": 14423614.0,
      "step": 6460
    },
    {
      "entropy": 5.611506032943725,
      "epoch": 0.7261189419891054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004957358186280839,
      "loss": 5.5184,
      "mean_token_accuracy": 0.17583573907613753,
      "num_tokens": 14434830.0,
      "step": 6465
    },
    {
      "entropy": 5.586643171310425,
      "epoch": 0.726680518897063,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004957280357198026,
      "loss": 5.5673,
      "mean_token_accuracy": 0.18017499148845673,
      "num_tokens": 14445015.0,
      "step": 6470
    },
    {
      "entropy": 5.641062784194946,
      "epoch": 0.7272420958050205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004957202457834551,
      "loss": 5.6303,
      "mean_token_accuracy": 0.17067774534225463,
      "num_tokens": 14455809.0,
      "step": 6475
    },
    {
      "entropy": 5.593528985977173,
      "epoch": 0.7278036727129781,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004957124488192893,
      "loss": 5.4163,
      "mean_token_accuracy": 0.1925433471798897,
      "num_tokens": 14466565.0,
      "step": 6480
    },
    {
      "entropy": 5.601392316818237,
      "epoch": 0.7283652496209356,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004957046448275535,
      "loss": 5.5142,
      "mean_token_accuracy": 0.18903954476118087,
      "num_tokens": 14477383.0,
      "step": 6485
    },
    {
      "entropy": 5.575949573516846,
      "epoch": 0.7289268265288932,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004956968338084961,
      "loss": 5.5132,
      "mean_token_accuracy": 0.17813279926776887,
      "num_tokens": 14489968.0,
      "step": 6490
    },
    {
      "entropy": 5.615130090713501,
      "epoch": 0.7294884034368507,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956890157623661,
      "loss": 5.5636,
      "mean_token_accuracy": 0.17915889322757722,
      "num_tokens": 14500808.0,
      "step": 6495
    },
    {
      "entropy": 5.597889089584351,
      "epoch": 0.7300499803448082,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004956811906894122,
      "loss": 5.5657,
      "mean_token_accuracy": 0.17833685725927353,
      "num_tokens": 14511444.0,
      "step": 6500
    },
    {
      "entropy": 5.571704912185669,
      "epoch": 0.7306115572527657,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004956733585898836,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18077786415815353,
      "num_tokens": 14522562.0,
      "step": 6505
    },
    {
      "entropy": 5.524119329452515,
      "epoch": 0.7311731341607233,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004956655194640297,
      "loss": 5.5188,
      "mean_token_accuracy": 0.18095792979001998,
      "num_tokens": 14533497.0,
      "step": 6510
    },
    {
      "entropy": 5.582210826873779,
      "epoch": 0.7317347110686808,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004956576733121,
      "loss": 5.5322,
      "mean_token_accuracy": 0.17469650954008104,
      "num_tokens": 14545637.0,
      "step": 6515
    },
    {
      "entropy": 5.715583658218383,
      "epoch": 0.7322962879766384,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004956498201343444,
      "loss": 5.5402,
      "mean_token_accuracy": 0.1738186165690422,
      "num_tokens": 14556826.0,
      "step": 6520
    },
    {
      "entropy": 5.492493724822998,
      "epoch": 0.7328578648845959,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000495641959931013,
      "loss": 5.4765,
      "mean_token_accuracy": 0.1791041299700737,
      "num_tokens": 14567315.0,
      "step": 6525
    },
    {
      "entropy": 5.594206142425537,
      "epoch": 0.7334194417925535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004956340927023561,
      "loss": 5.5462,
      "mean_token_accuracy": 0.17821876108646392,
      "num_tokens": 14577685.0,
      "step": 6530
    },
    {
      "entropy": 5.595286321640015,
      "epoch": 0.733981018700511,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000495626218448624,
      "loss": 5.5066,
      "mean_token_accuracy": 0.18141655027866363,
      "num_tokens": 14588885.0,
      "step": 6535
    },
    {
      "entropy": 5.599929857254028,
      "epoch": 0.7345425956084686,
      "grad_norm": 1.125,
      "learning_rate": 0.0004956183371700676,
      "loss": 5.5172,
      "mean_token_accuracy": 0.18136245310306548,
      "num_tokens": 14599774.0,
      "step": 6540
    },
    {
      "entropy": 5.6762114524841305,
      "epoch": 0.7351041725164261,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004956104488669377,
      "loss": 5.5702,
      "mean_token_accuracy": 0.17265312671661376,
      "num_tokens": 14612101.0,
      "step": 6545
    },
    {
      "entropy": 5.664939451217651,
      "epoch": 0.7356657494243837,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004956025535394856,
      "loss": 5.6306,
      "mean_token_accuracy": 0.16337666809558868,
      "num_tokens": 14622151.0,
      "step": 6550
    },
    {
      "entropy": 5.614804124832153,
      "epoch": 0.7362273263323412,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955946511879626,
      "loss": 5.5305,
      "mean_token_accuracy": 0.17403078824281693,
      "num_tokens": 14634631.0,
      "step": 6555
    },
    {
      "entropy": 5.541925621032715,
      "epoch": 0.7367889032402988,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955867418126205,
      "loss": 5.4724,
      "mean_token_accuracy": 0.18178219348192215,
      "num_tokens": 14645429.0,
      "step": 6560
    },
    {
      "entropy": 5.543210077285766,
      "epoch": 0.7373504801482563,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004955788254137108,
      "loss": 5.484,
      "mean_token_accuracy": 0.1796364888548851,
      "num_tokens": 14657309.0,
      "step": 6565
    },
    {
      "entropy": 5.5579516887664795,
      "epoch": 0.7379120570562139,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004955709019914857,
      "loss": 5.4095,
      "mean_token_accuracy": 0.18526892811059953,
      "num_tokens": 14668578.0,
      "step": 6570
    },
    {
      "entropy": 5.614055633544922,
      "epoch": 0.7384736339641714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004955629715461977,
      "loss": 5.6268,
      "mean_token_accuracy": 0.1746178910136223,
      "num_tokens": 14679120.0,
      "step": 6575
    },
    {
      "entropy": 5.692960786819458,
      "epoch": 0.739035210872129,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004955550340780991,
      "loss": 5.5816,
      "mean_token_accuracy": 0.17088816463947296,
      "num_tokens": 14690849.0,
      "step": 6580
    },
    {
      "entropy": 5.616451787948608,
      "epoch": 0.7395967877800865,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004955470895874427,
      "loss": 5.5417,
      "mean_token_accuracy": 0.18024831265211105,
      "num_tokens": 14702431.0,
      "step": 6585
    },
    {
      "entropy": 5.542486524581909,
      "epoch": 0.7401583646880441,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955391380744813,
      "loss": 5.4475,
      "mean_token_accuracy": 0.1850384771823883,
      "num_tokens": 14714271.0,
      "step": 6590
    },
    {
      "entropy": 5.574007987976074,
      "epoch": 0.7407199415960015,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004955311795394684,
      "loss": 5.5864,
      "mean_token_accuracy": 0.1800689071416855,
      "num_tokens": 14724796.0,
      "step": 6595
    },
    {
      "entropy": 5.607455253601074,
      "epoch": 0.7412815185039591,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495523213982657,
      "loss": 5.5087,
      "mean_token_accuracy": 0.17894099205732344,
      "num_tokens": 14735331.0,
      "step": 6600
    },
    {
      "entropy": 5.659684324264527,
      "epoch": 0.7418430954119166,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004955152414043011,
      "loss": 5.5451,
      "mean_token_accuracy": 0.181105837225914,
      "num_tokens": 14748595.0,
      "step": 6605
    },
    {
      "entropy": 5.619999551773072,
      "epoch": 0.7424046723198742,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004955072618046543,
      "loss": 5.4903,
      "mean_token_accuracy": 0.1820297509431839,
      "num_tokens": 14759501.0,
      "step": 6610
    },
    {
      "entropy": 5.541188383102417,
      "epoch": 0.7429662492278317,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004954992751839708,
      "loss": 5.5354,
      "mean_token_accuracy": 0.1811933383345604,
      "num_tokens": 14769752.0,
      "step": 6615
    },
    {
      "entropy": 5.606895542144775,
      "epoch": 0.7435278261357893,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004954912815425048,
      "loss": 5.4838,
      "mean_token_accuracy": 0.17985673099756241,
      "num_tokens": 14781166.0,
      "step": 6620
    },
    {
      "entropy": 5.516903352737427,
      "epoch": 0.7440894030437468,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000495483280880511,
      "loss": 5.4006,
      "mean_token_accuracy": 0.18454970866441728,
      "num_tokens": 14790536.0,
      "step": 6625
    },
    {
      "entropy": 5.665596151351929,
      "epoch": 0.7446509799517044,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000495475273198244,
      "loss": 5.6088,
      "mean_token_accuracy": 0.17387449145317077,
      "num_tokens": 14801673.0,
      "step": 6630
    },
    {
      "entropy": 5.586531305313111,
      "epoch": 0.7452125568596619,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004954672584959588,
      "loss": 5.4956,
      "mean_token_accuracy": 0.18119197189807892,
      "num_tokens": 14812345.0,
      "step": 6635
    },
    {
      "entropy": 5.6696093559265135,
      "epoch": 0.7457741337676195,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004954592367739104,
      "loss": 5.6632,
      "mean_token_accuracy": 0.17205969691276551,
      "num_tokens": 14823620.0,
      "step": 6640
    },
    {
      "entropy": 5.647694873809814,
      "epoch": 0.746335710675577,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004954512080323546,
      "loss": 5.4731,
      "mean_token_accuracy": 0.18358333855867387,
      "num_tokens": 14833487.0,
      "step": 6645
    },
    {
      "entropy": 5.41290020942688,
      "epoch": 0.7468972875835346,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004954431722715468,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18748193979263306,
      "num_tokens": 14844421.0,
      "step": 6650
    },
    {
      "entropy": 5.454937219619751,
      "epoch": 0.7474588644914921,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004954351294917428,
      "loss": 5.5029,
      "mean_token_accuracy": 0.1878804534673691,
      "num_tokens": 14855525.0,
      "step": 6655
    },
    {
      "entropy": 5.714896631240845,
      "epoch": 0.7480204413994497,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004954270796931989,
      "loss": 5.5735,
      "mean_token_accuracy": 0.17403673976659775,
      "num_tokens": 14866717.0,
      "step": 6660
    },
    {
      "entropy": 5.518547773361206,
      "epoch": 0.7485820183074072,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004954190228761712,
      "loss": 5.4693,
      "mean_token_accuracy": 0.1776670604944229,
      "num_tokens": 14877372.0,
      "step": 6665
    },
    {
      "entropy": 5.560596323013305,
      "epoch": 0.7491435952153648,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004954109590409164,
      "loss": 5.5229,
      "mean_token_accuracy": 0.18132778257131577,
      "num_tokens": 14888618.0,
      "step": 6670
    },
    {
      "entropy": 5.559640502929687,
      "epoch": 0.7497051721233223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004954028881876911,
      "loss": 5.5302,
      "mean_token_accuracy": 0.18338897079229355,
      "num_tokens": 14900421.0,
      "step": 6675
    },
    {
      "entropy": 5.6074700355529785,
      "epoch": 0.7502667490312799,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004953948103167524,
      "loss": 5.517,
      "mean_token_accuracy": 0.1782395139336586,
      "num_tokens": 14910634.0,
      "step": 6680
    },
    {
      "entropy": 5.709167766571045,
      "epoch": 0.7508283259392374,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004953867254283575,
      "loss": 5.5577,
      "mean_token_accuracy": 0.17521576583385468,
      "num_tokens": 14921408.0,
      "step": 6685
    },
    {
      "entropy": 5.575962209701538,
      "epoch": 0.7513899028471949,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004953786335227636,
      "loss": 5.5314,
      "mean_token_accuracy": 0.1847861796617508,
      "num_tokens": 14933435.0,
      "step": 6690
    },
    {
      "entropy": 5.556945562362671,
      "epoch": 0.7519514797551524,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953705346002287,
      "loss": 5.565,
      "mean_token_accuracy": 0.17538581043481827,
      "num_tokens": 14946488.0,
      "step": 6695
    },
    {
      "entropy": 5.583792495727539,
      "epoch": 0.75251305666311,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004953624286610106,
      "loss": 5.4344,
      "mean_token_accuracy": 0.1829975426197052,
      "num_tokens": 14958053.0,
      "step": 6700
    },
    {
      "entropy": 5.611514616012573,
      "epoch": 0.7530746335710675,
      "grad_norm": 1.0,
      "learning_rate": 0.0004953543157053672,
      "loss": 5.503,
      "mean_token_accuracy": 0.17879436612129213,
      "num_tokens": 14968922.0,
      "step": 6705
    },
    {
      "entropy": 5.613012504577637,
      "epoch": 0.7536362104790251,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004953461957335569,
      "loss": 5.5153,
      "mean_token_accuracy": 0.18105665296316148,
      "num_tokens": 14979541.0,
      "step": 6710
    },
    {
      "entropy": 5.573808383941651,
      "epoch": 0.7541977873869826,
      "grad_norm": 1.125,
      "learning_rate": 0.0004953380687458382,
      "loss": 5.545,
      "mean_token_accuracy": 0.1735311210155487,
      "num_tokens": 14990009.0,
      "step": 6715
    },
    {
      "entropy": 5.576241159439087,
      "epoch": 0.7547593642949402,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004953299347424701,
      "loss": 5.5385,
      "mean_token_accuracy": 0.17533133774995804,
      "num_tokens": 15001712.0,
      "step": 6720
    },
    {
      "entropy": 5.679225587844849,
      "epoch": 0.7553209412028977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004953217937237114,
      "loss": 5.4814,
      "mean_token_accuracy": 0.18626639991998672,
      "num_tokens": 15012574.0,
      "step": 6725
    },
    {
      "entropy": 5.482853269577026,
      "epoch": 0.7558825181108553,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004953136456898212,
      "loss": 5.4957,
      "mean_token_accuracy": 0.18412776589393615,
      "num_tokens": 15023034.0,
      "step": 6730
    },
    {
      "entropy": 5.637028312683105,
      "epoch": 0.7564440950188128,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004953054906410593,
      "loss": 5.5307,
      "mean_token_accuracy": 0.1769053280353546,
      "num_tokens": 15033925.0,
      "step": 6735
    },
    {
      "entropy": 5.604602479934693,
      "epoch": 0.7570056719267704,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004952973285776848,
      "loss": 5.5212,
      "mean_token_accuracy": 0.18264247179031373,
      "num_tokens": 15044947.0,
      "step": 6740
    },
    {
      "entropy": 5.6245598793029785,
      "epoch": 0.7575672488347279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004952891594999582,
      "loss": 5.4278,
      "mean_token_accuracy": 0.18887935429811478,
      "num_tokens": 15054784.0,
      "step": 6745
    },
    {
      "entropy": 5.507697010040284,
      "epoch": 0.7581288257426855,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952809834081392,
      "loss": 5.5199,
      "mean_token_accuracy": 0.1743734985589981,
      "num_tokens": 15066272.0,
      "step": 6750
    },
    {
      "entropy": 5.631846618652344,
      "epoch": 0.758690402650643,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004952728003024883,
      "loss": 5.5649,
      "mean_token_accuracy": 0.17555858045816422,
      "num_tokens": 15079504.0,
      "step": 6755
    },
    {
      "entropy": 5.620772886276245,
      "epoch": 0.7592519795586006,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952646101832659,
      "loss": 5.5301,
      "mean_token_accuracy": 0.1824829548597336,
      "num_tokens": 15089623.0,
      "step": 6760
    },
    {
      "entropy": 5.506829786300659,
      "epoch": 0.7598135564665581,
      "grad_norm": 1.125,
      "learning_rate": 0.0004952564130507331,
      "loss": 5.4287,
      "mean_token_accuracy": 0.17793408483266832,
      "num_tokens": 15100816.0,
      "step": 6765
    },
    {
      "entropy": 5.564327001571655,
      "epoch": 0.7603751333745157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004952482089051506,
      "loss": 5.5414,
      "mean_token_accuracy": 0.18410272747278214,
      "num_tokens": 15111697.0,
      "step": 6770
    },
    {
      "entropy": 5.68920578956604,
      "epoch": 0.7609367102824732,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004952399977467796,
      "loss": 5.5962,
      "mean_token_accuracy": 0.1749215081334114,
      "num_tokens": 15122869.0,
      "step": 6775
    },
    {
      "entropy": 5.5698570728302,
      "epoch": 0.7614982871904308,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004952317795758817,
      "loss": 5.4012,
      "mean_token_accuracy": 0.18042394667863845,
      "num_tokens": 15134709.0,
      "step": 6780
    },
    {
      "entropy": 5.5207994937896725,
      "epoch": 0.7620598640983883,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004952235543927185,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18242030143737792,
      "num_tokens": 15145319.0,
      "step": 6785
    },
    {
      "entropy": 5.504973936080932,
      "epoch": 0.7626214410063458,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000495215322197552,
      "loss": 5.4643,
      "mean_token_accuracy": 0.17559018433094026,
      "num_tokens": 15156463.0,
      "step": 6790
    },
    {
      "entropy": 5.661270618438721,
      "epoch": 0.7631830179143033,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004952070829906442,
      "loss": 5.6021,
      "mean_token_accuracy": 0.1682196795940399,
      "num_tokens": 15167618.0,
      "step": 6795
    },
    {
      "entropy": 5.6250732421875,
      "epoch": 0.7637445948222609,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004951988367722574,
      "loss": 5.539,
      "mean_token_accuracy": 0.1833435922861099,
      "num_tokens": 15179312.0,
      "step": 6800
    },
    {
      "entropy": 5.5221010684967045,
      "epoch": 0.7643061717302184,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004951905835426544,
      "loss": 5.4799,
      "mean_token_accuracy": 0.1866574168205261,
      "num_tokens": 15191370.0,
      "step": 6805
    },
    {
      "entropy": 5.628694105148315,
      "epoch": 0.764867748638176,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004951823233020976,
      "loss": 5.4319,
      "mean_token_accuracy": 0.1868742287158966,
      "num_tokens": 15201060.0,
      "step": 6810
    },
    {
      "entropy": 5.536026668548584,
      "epoch": 0.7654293255461335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004951740560508504,
      "loss": 5.3741,
      "mean_token_accuracy": 0.18512472361326218,
      "num_tokens": 15212186.0,
      "step": 6815
    },
    {
      "entropy": 5.498651123046875,
      "epoch": 0.7659909024540911,
      "grad_norm": 1.25,
      "learning_rate": 0.0004951657817891758,
      "loss": 5.3894,
      "mean_token_accuracy": 0.189059054851532,
      "num_tokens": 15223301.0,
      "step": 6820
    },
    {
      "entropy": 5.561225461959839,
      "epoch": 0.7665524793620486,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951575005173373,
      "loss": 5.5423,
      "mean_token_accuracy": 0.18042814135551452,
      "num_tokens": 15234475.0,
      "step": 6825
    },
    {
      "entropy": 5.6078619956970215,
      "epoch": 0.7671140562700062,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004951492122355986,
      "loss": 5.5286,
      "mean_token_accuracy": 0.17910326421260833,
      "num_tokens": 15245877.0,
      "step": 6830
    },
    {
      "entropy": 5.610657024383545,
      "epoch": 0.7676756331779637,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004951409169442236,
      "loss": 5.5098,
      "mean_token_accuracy": 0.1830678626894951,
      "num_tokens": 15257375.0,
      "step": 6835
    },
    {
      "entropy": 5.592937183380127,
      "epoch": 0.7682372100859213,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004951326146434765,
      "loss": 5.5291,
      "mean_token_accuracy": 0.18117677718400954,
      "num_tokens": 15268079.0,
      "step": 6840
    },
    {
      "entropy": 5.53626708984375,
      "epoch": 0.7687987869938788,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004951243053336214,
      "loss": 5.5113,
      "mean_token_accuracy": 0.1808534398674965,
      "num_tokens": 15278678.0,
      "step": 6845
    },
    {
      "entropy": 5.673134231567383,
      "epoch": 0.7693603639018364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004951159890149232,
      "loss": 5.621,
      "mean_token_accuracy": 0.17812574207782744,
      "num_tokens": 15290080.0,
      "step": 6850
    },
    {
      "entropy": 5.642621374130249,
      "epoch": 0.7699219408097939,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004951076656876466,
      "loss": 5.5435,
      "mean_token_accuracy": 0.17185832262039186,
      "num_tokens": 15300602.0,
      "step": 6855
    },
    {
      "entropy": 5.628775787353516,
      "epoch": 0.7704835177177515,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004950993353520564,
      "loss": 5.5439,
      "mean_token_accuracy": 0.18035327643156052,
      "num_tokens": 15312823.0,
      "step": 6860
    },
    {
      "entropy": 5.615487957000733,
      "epoch": 0.771045094625709,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495090998008418,
      "loss": 5.5572,
      "mean_token_accuracy": 0.1737877056002617,
      "num_tokens": 15323870.0,
      "step": 6865
    },
    {
      "entropy": 5.540188217163086,
      "epoch": 0.7716066715336666,
      "grad_norm": 1.03125,
      "learning_rate": 0.000495082653656997,
      "loss": 5.4444,
      "mean_token_accuracy": 0.1869006186723709,
      "num_tokens": 15334887.0,
      "step": 6870
    },
    {
      "entropy": 5.6197319507598875,
      "epoch": 0.7721682484416241,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004950743022980587,
      "loss": 5.544,
      "mean_token_accuracy": 0.1764238730072975,
      "num_tokens": 15345871.0,
      "step": 6875
    },
    {
      "entropy": 5.478558778762817,
      "epoch": 0.7727298253495817,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004950659439318696,
      "loss": 5.4073,
      "mean_token_accuracy": 0.18785117268562318,
      "num_tokens": 15356530.0,
      "step": 6880
    },
    {
      "entropy": 5.608475160598755,
      "epoch": 0.7732914022575391,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004950575785586952,
      "loss": 5.4735,
      "mean_token_accuracy": 0.182441870868206,
      "num_tokens": 15368756.0,
      "step": 6885
    },
    {
      "entropy": 5.638404178619385,
      "epoch": 0.7738529791654967,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004950492061788023,
      "loss": 5.5485,
      "mean_token_accuracy": 0.1835839956998825,
      "num_tokens": 15379450.0,
      "step": 6890
    },
    {
      "entropy": 5.506632471084595,
      "epoch": 0.7744145560734542,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004950408267924573,
      "loss": 5.4717,
      "mean_token_accuracy": 0.16710675954818727,
      "num_tokens": 15389501.0,
      "step": 6895
    },
    {
      "entropy": 5.434316825866699,
      "epoch": 0.7749761329814118,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004950324403999271,
      "loss": 5.4733,
      "mean_token_accuracy": 0.18077814429998398,
      "num_tokens": 15400608.0,
      "step": 6900
    },
    {
      "entropy": 5.6977643966674805,
      "epoch": 0.7755377098893693,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004950240470014785,
      "loss": 5.5596,
      "mean_token_accuracy": 0.17342529743909835,
      "num_tokens": 15411709.0,
      "step": 6905
    },
    {
      "entropy": 5.548699808120728,
      "epoch": 0.7760992867973269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004950156465973791,
      "loss": 5.4182,
      "mean_token_accuracy": 0.18748781234025955,
      "num_tokens": 15422292.0,
      "step": 6910
    },
    {
      "entropy": 5.5851311683654785,
      "epoch": 0.7766608637052844,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000495007239187896,
      "loss": 5.5694,
      "mean_token_accuracy": 0.17563919872045516,
      "num_tokens": 15434439.0,
      "step": 6915
    },
    {
      "entropy": 5.590206480026245,
      "epoch": 0.777222440613242,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004949988247732972,
      "loss": 5.4984,
      "mean_token_accuracy": 0.17924798429012298,
      "num_tokens": 15445480.0,
      "step": 6920
    },
    {
      "entropy": 5.522445249557495,
      "epoch": 0.7777840175211995,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004949904033538504,
      "loss": 5.4748,
      "mean_token_accuracy": 0.1830499142408371,
      "num_tokens": 15456960.0,
      "step": 6925
    },
    {
      "entropy": 5.4802093505859375,
      "epoch": 0.7783455944291571,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000494981974929824,
      "loss": 5.4997,
      "mean_token_accuracy": 0.1799062415957451,
      "num_tokens": 15468169.0,
      "step": 6930
    },
    {
      "entropy": 5.649336147308349,
      "epoch": 0.7789071713371146,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494973539501486,
      "loss": 5.4657,
      "mean_token_accuracy": 0.18194984197616576,
      "num_tokens": 15478740.0,
      "step": 6935
    },
    {
      "entropy": 5.588720083236694,
      "epoch": 0.7794687482450722,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004949650970691052,
      "loss": 5.5774,
      "mean_token_accuracy": 0.1715809226036072,
      "num_tokens": 15489705.0,
      "step": 6940
    },
    {
      "entropy": 5.5121870040893555,
      "epoch": 0.7800303251530297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004949566476329505,
      "loss": 5.4014,
      "mean_token_accuracy": 0.1816771224141121,
      "num_tokens": 15500549.0,
      "step": 6945
    },
    {
      "entropy": 5.713291931152344,
      "epoch": 0.7805919020609873,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004949481911932908,
      "loss": 5.5491,
      "mean_token_accuracy": 0.17817943543195724,
      "num_tokens": 15511124.0,
      "step": 6950
    },
    {
      "entropy": 5.515682363510132,
      "epoch": 0.7811534789689448,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004949397277503954,
      "loss": 5.448,
      "mean_token_accuracy": 0.18309102356433868,
      "num_tokens": 15521810.0,
      "step": 6955
    },
    {
      "entropy": 5.529624891281128,
      "epoch": 0.7817150558769024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004949312573045338,
      "loss": 5.4863,
      "mean_token_accuracy": 0.18091827929019927,
      "num_tokens": 15533108.0,
      "step": 6960
    },
    {
      "entropy": 5.608380746841431,
      "epoch": 0.7822766327848599,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004949227798559757,
      "loss": 5.5298,
      "mean_token_accuracy": 0.17911394834518432,
      "num_tokens": 15544841.0,
      "step": 6965
    },
    {
      "entropy": 5.551511478424072,
      "epoch": 0.7828382096928175,
      "grad_norm": 1.125,
      "learning_rate": 0.0004949142954049909,
      "loss": 5.4989,
      "mean_token_accuracy": 0.1772698223590851,
      "num_tokens": 15556297.0,
      "step": 6970
    },
    {
      "entropy": 5.561290454864502,
      "epoch": 0.783399786600775,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004949058039518496,
      "loss": 5.4019,
      "mean_token_accuracy": 0.18467747122049333,
      "num_tokens": 15566607.0,
      "step": 6975
    },
    {
      "entropy": 5.6568274974823,
      "epoch": 0.7839613635087325,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004948973054968222,
      "loss": 5.5661,
      "mean_token_accuracy": 0.1753258228302002,
      "num_tokens": 15577435.0,
      "step": 6980
    },
    {
      "entropy": 5.646487283706665,
      "epoch": 0.78452294041669,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004948888000401794,
      "loss": 5.4755,
      "mean_token_accuracy": 0.18313849419355394,
      "num_tokens": 15587844.0,
      "step": 6985
    },
    {
      "entropy": 5.599114179611206,
      "epoch": 0.7850845173246476,
      "grad_norm": 1.4375,
      "learning_rate": 0.0004948802875821918,
      "loss": 5.5405,
      "mean_token_accuracy": 0.17978910058736802,
      "num_tokens": 15599586.0,
      "step": 6990
    },
    {
      "entropy": 5.4770026206970215,
      "epoch": 0.7856460942326051,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004948717681231308,
      "loss": 5.4258,
      "mean_token_accuracy": 0.17992580831050872,
      "num_tokens": 15610838.0,
      "step": 6995
    },
    {
      "entropy": 5.607249164581299,
      "epoch": 0.7862076711405627,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004948632416632672,
      "loss": 5.512,
      "mean_token_accuracy": 0.18194066733121872,
      "num_tokens": 15622652.0,
      "step": 7000
    },
    {
      "entropy": 5.701489400863648,
      "epoch": 0.7867692480485202,
      "grad_norm": 1.125,
      "learning_rate": 0.0004948547082028728,
      "loss": 5.5976,
      "mean_token_accuracy": 0.17344842851161957,
      "num_tokens": 15634280.0,
      "step": 7005
    },
    {
      "entropy": 5.504639339447022,
      "epoch": 0.7873308249564778,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004948461677422193,
      "loss": 5.3308,
      "mean_token_accuracy": 0.19431402832269667,
      "num_tokens": 15644088.0,
      "step": 7010
    },
    {
      "entropy": 5.539066553115845,
      "epoch": 0.7878924018644353,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004948376202815785,
      "loss": 5.5837,
      "mean_token_accuracy": 0.1697140157222748,
      "num_tokens": 15654968.0,
      "step": 7015
    },
    {
      "entropy": 5.555192279815674,
      "epoch": 0.7884539787723929,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004948290658212225,
      "loss": 5.4843,
      "mean_token_accuracy": 0.18228119760751724,
      "num_tokens": 15666758.0,
      "step": 7020
    },
    {
      "entropy": 5.4680994033813475,
      "epoch": 0.7890155556803504,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948205043614239,
      "loss": 5.3367,
      "mean_token_accuracy": 0.19239940792322158,
      "num_tokens": 15677295.0,
      "step": 7025
    },
    {
      "entropy": 5.579523754119873,
      "epoch": 0.789577132588308,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004948119359024551,
      "loss": 5.5494,
      "mean_token_accuracy": 0.17890748232603074,
      "num_tokens": 15689508.0,
      "step": 7030
    },
    {
      "entropy": 5.60092945098877,
      "epoch": 0.7901387094962655,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004948033604445891,
      "loss": 5.4968,
      "mean_token_accuracy": 0.17887441962957382,
      "num_tokens": 15701289.0,
      "step": 7035
    },
    {
      "entropy": 5.519186878204346,
      "epoch": 0.7907002864042231,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004947947779880988,
      "loss": 5.4419,
      "mean_token_accuracy": 0.17764632552862167,
      "num_tokens": 15711579.0,
      "step": 7040
    },
    {
      "entropy": 5.604786109924317,
      "epoch": 0.7912618633121806,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004947861885332576,
      "loss": 5.5304,
      "mean_token_accuracy": 0.18105943948030473,
      "num_tokens": 15722629.0,
      "step": 7045
    },
    {
      "entropy": 5.5438587188720705,
      "epoch": 0.7918234402201382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004947775920803388,
      "loss": 5.458,
      "mean_token_accuracy": 0.17556296586990355,
      "num_tokens": 15735818.0,
      "step": 7050
    },
    {
      "entropy": 5.531337594985962,
      "epoch": 0.7923850171280957,
      "grad_norm": 1.125,
      "learning_rate": 0.0004947689886296162,
      "loss": 5.5279,
      "mean_token_accuracy": 0.1775309294462204,
      "num_tokens": 15746914.0,
      "step": 7055
    },
    {
      "entropy": 5.5272585391998295,
      "epoch": 0.7929465940360533,
      "grad_norm": 1.265625,
      "learning_rate": 0.000494760378181364,
      "loss": 5.4323,
      "mean_token_accuracy": 0.17647951543331147,
      "num_tokens": 15757597.0,
      "step": 7060
    },
    {
      "entropy": 5.517232227325439,
      "epoch": 0.7935081709440108,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004947517607358559,
      "loss": 5.3628,
      "mean_token_accuracy": 0.18752047717571257,
      "num_tokens": 15768860.0,
      "step": 7065
    },
    {
      "entropy": 5.55667552947998,
      "epoch": 0.7940697478519684,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004947431362933666,
      "loss": 5.4987,
      "mean_token_accuracy": 0.17675920724868774,
      "num_tokens": 15779476.0,
      "step": 7070
    },
    {
      "entropy": 5.558329248428345,
      "epoch": 0.7946313247599258,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004947345048541706,
      "loss": 5.5288,
      "mean_token_accuracy": 0.17560510635375975,
      "num_tokens": 15792142.0,
      "step": 7075
    },
    {
      "entropy": 5.556995153427124,
      "epoch": 0.7951929016678834,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004947258664185427,
      "loss": 5.5062,
      "mean_token_accuracy": 0.17924264967441558,
      "num_tokens": 15802307.0,
      "step": 7080
    },
    {
      "entropy": 5.542277002334595,
      "epoch": 0.7957544785758409,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004947172209867581,
      "loss": 5.4737,
      "mean_token_accuracy": 0.18031045496463777,
      "num_tokens": 15812986.0,
      "step": 7085
    },
    {
      "entropy": 5.540415239334107,
      "epoch": 0.7963160554837985,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004947085685590919,
      "loss": 5.4312,
      "mean_token_accuracy": 0.1854272723197937,
      "num_tokens": 15823938.0,
      "step": 7090
    },
    {
      "entropy": 5.503327417373657,
      "epoch": 0.796877632391756,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004946999091358196,
      "loss": 5.3951,
      "mean_token_accuracy": 0.18484974205493926,
      "num_tokens": 15835813.0,
      "step": 7095
    },
    {
      "entropy": 5.578824234008789,
      "epoch": 0.7974392092997136,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494691242717217,
      "loss": 5.4656,
      "mean_token_accuracy": 0.1857539013028145,
      "num_tokens": 15846203.0,
      "step": 7100
    },
    {
      "entropy": 5.5485265254974365,
      "epoch": 0.7980007862076711,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946825693035601,
      "loss": 5.4344,
      "mean_token_accuracy": 0.18856938183307648,
      "num_tokens": 15856919.0,
      "step": 7105
    },
    {
      "entropy": 5.515488862991333,
      "epoch": 0.7985623631156287,
      "grad_norm": 1.109375,
      "learning_rate": 0.000494673888895125,
      "loss": 5.5664,
      "mean_token_accuracy": 0.17663007527589797,
      "num_tokens": 15867149.0,
      "step": 7110
    },
    {
      "entropy": 5.4877420425415036,
      "epoch": 0.7991239400235862,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004946652014921881,
      "loss": 5.432,
      "mean_token_accuracy": 0.1861245095729828,
      "num_tokens": 15877229.0,
      "step": 7115
    },
    {
      "entropy": 5.591596460342407,
      "epoch": 0.7996855169315438,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946565070950259,
      "loss": 5.4269,
      "mean_token_accuracy": 0.17961910218000413,
      "num_tokens": 15888799.0,
      "step": 7120
    },
    {
      "entropy": 5.502090787887573,
      "epoch": 0.8002470938395013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946478057039153,
      "loss": 5.4127,
      "mean_token_accuracy": 0.19258841425180434,
      "num_tokens": 15900144.0,
      "step": 7125
    },
    {
      "entropy": 5.589698314666748,
      "epoch": 0.8008086707474589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946390973191335,
      "loss": 5.593,
      "mean_token_accuracy": 0.17292029708623885,
      "num_tokens": 15912030.0,
      "step": 7130
    },
    {
      "entropy": 5.584044122695923,
      "epoch": 0.8013702476554164,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946303819409575,
      "loss": 5.4584,
      "mean_token_accuracy": 0.18304266929626464,
      "num_tokens": 15923606.0,
      "step": 7135
    },
    {
      "entropy": 5.531179523468017,
      "epoch": 0.801931824563374,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004946216595696652,
      "loss": 5.4577,
      "mean_token_accuracy": 0.1842547670006752,
      "num_tokens": 15935077.0,
      "step": 7140
    },
    {
      "entropy": 5.466410350799561,
      "epoch": 0.8024934014713315,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004946129302055338,
      "loss": 5.3902,
      "mean_token_accuracy": 0.18604701906442642,
      "num_tokens": 15945331.0,
      "step": 7145
    },
    {
      "entropy": 5.560829782485962,
      "epoch": 0.8030549783792891,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004946041938488417,
      "loss": 5.4833,
      "mean_token_accuracy": 0.17986607700586318,
      "num_tokens": 15957031.0,
      "step": 7150
    },
    {
      "entropy": 5.587395668029785,
      "epoch": 0.8036165552872466,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004945954504998668,
      "loss": 5.5288,
      "mean_token_accuracy": 0.18224121928215026,
      "num_tokens": 15967545.0,
      "step": 7155
    },
    {
      "entropy": 5.656823635101318,
      "epoch": 0.8041781321952042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004945867001588877,
      "loss": 5.5712,
      "mean_token_accuracy": 0.16716524362564086,
      "num_tokens": 15978700.0,
      "step": 7160
    },
    {
      "entropy": 5.490016794204712,
      "epoch": 0.8047397091031617,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945779428261827,
      "loss": 5.4743,
      "mean_token_accuracy": 0.1853304013609886,
      "num_tokens": 15989712.0,
      "step": 7165
    },
    {
      "entropy": 5.690036058425903,
      "epoch": 0.8053012860111193,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000494569178502031,
      "loss": 5.6535,
      "mean_token_accuracy": 0.17617819905281068,
      "num_tokens": 16002380.0,
      "step": 7170
    },
    {
      "entropy": 5.655481195449829,
      "epoch": 0.8058628629190767,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004945604071867114,
      "loss": 5.4413,
      "mean_token_accuracy": 0.1773824855685234,
      "num_tokens": 16013749.0,
      "step": 7175
    },
    {
      "entropy": 5.570797538757324,
      "epoch": 0.8064244398270343,
      "grad_norm": 1.0,
      "learning_rate": 0.0004945516288805035,
      "loss": 5.6095,
      "mean_token_accuracy": 0.17374878674745559,
      "num_tokens": 16025131.0,
      "step": 7180
    },
    {
      "entropy": 5.614803266525269,
      "epoch": 0.8069860167349918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004945428435836864,
      "loss": 5.4795,
      "mean_token_accuracy": 0.1821274533867836,
      "num_tokens": 16037269.0,
      "step": 7185
    },
    {
      "entropy": 5.701347589492798,
      "epoch": 0.8075475936429494,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004945340512965401,
      "loss": 5.6682,
      "mean_token_accuracy": 0.17207725644111632,
      "num_tokens": 16050992.0,
      "step": 7190
    },
    {
      "entropy": 5.539309120178222,
      "epoch": 0.8081091705509069,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004945252520193443,
      "loss": 5.5066,
      "mean_token_accuracy": 0.17609120309352874,
      "num_tokens": 16062309.0,
      "step": 7195
    },
    {
      "entropy": 5.7033275127410885,
      "epoch": 0.8086707474588645,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004945164457523795,
      "loss": 5.539,
      "mean_token_accuracy": 0.17798758447170257,
      "num_tokens": 16073079.0,
      "step": 7200
    },
    {
      "entropy": 5.550238037109375,
      "epoch": 0.809232324366822,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004945076324959258,
      "loss": 5.4279,
      "mean_token_accuracy": 0.18517227917909623,
      "num_tokens": 16082885.0,
      "step": 7205
    },
    {
      "entropy": 5.3724476337432865,
      "epoch": 0.8097939012747796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004944988122502641,
      "loss": 5.3889,
      "mean_token_accuracy": 0.19014929682016374,
      "num_tokens": 16094173.0,
      "step": 7210
    },
    {
      "entropy": 5.52805986404419,
      "epoch": 0.8103554781827371,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000494489985015675,
      "loss": 5.439,
      "mean_token_accuracy": 0.1837968036532402,
      "num_tokens": 16106351.0,
      "step": 7215
    },
    {
      "entropy": 5.511557292938233,
      "epoch": 0.8109170550906947,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004944811507924396,
      "loss": 5.4244,
      "mean_token_accuracy": 0.1828888714313507,
      "num_tokens": 16116892.0,
      "step": 7220
    },
    {
      "entropy": 5.580978870391846,
      "epoch": 0.8114786319986522,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004944723095808393,
      "loss": 5.5735,
      "mean_token_accuracy": 0.17245373576879502,
      "num_tokens": 16128316.0,
      "step": 7225
    },
    {
      "entropy": 5.547515153884888,
      "epoch": 0.8120402089066098,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004944634613811555,
      "loss": 5.4063,
      "mean_token_accuracy": 0.18574077785015106,
      "num_tokens": 16139245.0,
      "step": 7230
    },
    {
      "entropy": 5.523660326004029,
      "epoch": 0.8126017858145673,
      "grad_norm": 1.078125,
      "learning_rate": 0.00049445460619367,
      "loss": 5.4442,
      "mean_token_accuracy": 0.18383751958608627,
      "num_tokens": 16150893.0,
      "step": 7235
    },
    {
      "entropy": 5.608372926712036,
      "epoch": 0.8131633627225249,
      "grad_norm": 1.8125,
      "learning_rate": 0.0004944457440186648,
      "loss": 5.5954,
      "mean_token_accuracy": 0.16272716075181962,
      "num_tokens": 16163430.0,
      "step": 7240
    },
    {
      "entropy": 5.695658922195435,
      "epoch": 0.8137249396304824,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004944368748564219,
      "loss": 5.4608,
      "mean_token_accuracy": 0.18355578631162645,
      "num_tokens": 16173954.0,
      "step": 7245
    },
    {
      "entropy": 5.534504842758179,
      "epoch": 0.81428651653844,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004944279987072238,
      "loss": 5.5219,
      "mean_token_accuracy": 0.17579175531864166,
      "num_tokens": 16185491.0,
      "step": 7250
    },
    {
      "entropy": 5.453217554092407,
      "epoch": 0.8148480934463975,
      "grad_norm": 1.125,
      "learning_rate": 0.0004944191155713531,
      "loss": 5.3823,
      "mean_token_accuracy": 0.19262163192033768,
      "num_tokens": 16195301.0,
      "step": 7255
    },
    {
      "entropy": 5.588571548461914,
      "epoch": 0.8154096703543551,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004944102254490928,
      "loss": 5.5071,
      "mean_token_accuracy": 0.17802703380584717,
      "num_tokens": 16206620.0,
      "step": 7260
    },
    {
      "entropy": 5.614008474349975,
      "epoch": 0.8159712472623126,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004944013283407257,
      "loss": 5.5808,
      "mean_token_accuracy": 0.17722658514976503,
      "num_tokens": 16218693.0,
      "step": 7265
    },
    {
      "entropy": 5.627243709564209,
      "epoch": 0.8165328241702701,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943924242465352,
      "loss": 5.479,
      "mean_token_accuracy": 0.1865539163351059,
      "num_tokens": 16229124.0,
      "step": 7270
    },
    {
      "entropy": 5.536802005767822,
      "epoch": 0.8170944010782276,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004943835131668049,
      "loss": 5.4441,
      "mean_token_accuracy": 0.18285898864269257,
      "num_tokens": 16239999.0,
      "step": 7275
    },
    {
      "entropy": 5.54294638633728,
      "epoch": 0.8176559779861852,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004943745951018184,
      "loss": 5.5336,
      "mean_token_accuracy": 0.18246021568775178,
      "num_tokens": 16251027.0,
      "step": 7280
    },
    {
      "entropy": 5.589165830612183,
      "epoch": 0.8182175548941427,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004943656700518597,
      "loss": 5.4917,
      "mean_token_accuracy": 0.18202891945838928,
      "num_tokens": 16262924.0,
      "step": 7285
    },
    {
      "entropy": 5.6963804244995115,
      "epoch": 0.8187791318021003,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004943567380172129,
      "loss": 5.5783,
      "mean_token_accuracy": 0.1787103682756424,
      "num_tokens": 16274481.0,
      "step": 7290
    },
    {
      "entropy": 5.572768259048462,
      "epoch": 0.8193407087100578,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004943477989981627,
      "loss": 5.5269,
      "mean_token_accuracy": 0.18289621770381928,
      "num_tokens": 16285214.0,
      "step": 7295
    },
    {
      "entropy": 5.599220180511475,
      "epoch": 0.8199022856180154,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943388529949933,
      "loss": 5.582,
      "mean_token_accuracy": 0.17892631590366365,
      "num_tokens": 16296077.0,
      "step": 7300
    },
    {
      "entropy": 5.642230701446533,
      "epoch": 0.8204638625259729,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004943299000079899,
      "loss": 5.4796,
      "mean_token_accuracy": 0.1827496901154518,
      "num_tokens": 16307514.0,
      "step": 7305
    },
    {
      "entropy": 5.673485326766968,
      "epoch": 0.8210254394339305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004943209400374373,
      "loss": 5.4862,
      "mean_token_accuracy": 0.18291833847761155,
      "num_tokens": 16318734.0,
      "step": 7310
    },
    {
      "entropy": 5.500394058227539,
      "epoch": 0.821587016341888,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000494311973083621,
      "loss": 5.5279,
      "mean_token_accuracy": 0.18114324957132338,
      "num_tokens": 16329057.0,
      "step": 7315
    },
    {
      "entropy": 5.554307222366333,
      "epoch": 0.8221485932498456,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004943029991468264,
      "loss": 5.4449,
      "mean_token_accuracy": 0.18278637677431106,
      "num_tokens": 16339646.0,
      "step": 7320
    },
    {
      "entropy": 5.6280481815338135,
      "epoch": 0.8227101701578031,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004942940182273393,
      "loss": 5.5349,
      "mean_token_accuracy": 0.17492180764675141,
      "num_tokens": 16349783.0,
      "step": 7325
    },
    {
      "entropy": 5.573806858062744,
      "epoch": 0.8232717470657607,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004942850303254455,
      "loss": 5.4408,
      "mean_token_accuracy": 0.18325956016778946,
      "num_tokens": 16361078.0,
      "step": 7330
    },
    {
      "entropy": 5.46482663154602,
      "epoch": 0.8238333239737182,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004942760354414313,
      "loss": 5.4668,
      "mean_token_accuracy": 0.1810009852051735,
      "num_tokens": 16371308.0,
      "step": 7335
    },
    {
      "entropy": 5.558573484420776,
      "epoch": 0.8243949008816758,
      "grad_norm": 1.125,
      "learning_rate": 0.0004942670335755831,
      "loss": 5.4977,
      "mean_token_accuracy": 0.18435774445533754,
      "num_tokens": 16382682.0,
      "step": 7340
    },
    {
      "entropy": 5.647301387786865,
      "epoch": 0.8249564777896333,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942580247281876,
      "loss": 5.5319,
      "mean_token_accuracy": 0.17990350127220153,
      "num_tokens": 16393715.0,
      "step": 7345
    },
    {
      "entropy": 5.471990966796875,
      "epoch": 0.8255180546975909,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004942490088995315,
      "loss": 5.4967,
      "mean_token_accuracy": 0.18317774385213853,
      "num_tokens": 16406332.0,
      "step": 7350
    },
    {
      "entropy": 5.552273082733154,
      "epoch": 0.8260796316055484,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004942399860899018,
      "loss": 5.4343,
      "mean_token_accuracy": 0.17803783118724822,
      "num_tokens": 16417702.0,
      "step": 7355
    },
    {
      "entropy": 5.636402893066406,
      "epoch": 0.826641208513506,
      "grad_norm": 1.0,
      "learning_rate": 0.0004942309562995862,
      "loss": 5.4714,
      "mean_token_accuracy": 0.17821657806634902,
      "num_tokens": 16428497.0,
      "step": 7360
    },
    {
      "entropy": 5.520858240127564,
      "epoch": 0.8272027854214634,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004942219195288716,
      "loss": 5.4722,
      "mean_token_accuracy": 0.18496584743261338,
      "num_tokens": 16440637.0,
      "step": 7365
    },
    {
      "entropy": 5.549655771255493,
      "epoch": 0.827764362329421,
      "grad_norm": 1.125,
      "learning_rate": 0.0004942128757780463,
      "loss": 5.4996,
      "mean_token_accuracy": 0.17600851505994797,
      "num_tokens": 16453213.0,
      "step": 7370
    },
    {
      "entropy": 5.607975721359253,
      "epoch": 0.8283259392373785,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004942038250473979,
      "loss": 5.5122,
      "mean_token_accuracy": 0.18347849547863007,
      "num_tokens": 16464345.0,
      "step": 7375
    },
    {
      "entropy": 5.51188178062439,
      "epoch": 0.8288875161453361,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004941947673372147,
      "loss": 5.3986,
      "mean_token_accuracy": 0.18549505174160003,
      "num_tokens": 16475399.0,
      "step": 7380
    },
    {
      "entropy": 5.499198007583618,
      "epoch": 0.8294490930532936,
      "grad_norm": 1.125,
      "learning_rate": 0.0004941857026477851,
      "loss": 5.3742,
      "mean_token_accuracy": 0.18233252912759781,
      "num_tokens": 16485439.0,
      "step": 7385
    },
    {
      "entropy": 5.474674320220947,
      "epoch": 0.8300106699612512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004941766309793978,
      "loss": 5.3844,
      "mean_token_accuracy": 0.18662626445293426,
      "num_tokens": 16496765.0,
      "step": 7390
    },
    {
      "entropy": 5.485547924041748,
      "epoch": 0.8305722468692087,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004941675523323414,
      "loss": 5.4124,
      "mean_token_accuracy": 0.18218796849250793,
      "num_tokens": 16509199.0,
      "step": 7395
    },
    {
      "entropy": 5.602842330932617,
      "epoch": 0.8311338237771663,
      "grad_norm": 1.125,
      "learning_rate": 0.0004941584667069052,
      "loss": 5.5736,
      "mean_token_accuracy": 0.1746901348233223,
      "num_tokens": 16520073.0,
      "step": 7400
    },
    {
      "entropy": 5.562895202636719,
      "epoch": 0.8316954006851238,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004941493741033784,
      "loss": 5.4554,
      "mean_token_accuracy": 0.18130480796098708,
      "num_tokens": 16530919.0,
      "step": 7405
    },
    {
      "entropy": 5.5857038497924805,
      "epoch": 0.8322569775930814,
      "grad_norm": 2.453125,
      "learning_rate": 0.0004941402745220507,
      "loss": 5.4677,
      "mean_token_accuracy": 0.1780529499053955,
      "num_tokens": 16542591.0,
      "step": 7410
    },
    {
      "entropy": 5.578108072280884,
      "epoch": 0.8328185545010389,
      "grad_norm": 1.0,
      "learning_rate": 0.0004941311679632115,
      "loss": 5.4764,
      "mean_token_accuracy": 0.18124143183231353,
      "num_tokens": 16554984.0,
      "step": 7415
    },
    {
      "entropy": 5.556028604507446,
      "epoch": 0.8333801314089965,
      "grad_norm": 1.125,
      "learning_rate": 0.000494122054427151,
      "loss": 5.3842,
      "mean_token_accuracy": 0.18570197522640228,
      "num_tokens": 16566598.0,
      "step": 7420
    },
    {
      "entropy": 5.5887126445770265,
      "epoch": 0.833941708316954,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004941129339141592,
      "loss": 5.5266,
      "mean_token_accuracy": 0.1773352026939392,
      "num_tokens": 16577868.0,
      "step": 7425
    },
    {
      "entropy": 5.532208108901978,
      "epoch": 0.8345032852249116,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004941038064245267,
      "loss": 5.4716,
      "mean_token_accuracy": 0.1792238563299179,
      "num_tokens": 16587891.0,
      "step": 7430
    },
    {
      "entropy": 5.540295124053955,
      "epoch": 0.8350648621328691,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000494094671958544,
      "loss": 5.4245,
      "mean_token_accuracy": 0.18846536576747894,
      "num_tokens": 16599924.0,
      "step": 7435
    },
    {
      "entropy": 5.48519458770752,
      "epoch": 0.8356264390408267,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004940855305165019,
      "loss": 5.3775,
      "mean_token_accuracy": 0.18963050693273545,
      "num_tokens": 16609958.0,
      "step": 7440
    },
    {
      "entropy": 5.574797058105469,
      "epoch": 0.8361880159487842,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004940763820986916,
      "loss": 5.505,
      "mean_token_accuracy": 0.17985743135213852,
      "num_tokens": 16621159.0,
      "step": 7445
    },
    {
      "entropy": 5.633857870101929,
      "epoch": 0.8367495928567418,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004940672267054044,
      "loss": 5.6022,
      "mean_token_accuracy": 0.1801896035671234,
      "num_tokens": 16632263.0,
      "step": 7450
    },
    {
      "entropy": 5.534639835357666,
      "epoch": 0.8373111697646993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940580643369316,
      "loss": 5.4772,
      "mean_token_accuracy": 0.18106465190649032,
      "num_tokens": 16644117.0,
      "step": 7455
    },
    {
      "entropy": 5.4652656555175785,
      "epoch": 0.8378727466726568,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004940488949935651,
      "loss": 5.3964,
      "mean_token_accuracy": 0.18824367672204972,
      "num_tokens": 16654529.0,
      "step": 7460
    },
    {
      "entropy": 5.496886730194092,
      "epoch": 0.8384343235806143,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004940397186755968,
      "loss": 5.4067,
      "mean_token_accuracy": 0.18636858761310576,
      "num_tokens": 16666191.0,
      "step": 7465
    },
    {
      "entropy": 5.55779185295105,
      "epoch": 0.8389959004885719,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004940305353833191,
      "loss": 5.3904,
      "mean_token_accuracy": 0.18847827166318892,
      "num_tokens": 16677252.0,
      "step": 7470
    },
    {
      "entropy": 5.480805492401123,
      "epoch": 0.8395574773965294,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000494021345117024,
      "loss": 5.4425,
      "mean_token_accuracy": 0.18266780376434327,
      "num_tokens": 16689487.0,
      "step": 7475
    },
    {
      "entropy": 5.428728103637695,
      "epoch": 0.840119054304487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004940121478770044,
      "loss": 5.3337,
      "mean_token_accuracy": 0.18980036675930023,
      "num_tokens": 16699955.0,
      "step": 7480
    },
    {
      "entropy": 5.458487129211425,
      "epoch": 0.8406806312124445,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004940029436635531,
      "loss": 5.4081,
      "mean_token_accuracy": 0.17810123562812805,
      "num_tokens": 16711071.0,
      "step": 7485
    },
    {
      "entropy": 5.489441776275635,
      "epoch": 0.8412422081204021,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004939937324769631,
      "loss": 5.4067,
      "mean_token_accuracy": 0.1881688728928566,
      "num_tokens": 16722115.0,
      "step": 7490
    },
    {
      "entropy": 5.56087064743042,
      "epoch": 0.8418037850283596,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004939845143175276,
      "loss": 5.4482,
      "mean_token_accuracy": 0.1789998859167099,
      "num_tokens": 16733146.0,
      "step": 7495
    },
    {
      "entropy": 5.484468936920166,
      "epoch": 0.8423653619363172,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004939752891855404,
      "loss": 5.4175,
      "mean_token_accuracy": 0.1852555364370346,
      "num_tokens": 16745543.0,
      "step": 7500
    },
    {
      "entropy": 5.543926095962524,
      "epoch": 0.8429269388442747,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939660570812951,
      "loss": 5.5012,
      "mean_token_accuracy": 0.18036285638809205,
      "num_tokens": 16755782.0,
      "step": 7505
    },
    {
      "entropy": 5.601821517944336,
      "epoch": 0.8434885157522323,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004939568180050855,
      "loss": 5.5139,
      "mean_token_accuracy": 0.17965976297855377,
      "num_tokens": 16767509.0,
      "step": 7510
    },
    {
      "entropy": 5.489942646026611,
      "epoch": 0.8440500926601898,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004939475719572059,
      "loss": 5.3845,
      "mean_token_accuracy": 0.18219635635614395,
      "num_tokens": 16779257.0,
      "step": 7515
    },
    {
      "entropy": 5.4493317127227785,
      "epoch": 0.8446116695681474,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004939383189379507,
      "loss": 5.4323,
      "mean_token_accuracy": 0.18184877187013626,
      "num_tokens": 16790769.0,
      "step": 7520
    },
    {
      "entropy": 5.522083950042725,
      "epoch": 0.8451732464761049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004939290589476145,
      "loss": 5.4275,
      "mean_token_accuracy": 0.18631497025489807,
      "num_tokens": 16800054.0,
      "step": 7525
    },
    {
      "entropy": 5.6331078052520756,
      "epoch": 0.8457348233840625,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004939197919864922,
      "loss": 5.5246,
      "mean_token_accuracy": 0.17630529552698135,
      "num_tokens": 16811485.0,
      "step": 7530
    },
    {
      "entropy": 5.6258291721344,
      "epoch": 0.84629640029202,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004939105180548788,
      "loss": 5.5823,
      "mean_token_accuracy": 0.18421612083911895,
      "num_tokens": 16822683.0,
      "step": 7535
    },
    {
      "entropy": 5.619016456604004,
      "epoch": 0.8468579771999776,
      "grad_norm": 1.125,
      "learning_rate": 0.0004939012371530696,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1803028404712677,
      "num_tokens": 16834200.0,
      "step": 7540
    },
    {
      "entropy": 5.604679012298584,
      "epoch": 0.8474195541079351,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938919492813601,
      "loss": 5.4986,
      "mean_token_accuracy": 0.17613030672073365,
      "num_tokens": 16846399.0,
      "step": 7545
    },
    {
      "entropy": 5.614399719238281,
      "epoch": 0.8479811310158927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938826544400461,
      "loss": 5.6407,
      "mean_token_accuracy": 0.17467446774244308,
      "num_tokens": 16858322.0,
      "step": 7550
    },
    {
      "entropy": 5.5493834018707275,
      "epoch": 0.8485427079238502,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004938733526294234,
      "loss": 5.4735,
      "mean_token_accuracy": 0.1825816497206688,
      "num_tokens": 16869537.0,
      "step": 7555
    },
    {
      "entropy": 5.691517639160156,
      "epoch": 0.8491042848318077,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004938640438497883,
      "loss": 5.5038,
      "mean_token_accuracy": 0.1843949541449547,
      "num_tokens": 16881566.0,
      "step": 7560
    },
    {
      "entropy": 5.4563212394714355,
      "epoch": 0.8496658617397652,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004938547281014372,
      "loss": 5.4316,
      "mean_token_accuracy": 0.18823107182979584,
      "num_tokens": 16892959.0,
      "step": 7565
    },
    {
      "entropy": 5.5974513530731205,
      "epoch": 0.8502274386477228,
      "grad_norm": 1.125,
      "learning_rate": 0.0004938454053846666,
      "loss": 5.4844,
      "mean_token_accuracy": 0.17420702278614045,
      "num_tokens": 16904809.0,
      "step": 7570
    },
    {
      "entropy": 5.612712335586548,
      "epoch": 0.8507890155556803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938360756997735,
      "loss": 5.507,
      "mean_token_accuracy": 0.17359740287065506,
      "num_tokens": 16916969.0,
      "step": 7575
    },
    {
      "entropy": 5.512321090698242,
      "epoch": 0.8513505924636379,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004938267390470547,
      "loss": 5.4074,
      "mean_token_accuracy": 0.18779443800449372,
      "num_tokens": 16928253.0,
      "step": 7580
    },
    {
      "entropy": 5.498369884490967,
      "epoch": 0.8519121693715954,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004938173954268079,
      "loss": 5.4415,
      "mean_token_accuracy": 0.17914493680000304,
      "num_tokens": 16939875.0,
      "step": 7585
    },
    {
      "entropy": 5.662486600875854,
      "epoch": 0.852473746279553,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004938080448393302,
      "loss": 5.5612,
      "mean_token_accuracy": 0.17964935004711152,
      "num_tokens": 16950173.0,
      "step": 7590
    },
    {
      "entropy": 5.537886667251587,
      "epoch": 0.8530353231875105,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004937986872849195,
      "loss": 5.4761,
      "mean_token_accuracy": 0.18149444758892058,
      "num_tokens": 16960933.0,
      "step": 7595
    },
    {
      "entropy": 5.488050746917724,
      "epoch": 0.8535969000954681,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004937893227638738,
      "loss": 5.3844,
      "mean_token_accuracy": 0.18457337468862534,
      "num_tokens": 16972284.0,
      "step": 7600
    },
    {
      "entropy": 5.519171905517578,
      "epoch": 0.8541584770034256,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004937799512764912,
      "loss": 5.4069,
      "mean_token_accuracy": 0.1892599418759346,
      "num_tokens": 16983541.0,
      "step": 7605
    },
    {
      "entropy": 5.601930427551269,
      "epoch": 0.8547200539113832,
      "grad_norm": 1.109375,
      "learning_rate": 0.00049377057282307,
      "loss": 5.5423,
      "mean_token_accuracy": 0.17711724936962128,
      "num_tokens": 16993773.0,
      "step": 7610
    },
    {
      "entropy": 5.629774951934815,
      "epoch": 0.8552816308193407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000493761187403909,
      "loss": 5.5274,
      "mean_token_accuracy": 0.1779188483953476,
      "num_tokens": 17005287.0,
      "step": 7615
    },
    {
      "entropy": 5.60647406578064,
      "epoch": 0.8558432077272983,
      "grad_norm": 1.15625,
      "learning_rate": 0.000493751795019307,
      "loss": 5.4671,
      "mean_token_accuracy": 0.18144695907831193,
      "num_tokens": 17015664.0,
      "step": 7620
    },
    {
      "entropy": 5.514392328262329,
      "epoch": 0.8564047846352558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004937423956695629,
      "loss": 5.3805,
      "mean_token_accuracy": 0.18107551336288452,
      "num_tokens": 17026876.0,
      "step": 7625
    },
    {
      "entropy": 5.44736680984497,
      "epoch": 0.8569663615432134,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004937329893549762,
      "loss": 5.4286,
      "mean_token_accuracy": 0.19533413350582124,
      "num_tokens": 17037321.0,
      "step": 7630
    },
    {
      "entropy": 5.560785102844238,
      "epoch": 0.8575279384511709,
      "grad_norm": 1.125,
      "learning_rate": 0.0004937235760758462,
      "loss": 5.511,
      "mean_token_accuracy": 0.18063998371362686,
      "num_tokens": 17047506.0,
      "step": 7635
    },
    {
      "entropy": 5.633035659790039,
      "epoch": 0.8580895153591285,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004937141558324726,
      "loss": 5.5313,
      "mean_token_accuracy": 0.17728657126426697,
      "num_tokens": 17059710.0,
      "step": 7640
    },
    {
      "entropy": 5.6412190914154055,
      "epoch": 0.858651092267086,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004937047286251556,
      "loss": 5.5482,
      "mean_token_accuracy": 0.17355645149946214,
      "num_tokens": 17070956.0,
      "step": 7645
    },
    {
      "entropy": 5.607813215255737,
      "epoch": 0.8592126691750436,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004936952944541952,
      "loss": 5.527,
      "mean_token_accuracy": 0.18062603622674941,
      "num_tokens": 17081404.0,
      "step": 7650
    },
    {
      "entropy": 5.659487533569336,
      "epoch": 0.859774246083001,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004936858533198917,
      "loss": 5.5392,
      "mean_token_accuracy": 0.17470980137586595,
      "num_tokens": 17092692.0,
      "step": 7655
    },
    {
      "entropy": 5.523286771774292,
      "epoch": 0.8603358229909586,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000493676405222546,
      "loss": 5.3869,
      "mean_token_accuracy": 0.18214413076639174,
      "num_tokens": 17103730.0,
      "step": 7660
    },
    {
      "entropy": 5.579022169113159,
      "epoch": 0.8608973998989161,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004936669501624586,
      "loss": 5.4592,
      "mean_token_accuracy": 0.1768531545996666,
      "num_tokens": 17115086.0,
      "step": 7665
    },
    {
      "entropy": 5.546690607070923,
      "epoch": 0.8614589768068737,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004936574881399308,
      "loss": 5.4663,
      "mean_token_accuracy": 0.18444371074438096,
      "num_tokens": 17125928.0,
      "step": 7670
    },
    {
      "entropy": 5.564424180984497,
      "epoch": 0.8620205537148312,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004936480191552637,
      "loss": 5.4762,
      "mean_token_accuracy": 0.18243034034967423,
      "num_tokens": 17137328.0,
      "step": 7675
    },
    {
      "entropy": 5.586181926727295,
      "epoch": 0.8625821306227888,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004936385432087589,
      "loss": 5.4764,
      "mean_token_accuracy": 0.1780944973230362,
      "num_tokens": 17148274.0,
      "step": 7680
    },
    {
      "entropy": 5.560336494445801,
      "epoch": 0.8631437075307463,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004936290603007181,
      "loss": 5.4969,
      "mean_token_accuracy": 0.17892249971628188,
      "num_tokens": 17158834.0,
      "step": 7685
    },
    {
      "entropy": 5.555933666229248,
      "epoch": 0.8637052844387039,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004936195704314432,
      "loss": 5.3564,
      "mean_token_accuracy": 0.18445111215114593,
      "num_tokens": 17169276.0,
      "step": 7690
    },
    {
      "entropy": 5.483067417144776,
      "epoch": 0.8642668613466614,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004936100736012363,
      "loss": 5.4094,
      "mean_token_accuracy": 0.18652159124612808,
      "num_tokens": 17180642.0,
      "step": 7695
    },
    {
      "entropy": 5.496457576751709,
      "epoch": 0.864828438254619,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004936005698104,
      "loss": 5.4739,
      "mean_token_accuracy": 0.1882736012339592,
      "num_tokens": 17192477.0,
      "step": 7700
    },
    {
      "entropy": 5.683662748336792,
      "epoch": 0.8653900151625765,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004935910590592367,
      "loss": 5.6081,
      "mean_token_accuracy": 0.1802874758839607,
      "num_tokens": 17202653.0,
      "step": 7705
    },
    {
      "entropy": 5.456462287902832,
      "epoch": 0.8659515920705341,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004935815413480494,
      "loss": 5.3447,
      "mean_token_accuracy": 0.19504539519548417,
      "num_tokens": 17212607.0,
      "step": 7710
    },
    {
      "entropy": 5.533054304122925,
      "epoch": 0.8665131689784916,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000493572016677141,
      "loss": 5.4907,
      "mean_token_accuracy": 0.18590737581253053,
      "num_tokens": 17223701.0,
      "step": 7715
    },
    {
      "entropy": 5.591048717498779,
      "epoch": 0.8670747458864492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004935624850468148,
      "loss": 5.3938,
      "mean_token_accuracy": 0.18450769633054734,
      "num_tokens": 17234878.0,
      "step": 7720
    },
    {
      "entropy": 5.526391553878784,
      "epoch": 0.8676363227944067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004935529464573742,
      "loss": 5.4661,
      "mean_token_accuracy": 0.18419525772333145,
      "num_tokens": 17245813.0,
      "step": 7725
    },
    {
      "entropy": 5.57021656036377,
      "epoch": 0.8681978997023643,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004935434009091232,
      "loss": 5.5264,
      "mean_token_accuracy": 0.18249285221099854,
      "num_tokens": 17257378.0,
      "step": 7730
    },
    {
      "entropy": 5.511587762832642,
      "epoch": 0.8687594766103218,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004935338484023655,
      "loss": 5.4499,
      "mean_token_accuracy": 0.18388038873672485,
      "num_tokens": 17269108.0,
      "step": 7735
    },
    {
      "entropy": 5.583608102798462,
      "epoch": 0.8693210535182794,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004935242889374052,
      "loss": 5.4516,
      "mean_token_accuracy": 0.18015028089284896,
      "num_tokens": 17280643.0,
      "step": 7740
    },
    {
      "entropy": 5.68832311630249,
      "epoch": 0.8698826304262369,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004935147225145469,
      "loss": 5.5147,
      "mean_token_accuracy": 0.18133498579263688,
      "num_tokens": 17291236.0,
      "step": 7745
    },
    {
      "entropy": 5.624429750442505,
      "epoch": 0.8704442073341944,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004935051491340951,
      "loss": 5.5803,
      "mean_token_accuracy": 0.1733072206377983,
      "num_tokens": 17303547.0,
      "step": 7750
    },
    {
      "entropy": 5.653601026535034,
      "epoch": 0.8710057842421519,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004934955687963546,
      "loss": 5.5516,
      "mean_token_accuracy": 0.17682304829359055,
      "num_tokens": 17314803.0,
      "step": 7755
    },
    {
      "entropy": 5.541751766204834,
      "epoch": 0.8715673611501095,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004934859815016304,
      "loss": 5.4321,
      "mean_token_accuracy": 0.18183032125234605,
      "num_tokens": 17326009.0,
      "step": 7760
    },
    {
      "entropy": 5.583894157409668,
      "epoch": 0.872128938058067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004934763872502278,
      "loss": 5.4441,
      "mean_token_accuracy": 0.18149555027484893,
      "num_tokens": 17337078.0,
      "step": 7765
    },
    {
      "entropy": 5.613049077987671,
      "epoch": 0.8726905149660246,
      "grad_norm": 1.125,
      "learning_rate": 0.0004934667860424523,
      "loss": 5.5476,
      "mean_token_accuracy": 0.17916815876960754,
      "num_tokens": 17349266.0,
      "step": 7770
    },
    {
      "entropy": 5.446396732330323,
      "epoch": 0.8732520918739821,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004934571778786096,
      "loss": 5.3544,
      "mean_token_accuracy": 0.1859301820397377,
      "num_tokens": 17360327.0,
      "step": 7775
    },
    {
      "entropy": 5.616374063491821,
      "epoch": 0.8738136687819397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004934475627590057,
      "loss": 5.5105,
      "mean_token_accuracy": 0.17509843558073043,
      "num_tokens": 17372727.0,
      "step": 7780
    },
    {
      "entropy": 5.6568365573883055,
      "epoch": 0.8743752456898972,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004934379406839466,
      "loss": 5.5699,
      "mean_token_accuracy": 0.1762222021818161,
      "num_tokens": 17384160.0,
      "step": 7785
    },
    {
      "entropy": 5.543169689178467,
      "epoch": 0.8749368225978548,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004934283116537387,
      "loss": 5.4789,
      "mean_token_accuracy": 0.18203620612621307,
      "num_tokens": 17396009.0,
      "step": 7790
    },
    {
      "entropy": 5.58352484703064,
      "epoch": 0.8754983995058123,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004934186756686888,
      "loss": 5.5019,
      "mean_token_accuracy": 0.18040887862443925,
      "num_tokens": 17406903.0,
      "step": 7795
    },
    {
      "entropy": 5.532858848571777,
      "epoch": 0.8760599764137699,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004934090327291034,
      "loss": 5.3545,
      "mean_token_accuracy": 0.18683639913797379,
      "num_tokens": 17417694.0,
      "step": 7800
    },
    {
      "entropy": 5.512098360061645,
      "epoch": 0.8766215533217274,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004933993828352898,
      "loss": 5.5861,
      "mean_token_accuracy": 0.1832665055990219,
      "num_tokens": 17428731.0,
      "step": 7805
    },
    {
      "entropy": 5.538304042816162,
      "epoch": 0.877183130229685,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004933897259875551,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18505916148424148,
      "num_tokens": 17439242.0,
      "step": 7810
    },
    {
      "entropy": 5.573413228988647,
      "epoch": 0.8777447071376425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004933800621862069,
      "loss": 5.4533,
      "mean_token_accuracy": 0.18363437950611114,
      "num_tokens": 17449980.0,
      "step": 7815
    },
    {
      "entropy": 5.633233833312988,
      "epoch": 0.8783062840456001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004933703914315527,
      "loss": 5.5467,
      "mean_token_accuracy": 0.17176835238933563,
      "num_tokens": 17464140.0,
      "step": 7820
    },
    {
      "entropy": 5.686376428604126,
      "epoch": 0.8788678609535576,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933607137239006,
      "loss": 5.5631,
      "mean_token_accuracy": 0.1781417980790138,
      "num_tokens": 17476246.0,
      "step": 7825
    },
    {
      "entropy": 5.5935876846313475,
      "epoch": 0.8794294378615152,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004933510290635587,
      "loss": 5.5453,
      "mean_token_accuracy": 0.16770657002925873,
      "num_tokens": 17488169.0,
      "step": 7830
    },
    {
      "entropy": 5.506634664535523,
      "epoch": 0.8799910147694727,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004933413374508353,
      "loss": 5.3804,
      "mean_token_accuracy": 0.18736401349306106,
      "num_tokens": 17498465.0,
      "step": 7835
    },
    {
      "entropy": 5.559989976882934,
      "epoch": 0.8805525916774303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004933316388860391,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1897400736808777,
      "num_tokens": 17509217.0,
      "step": 7840
    },
    {
      "entropy": 5.5196442127227785,
      "epoch": 0.8811141685853877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004933219333694787,
      "loss": 5.4229,
      "mean_token_accuracy": 0.18789440393447876,
      "num_tokens": 17519851.0,
      "step": 7845
    },
    {
      "entropy": 5.5505396842956545,
      "epoch": 0.8816757454933453,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004933122209014634,
      "loss": 5.4804,
      "mean_token_accuracy": 0.18657491058111192,
      "num_tokens": 17532144.0,
      "step": 7850
    },
    {
      "entropy": 5.558473539352417,
      "epoch": 0.8822373224013028,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004933025014823022,
      "loss": 5.3893,
      "mean_token_accuracy": 0.19165836572647094,
      "num_tokens": 17542929.0,
      "step": 7855
    },
    {
      "entropy": 5.574261283874511,
      "epoch": 0.8827988993092604,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932927751123048,
      "loss": 5.5689,
      "mean_token_accuracy": 0.17471055686473846,
      "num_tokens": 17554931.0,
      "step": 7860
    },
    {
      "entropy": 5.492510986328125,
      "epoch": 0.8833604762172179,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004932830417917808,
      "loss": 5.3746,
      "mean_token_accuracy": 0.17897113412618637,
      "num_tokens": 17566026.0,
      "step": 7865
    },
    {
      "entropy": 5.54307804107666,
      "epoch": 0.8839220531251755,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00049327330152104,
      "loss": 5.4819,
      "mean_token_accuracy": 0.18322187662124634,
      "num_tokens": 17578396.0,
      "step": 7870
    },
    {
      "entropy": 5.572889614105224,
      "epoch": 0.884483630033133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004932635543003927,
      "loss": 5.4725,
      "mean_token_accuracy": 0.1826783001422882,
      "num_tokens": 17590559.0,
      "step": 7875
    },
    {
      "entropy": 5.549179983139038,
      "epoch": 0.8850452069410906,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004932538001301491,
      "loss": 5.4368,
      "mean_token_accuracy": 0.18069201856851577,
      "num_tokens": 17600833.0,
      "step": 7880
    },
    {
      "entropy": 5.508893394470215,
      "epoch": 0.8856067838490481,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004932440390106199,
      "loss": 5.421,
      "mean_token_accuracy": 0.18312568813562394,
      "num_tokens": 17612166.0,
      "step": 7885
    },
    {
      "entropy": 5.611009645462036,
      "epoch": 0.8861683607570057,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932342709421159,
      "loss": 5.5451,
      "mean_token_accuracy": 0.1778656378388405,
      "num_tokens": 17623070.0,
      "step": 7890
    },
    {
      "entropy": 5.678601169586182,
      "epoch": 0.8867299376649632,
      "grad_norm": 1.078125,
      "learning_rate": 0.000493224495924948,
      "loss": 5.607,
      "mean_token_accuracy": 0.17940963357686995,
      "num_tokens": 17633319.0,
      "step": 7895
    },
    {
      "entropy": 5.538157796859741,
      "epoch": 0.8872915145729208,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004932147139594276,
      "loss": 5.4118,
      "mean_token_accuracy": 0.18473155200481414,
      "num_tokens": 17643818.0,
      "step": 7900
    },
    {
      "entropy": 5.541986608505249,
      "epoch": 0.8878530914808783,
      "grad_norm": 1.125,
      "learning_rate": 0.0004932049250458659,
      "loss": 5.4128,
      "mean_token_accuracy": 0.184093701839447,
      "num_tokens": 17655665.0,
      "step": 7905
    },
    {
      "entropy": 5.514326477050782,
      "epoch": 0.8884146683888359,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004931951291845749,
      "loss": 5.3412,
      "mean_token_accuracy": 0.1917750731110573,
      "num_tokens": 17666200.0,
      "step": 7910
    },
    {
      "entropy": 5.500199794769287,
      "epoch": 0.8889762452967934,
      "grad_norm": 1.125,
      "learning_rate": 0.0004931853263758663,
      "loss": 5.5377,
      "mean_token_accuracy": 0.17567701786756515,
      "num_tokens": 17677299.0,
      "step": 7915
    },
    {
      "entropy": 5.617593383789062,
      "epoch": 0.889537822204751,
      "grad_norm": 1.0,
      "learning_rate": 0.0004931755166200523,
      "loss": 5.3977,
      "mean_token_accuracy": 0.19115814864635466,
      "num_tokens": 17687870.0,
      "step": 7920
    },
    {
      "entropy": 5.480732679367065,
      "epoch": 0.8900993991127085,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004931656999174452,
      "loss": 5.4234,
      "mean_token_accuracy": 0.18501016199588777,
      "num_tokens": 17698853.0,
      "step": 7925
    },
    {
      "entropy": 5.513735246658325,
      "epoch": 0.8906609760206661,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004931558762683577,
      "loss": 5.4565,
      "mean_token_accuracy": 0.18694014400243758,
      "num_tokens": 17709897.0,
      "step": 7930
    },
    {
      "entropy": 5.5990739345550535,
      "epoch": 0.8912225529286236,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004931460456731024,
      "loss": 5.457,
      "mean_token_accuracy": 0.1840829446911812,
      "num_tokens": 17720974.0,
      "step": 7935
    },
    {
      "entropy": 5.638696908950806,
      "epoch": 0.8917841298365812,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004931362081319925,
      "loss": 5.609,
      "mean_token_accuracy": 0.17928121984004974,
      "num_tokens": 17734449.0,
      "step": 7940
    },
    {
      "entropy": 5.488146877288818,
      "epoch": 0.8923457067445386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004931263636453409,
      "loss": 5.3485,
      "mean_token_accuracy": 0.18492533564567565,
      "num_tokens": 17744735.0,
      "step": 7945
    },
    {
      "entropy": 5.478331613540649,
      "epoch": 0.8929072836524962,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004931165122134616,
      "loss": 5.3693,
      "mean_token_accuracy": 0.1868396982550621,
      "num_tokens": 17755641.0,
      "step": 7950
    },
    {
      "entropy": 5.519529724121094,
      "epoch": 0.8934688605604537,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004931066538366677,
      "loss": 5.4927,
      "mean_token_accuracy": 0.17717790752649307,
      "num_tokens": 17767621.0,
      "step": 7955
    },
    {
      "entropy": 5.677061939239502,
      "epoch": 0.8940304374684113,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004930967885152734,
      "loss": 5.5796,
      "mean_token_accuracy": 0.17662698030471802,
      "num_tokens": 17779507.0,
      "step": 7960
    },
    {
      "entropy": 5.510627222061157,
      "epoch": 0.8945920143763688,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004930869162495929,
      "loss": 5.4236,
      "mean_token_accuracy": 0.1815606653690338,
      "num_tokens": 17790371.0,
      "step": 7965
    },
    {
      "entropy": 5.547703981399536,
      "epoch": 0.8951535912843264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004930770370399403,
      "loss": 5.4445,
      "mean_token_accuracy": 0.18780944645404815,
      "num_tokens": 17800903.0,
      "step": 7970
    },
    {
      "entropy": 5.648083066940307,
      "epoch": 0.8957151681922839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004930671508866303,
      "loss": 5.6305,
      "mean_token_accuracy": 0.1707410454750061,
      "num_tokens": 17812053.0,
      "step": 7975
    },
    {
      "entropy": 5.556245565414429,
      "epoch": 0.8962767451002415,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004930572577899776,
      "loss": 5.3557,
      "mean_token_accuracy": 0.18426816314458846,
      "num_tokens": 17822113.0,
      "step": 7980
    },
    {
      "entropy": 5.487072229385376,
      "epoch": 0.896838322008199,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004930473577502972,
      "loss": 5.4065,
      "mean_token_accuracy": 0.1852512001991272,
      "num_tokens": 17832741.0,
      "step": 7985
    },
    {
      "entropy": 5.3799156665802,
      "epoch": 0.8973998989161566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004930374507679044,
      "loss": 5.3393,
      "mean_token_accuracy": 0.1804100200533867,
      "num_tokens": 17842968.0,
      "step": 7990
    },
    {
      "entropy": 5.666249465942383,
      "epoch": 0.8979614758241141,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004930275368431146,
      "loss": 5.5966,
      "mean_token_accuracy": 0.17014015316963196,
      "num_tokens": 17854295.0,
      "step": 7995
    },
    {
      "entropy": 5.729411935806274,
      "epoch": 0.8985230527320717,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004930176159762435,
      "loss": 5.526,
      "mean_token_accuracy": 0.18004802167415618,
      "num_tokens": 17865597.0,
      "step": 8000
    },
    {
      "entropy": 5.370964527130127,
      "epoch": 0.8990846296400292,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004930076881676069,
      "loss": 5.3041,
      "mean_token_accuracy": 0.1884951815009117,
      "num_tokens": 17877405.0,
      "step": 8005
    },
    {
      "entropy": 5.668585300445557,
      "epoch": 0.8996462065479868,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004929977534175211,
      "loss": 5.5024,
      "mean_token_accuracy": 0.18293628245592117,
      "num_tokens": 17889273.0,
      "step": 8010
    },
    {
      "entropy": 5.490920877456665,
      "epoch": 0.9002077834559443,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004929878117263021,
      "loss": 5.4114,
      "mean_token_accuracy": 0.1900405317544937,
      "num_tokens": 17899761.0,
      "step": 8015
    },
    {
      "entropy": 5.486945390701294,
      "epoch": 0.9007693603639019,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004929778630942668,
      "loss": 5.4173,
      "mean_token_accuracy": 0.18127040714025497,
      "num_tokens": 17910607.0,
      "step": 8020
    },
    {
      "entropy": 5.619477987289429,
      "epoch": 0.9013309372718594,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004929679075217317,
      "loss": 5.4752,
      "mean_token_accuracy": 0.17835913151502608,
      "num_tokens": 17922538.0,
      "step": 8025
    },
    {
      "entropy": 5.557470512390137,
      "epoch": 0.901892514179817,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004929579450090139,
      "loss": 5.4916,
      "mean_token_accuracy": 0.18290689289569856,
      "num_tokens": 17933313.0,
      "step": 8030
    },
    {
      "entropy": 5.5747209072113035,
      "epoch": 0.9024540910877745,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004929479755564305,
      "loss": 5.5871,
      "mean_token_accuracy": 0.1780872091650963,
      "num_tokens": 17944893.0,
      "step": 8035
    },
    {
      "entropy": 5.564822483062744,
      "epoch": 0.903015667995732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004929379991642991,
      "loss": 5.471,
      "mean_token_accuracy": 0.1782379612326622,
      "num_tokens": 17955019.0,
      "step": 8040
    },
    {
      "entropy": 5.657858514785767,
      "epoch": 0.9035772449036895,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004929280158329373,
      "loss": 5.497,
      "mean_token_accuracy": 0.18043357729911805,
      "num_tokens": 17965793.0,
      "step": 8045
    },
    {
      "entropy": 5.507900667190552,
      "epoch": 0.9041388218116471,
      "grad_norm": 1.21875,
      "learning_rate": 0.000492918025562663,
      "loss": 5.3321,
      "mean_token_accuracy": 0.18465083986520767,
      "num_tokens": 17975709.0,
      "step": 8050
    },
    {
      "entropy": 5.5629322052001955,
      "epoch": 0.9047003987196046,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004929080283537942,
      "loss": 5.5597,
      "mean_token_accuracy": 0.18016576319932937,
      "num_tokens": 17985405.0,
      "step": 8055
    },
    {
      "entropy": 5.622521877288818,
      "epoch": 0.9052619756275622,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004928980242066492,
      "loss": 5.5793,
      "mean_token_accuracy": 0.17631723433732988,
      "num_tokens": 17997105.0,
      "step": 8060
    },
    {
      "entropy": 5.601155948638916,
      "epoch": 0.9058235525355197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004928880131215467,
      "loss": 5.4693,
      "mean_token_accuracy": 0.18954723924398423,
      "num_tokens": 18007914.0,
      "step": 8065
    },
    {
      "entropy": 5.487916469573975,
      "epoch": 0.9063851294434773,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004928779950988053,
      "loss": 5.4823,
      "mean_token_accuracy": 0.1835661217570305,
      "num_tokens": 18018328.0,
      "step": 8070
    },
    {
      "entropy": 5.583480787277222,
      "epoch": 0.9069467063514348,
      "grad_norm": 1.125,
      "learning_rate": 0.000492867970138744,
      "loss": 5.4323,
      "mean_token_accuracy": 0.17852159291505815,
      "num_tokens": 18028568.0,
      "step": 8075
    },
    {
      "entropy": 5.5793781757354735,
      "epoch": 0.9075082832593924,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004928579382416821,
      "loss": 5.4453,
      "mean_token_accuracy": 0.18433612287044526,
      "num_tokens": 18038655.0,
      "step": 8080
    },
    {
      "entropy": 5.474751996994018,
      "epoch": 0.9080698601673499,
      "grad_norm": 1.125,
      "learning_rate": 0.000492847899407939,
      "loss": 5.3914,
      "mean_token_accuracy": 0.18396716564893723,
      "num_tokens": 18049716.0,
      "step": 8085
    },
    {
      "entropy": 5.455990314483643,
      "epoch": 0.9086314370753075,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004928378536378341,
      "loss": 5.3737,
      "mean_token_accuracy": 0.18509361743927003,
      "num_tokens": 18062530.0,
      "step": 8090
    },
    {
      "entropy": 5.545447874069214,
      "epoch": 0.909193013983265,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004928278009316877,
      "loss": 5.4721,
      "mean_token_accuracy": 0.17661340683698654,
      "num_tokens": 18073076.0,
      "step": 8095
    },
    {
      "entropy": 5.594818210601806,
      "epoch": 0.9097545908912226,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004928177412898195,
      "loss": 5.4322,
      "mean_token_accuracy": 0.1810903012752533,
      "num_tokens": 18083759.0,
      "step": 8100
    },
    {
      "entropy": 5.458002090454102,
      "epoch": 0.9103161677991801,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004928076747125499,
      "loss": 5.3797,
      "mean_token_accuracy": 0.18231212198734284,
      "num_tokens": 18095980.0,
      "step": 8105
    },
    {
      "entropy": 5.561630296707153,
      "epoch": 0.9108777447071377,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004927976012001995,
      "loss": 5.4627,
      "mean_token_accuracy": 0.18530253171920777,
      "num_tokens": 18106672.0,
      "step": 8110
    },
    {
      "entropy": 5.5100311756134035,
      "epoch": 0.9114393216150952,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004927875207530889,
      "loss": 5.4723,
      "mean_token_accuracy": 0.17796551883220674,
      "num_tokens": 18117481.0,
      "step": 8115
    },
    {
      "entropy": 5.515746688842773,
      "epoch": 0.9120008985230528,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004927774333715394,
      "loss": 5.4125,
      "mean_token_accuracy": 0.1871819466352463,
      "num_tokens": 18127602.0,
      "step": 8120
    },
    {
      "entropy": 5.627735900878906,
      "epoch": 0.9125624754310103,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004927673390558718,
      "loss": 5.4763,
      "mean_token_accuracy": 0.18249623030424117,
      "num_tokens": 18140109.0,
      "step": 8125
    },
    {
      "entropy": 5.539096021652222,
      "epoch": 0.9131240523389679,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004927572378064077,
      "loss": 5.3739,
      "mean_token_accuracy": 0.1930138126015663,
      "num_tokens": 18150271.0,
      "step": 8130
    },
    {
      "entropy": 5.552886009216309,
      "epoch": 0.9136856292469253,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004927471296234687,
      "loss": 5.4951,
      "mean_token_accuracy": 0.1850314050912857,
      "num_tokens": 18160746.0,
      "step": 8135
    },
    {
      "entropy": 5.598134851455688,
      "epoch": 0.9142472061548829,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004927370145073767,
      "loss": 5.4694,
      "mean_token_accuracy": 0.1848224326968193,
      "num_tokens": 18172271.0,
      "step": 8140
    },
    {
      "entropy": 5.504380750656128,
      "epoch": 0.9148087830628404,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004927268924584536,
      "loss": 5.4105,
      "mean_token_accuracy": 0.18245136141777038,
      "num_tokens": 18183116.0,
      "step": 8145
    },
    {
      "entropy": 5.45198016166687,
      "epoch": 0.915370359970798,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004927167634770219,
      "loss": 5.3676,
      "mean_token_accuracy": 0.1966984361410141,
      "num_tokens": 18195147.0,
      "step": 8150
    },
    {
      "entropy": 5.567431879043579,
      "epoch": 0.9159319368787555,
      "grad_norm": 1.3203125,
      "learning_rate": 0.000492706627563404,
      "loss": 5.4343,
      "mean_token_accuracy": 0.18496874570846558,
      "num_tokens": 18206217.0,
      "step": 8155
    },
    {
      "entropy": 5.51133165359497,
      "epoch": 0.9164935137867131,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004926964847179225,
      "loss": 5.4474,
      "mean_token_accuracy": 0.1788071185350418,
      "num_tokens": 18217434.0,
      "step": 8160
    },
    {
      "entropy": 5.548876333236694,
      "epoch": 0.9170550906946706,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004926863349409006,
      "loss": 5.4443,
      "mean_token_accuracy": 0.1806663379073143,
      "num_tokens": 18229277.0,
      "step": 8165
    },
    {
      "entropy": 5.515634250640869,
      "epoch": 0.9176166676026282,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004926761782326613,
      "loss": 5.3151,
      "mean_token_accuracy": 0.18781447410583496,
      "num_tokens": 18240320.0,
      "step": 8170
    },
    {
      "entropy": 5.441443252563476,
      "epoch": 0.9181782445105857,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004926660145935282,
      "loss": 5.3575,
      "mean_token_accuracy": 0.18915175497531891,
      "num_tokens": 18250380.0,
      "step": 8175
    },
    {
      "entropy": 5.56321530342102,
      "epoch": 0.9187398214185433,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004926558440238247,
      "loss": 5.5333,
      "mean_token_accuracy": 0.18192991465330124,
      "num_tokens": 18261703.0,
      "step": 8180
    },
    {
      "entropy": 5.609410524368286,
      "epoch": 0.9193013983265008,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004926456665238746,
      "loss": 5.4586,
      "mean_token_accuracy": 0.17719413638114928,
      "num_tokens": 18272472.0,
      "step": 8185
    },
    {
      "entropy": 5.6421808242797855,
      "epoch": 0.9198629752344584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004926354820940023,
      "loss": 5.5459,
      "mean_token_accuracy": 0.1825111448764801,
      "num_tokens": 18282168.0,
      "step": 8190
    },
    {
      "entropy": 5.6784868240356445,
      "epoch": 0.9204245521424159,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004926252907345317,
      "loss": 5.5336,
      "mean_token_accuracy": 0.1746810644865036,
      "num_tokens": 18293750.0,
      "step": 8195
    },
    {
      "entropy": 5.549223518371582,
      "epoch": 0.9209861290503735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004926150924457874,
      "loss": 5.4642,
      "mean_token_accuracy": 0.18377346247434617,
      "num_tokens": 18304568.0,
      "step": 8200
    },
    {
      "entropy": 5.450313091278076,
      "epoch": 0.921547705958331,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004926048872280943,
      "loss": 5.3869,
      "mean_token_accuracy": 0.18929603397846223,
      "num_tokens": 18315129.0,
      "step": 8205
    },
    {
      "entropy": 5.4483380794525145,
      "epoch": 0.9221092828662886,
      "grad_norm": 1.171875,
      "learning_rate": 0.000492594675081777,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18329922556877137,
      "num_tokens": 18324628.0,
      "step": 8210
    },
    {
      "entropy": 5.434116172790527,
      "epoch": 0.9226708597742461,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492584456007161,
      "loss": 5.3432,
      "mean_token_accuracy": 0.19294414818286895,
      "num_tokens": 18334899.0,
      "step": 8215
    },
    {
      "entropy": 5.5739617347717285,
      "epoch": 0.9232324366822037,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004925742300045716,
      "loss": 5.3752,
      "mean_token_accuracy": 0.18926191926002503,
      "num_tokens": 18345719.0,
      "step": 8220
    },
    {
      "entropy": 5.55813455581665,
      "epoch": 0.9237940135901612,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004925639970743341,
      "loss": 5.4471,
      "mean_token_accuracy": 0.18387410789728165,
      "num_tokens": 18355814.0,
      "step": 8225
    },
    {
      "entropy": 5.497466611862182,
      "epoch": 0.9243555904981187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004925537572167746,
      "loss": 5.5283,
      "mean_token_accuracy": 0.17901033759117127,
      "num_tokens": 18366746.0,
      "step": 8230
    },
    {
      "entropy": 5.48896803855896,
      "epoch": 0.9249171674060762,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004925435104322192,
      "loss": 5.3314,
      "mean_token_accuracy": 0.19519009441137314,
      "num_tokens": 18375846.0,
      "step": 8235
    },
    {
      "entropy": 5.471853160858155,
      "epoch": 0.9254787443140338,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000492533256720994,
      "loss": 5.4771,
      "mean_token_accuracy": 0.1840257540345192,
      "num_tokens": 18386850.0,
      "step": 8240
    },
    {
      "entropy": 5.472069358825683,
      "epoch": 0.9260403212219913,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004925229960834256,
      "loss": 5.3599,
      "mean_token_accuracy": 0.19172636568546295,
      "num_tokens": 18396347.0,
      "step": 8245
    },
    {
      "entropy": 5.517929315567017,
      "epoch": 0.9266018981299489,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004925127285198407,
      "loss": 5.3878,
      "mean_token_accuracy": 0.1863281473517418,
      "num_tokens": 18406583.0,
      "step": 8250
    },
    {
      "entropy": 5.633387565612793,
      "epoch": 0.9271634750379064,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000492502454030566,
      "loss": 5.5094,
      "mean_token_accuracy": 0.17560975253582,
      "num_tokens": 18418586.0,
      "step": 8255
    },
    {
      "entropy": 5.511312246322632,
      "epoch": 0.927725051945864,
      "grad_norm": 1.125,
      "learning_rate": 0.0004924921726159288,
      "loss": 5.3554,
      "mean_token_accuracy": 0.18426650762557983,
      "num_tokens": 18429752.0,
      "step": 8260
    },
    {
      "entropy": 5.55806188583374,
      "epoch": 0.9282866288538215,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004924818842762565,
      "loss": 5.4477,
      "mean_token_accuracy": 0.18170583993196487,
      "num_tokens": 18440663.0,
      "step": 8265
    },
    {
      "entropy": 5.453932857513427,
      "epoch": 0.9288482057617791,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004924715890118768,
      "loss": 5.4153,
      "mean_token_accuracy": 0.1861587256193161,
      "num_tokens": 18451497.0,
      "step": 8270
    },
    {
      "entropy": 5.518650197982788,
      "epoch": 0.9294097826697366,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004924612868231173,
      "loss": 5.3821,
      "mean_token_accuracy": 0.17913926094770433,
      "num_tokens": 18462794.0,
      "step": 8275
    },
    {
      "entropy": 5.5331353664398195,
      "epoch": 0.9299713595776942,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004924509777103059,
      "loss": 5.3812,
      "mean_token_accuracy": 0.18929198235273362,
      "num_tokens": 18472973.0,
      "step": 8280
    },
    {
      "entropy": 5.540349960327148,
      "epoch": 0.9305329364856517,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004924406616737712,
      "loss": 5.5451,
      "mean_token_accuracy": 0.18369444459676743,
      "num_tokens": 18484451.0,
      "step": 8285
    },
    {
      "entropy": 5.562240505218506,
      "epoch": 0.9310945133936093,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004924303387138415,
      "loss": 5.3996,
      "mean_token_accuracy": 0.1911138489842415,
      "num_tokens": 18496412.0,
      "step": 8290
    },
    {
      "entropy": 5.535116529464721,
      "epoch": 0.9316560903015668,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004924200088308455,
      "loss": 5.4378,
      "mean_token_accuracy": 0.1902432769536972,
      "num_tokens": 18508547.0,
      "step": 8295
    },
    {
      "entropy": 5.476241827011108,
      "epoch": 0.9322176672095244,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000492409672025112,
      "loss": 5.4702,
      "mean_token_accuracy": 0.18140693455934526,
      "num_tokens": 18518954.0,
      "step": 8300
    },
    {
      "entropy": 5.4479576587677006,
      "epoch": 0.9327792441174819,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004923993282969703,
      "loss": 5.3187,
      "mean_token_accuracy": 0.18697276413440705,
      "num_tokens": 18530444.0,
      "step": 8305
    },
    {
      "entropy": 5.557790231704712,
      "epoch": 0.9333408210254395,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004923889776467495,
      "loss": 5.4001,
      "mean_token_accuracy": 0.18215180486440657,
      "num_tokens": 18541612.0,
      "step": 8310
    },
    {
      "entropy": 5.56995735168457,
      "epoch": 0.933902397933397,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004923786200747795,
      "loss": 5.5528,
      "mean_token_accuracy": 0.17971559017896652,
      "num_tokens": 18553316.0,
      "step": 8315
    },
    {
      "entropy": 5.515476369857788,
      "epoch": 0.9344639748413546,
      "grad_norm": 1.484375,
      "learning_rate": 0.00049236825558139,
      "loss": 5.4118,
      "mean_token_accuracy": 0.18572491854429246,
      "num_tokens": 18564468.0,
      "step": 8320
    },
    {
      "entropy": 5.425571727752685,
      "epoch": 0.9350255517493121,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004923578841669108,
      "loss": 5.3503,
      "mean_token_accuracy": 0.18744342923164367,
      "num_tokens": 18575536.0,
      "step": 8325
    },
    {
      "entropy": 5.511794376373291,
      "epoch": 0.9355871286572696,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004923475058316723,
      "loss": 5.4808,
      "mean_token_accuracy": 0.17326518446207045,
      "num_tokens": 18586906.0,
      "step": 8330
    },
    {
      "entropy": 5.679193449020386,
      "epoch": 0.9361487055652271,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004923371205760049,
      "loss": 5.4678,
      "mean_token_accuracy": 0.18397717624902726,
      "num_tokens": 18599064.0,
      "step": 8335
    },
    {
      "entropy": 5.540845155715942,
      "epoch": 0.9367102824731847,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004923267284002394,
      "loss": 5.4909,
      "mean_token_accuracy": 0.18058904707431794,
      "num_tokens": 18610057.0,
      "step": 8340
    },
    {
      "entropy": 5.458007192611694,
      "epoch": 0.9372718593811422,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004923163293047065,
      "loss": 5.3221,
      "mean_token_accuracy": 0.18888463824987411,
      "num_tokens": 18620008.0,
      "step": 8345
    },
    {
      "entropy": 5.4577137470245365,
      "epoch": 0.9378334362890998,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004923059232897374,
      "loss": 5.322,
      "mean_token_accuracy": 0.1848251387476921,
      "num_tokens": 18630599.0,
      "step": 8350
    },
    {
      "entropy": 5.483434104919434,
      "epoch": 0.9383950131970573,
      "grad_norm": 1.6953125,
      "learning_rate": 0.0004922955103556633,
      "loss": 5.5001,
      "mean_token_accuracy": 0.18597798347473143,
      "num_tokens": 18643898.0,
      "step": 8355
    },
    {
      "entropy": 5.62312502861023,
      "epoch": 0.9389565901050149,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000492285090502816,
      "loss": 5.4854,
      "mean_token_accuracy": 0.17609167098999023,
      "num_tokens": 18654760.0,
      "step": 8360
    },
    {
      "entropy": 5.58647723197937,
      "epoch": 0.9395181670129724,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004922746637315272,
      "loss": 5.4479,
      "mean_token_accuracy": 0.1818562254309654,
      "num_tokens": 18666036.0,
      "step": 8365
    },
    {
      "entropy": 5.551207590103149,
      "epoch": 0.94007974392093,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004922642300421288,
      "loss": 5.4886,
      "mean_token_accuracy": 0.18060690760612488,
      "num_tokens": 18677806.0,
      "step": 8370
    },
    {
      "entropy": 5.606098413467407,
      "epoch": 0.9406413208288875,
      "grad_norm": 1.078125,
      "learning_rate": 0.000492253789434953,
      "loss": 5.6054,
      "mean_token_accuracy": 0.17770332247018814,
      "num_tokens": 18688971.0,
      "step": 8375
    },
    {
      "entropy": 5.572735691070557,
      "epoch": 0.9412028977368451,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004922433419103323,
      "loss": 5.4755,
      "mean_token_accuracy": 0.1787707731127739,
      "num_tokens": 18700153.0,
      "step": 8380
    },
    {
      "entropy": 5.657432889938354,
      "epoch": 0.9417644746448026,
      "grad_norm": 1.125,
      "learning_rate": 0.0004922328874685993,
      "loss": 5.5224,
      "mean_token_accuracy": 0.1776360183954239,
      "num_tokens": 18711935.0,
      "step": 8385
    },
    {
      "entropy": 5.415850734710693,
      "epoch": 0.9423260515527602,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004922224261100869,
      "loss": 5.3591,
      "mean_token_accuracy": 0.1885710507631302,
      "num_tokens": 18722694.0,
      "step": 8390
    },
    {
      "entropy": 5.554447746276855,
      "epoch": 0.9428876284607177,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004922119578351283,
      "loss": 5.508,
      "mean_token_accuracy": 0.18703244179487227,
      "num_tokens": 18733443.0,
      "step": 8395
    },
    {
      "entropy": 5.595880460739136,
      "epoch": 0.9434492053686753,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004922014826440566,
      "loss": 5.4167,
      "mean_token_accuracy": 0.18218420296907425,
      "num_tokens": 18743965.0,
      "step": 8400
    },
    {
      "entropy": 5.532165384292602,
      "epoch": 0.9440107822766328,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004921910005372056,
      "loss": 5.5046,
      "mean_token_accuracy": 0.1802481397986412,
      "num_tokens": 18755712.0,
      "step": 8405
    },
    {
      "entropy": 5.552771329879761,
      "epoch": 0.9445723591845904,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004921805115149088,
      "loss": 5.4193,
      "mean_token_accuracy": 0.18201478719711303,
      "num_tokens": 18766756.0,
      "step": 8410
    },
    {
      "entropy": 5.550810194015503,
      "epoch": 0.9451339360925479,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004921700155775002,
      "loss": 5.5178,
      "mean_token_accuracy": 0.17482238113880158,
      "num_tokens": 18776932.0,
      "step": 8415
    },
    {
      "entropy": 5.616715955734253,
      "epoch": 0.9456955130005055,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004921595127253142,
      "loss": 5.4323,
      "mean_token_accuracy": 0.18388253301382065,
      "num_tokens": 18787197.0,
      "step": 8420
    },
    {
      "entropy": 5.466071605682373,
      "epoch": 0.9462570899084629,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004921490029586852,
      "loss": 5.4013,
      "mean_token_accuracy": 0.1836030140519142,
      "num_tokens": 18797861.0,
      "step": 8425
    },
    {
      "entropy": 5.503315782546997,
      "epoch": 0.9468186668164205,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004921384862779475,
      "loss": 5.3723,
      "mean_token_accuracy": 0.18375704437494278,
      "num_tokens": 18809111.0,
      "step": 8430
    },
    {
      "entropy": 5.498533391952515,
      "epoch": 0.947380243724378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004921279626834363,
      "loss": 5.417,
      "mean_token_accuracy": 0.19604480713605882,
      "num_tokens": 18820236.0,
      "step": 8435
    },
    {
      "entropy": 5.475078010559082,
      "epoch": 0.9479418206323356,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004921174321754865,
      "loss": 5.3997,
      "mean_token_accuracy": 0.18670594841241836,
      "num_tokens": 18831942.0,
      "step": 8440
    },
    {
      "entropy": 5.543049955368042,
      "epoch": 0.9485033975402931,
      "grad_norm": 1.125,
      "learning_rate": 0.0004921068947544335,
      "loss": 5.4277,
      "mean_token_accuracy": 0.18354681134223938,
      "num_tokens": 18843271.0,
      "step": 8445
    },
    {
      "entropy": 5.486358690261841,
      "epoch": 0.9490649744482507,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004920963504206128,
      "loss": 5.4754,
      "mean_token_accuracy": 0.18712944388389588,
      "num_tokens": 18856285.0,
      "step": 8450
    },
    {
      "entropy": 5.600731706619262,
      "epoch": 0.9496265513562082,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00049208579917436,
      "loss": 5.4861,
      "mean_token_accuracy": 0.18112098723649978,
      "num_tokens": 18867203.0,
      "step": 8455
    },
    {
      "entropy": 5.590496826171875,
      "epoch": 0.9501881282641658,
      "grad_norm": 1.125,
      "learning_rate": 0.0004920752410160113,
      "loss": 5.5516,
      "mean_token_accuracy": 0.17466073781251906,
      "num_tokens": 18878598.0,
      "step": 8460
    },
    {
      "entropy": 5.609384059906006,
      "epoch": 0.9507497051721233,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004920646759459026,
      "loss": 5.441,
      "mean_token_accuracy": 0.18235270082950591,
      "num_tokens": 18890233.0,
      "step": 8465
    },
    {
      "entropy": 5.618962192535401,
      "epoch": 0.9513112820800809,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004920541039643705,
      "loss": 5.5332,
      "mean_token_accuracy": 0.17830730676651002,
      "num_tokens": 18900362.0,
      "step": 8470
    },
    {
      "entropy": 5.547064304351807,
      "epoch": 0.9518728589880384,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004920435250717515,
      "loss": 5.4501,
      "mean_token_accuracy": 0.18904222100973128,
      "num_tokens": 18911924.0,
      "step": 8475
    },
    {
      "entropy": 5.547458219528198,
      "epoch": 0.952434435895996,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004920329392683826,
      "loss": 5.4669,
      "mean_token_accuracy": 0.18137711137533188,
      "num_tokens": 18923931.0,
      "step": 8480
    },
    {
      "entropy": 5.546305179595947,
      "epoch": 0.9529960128039535,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004920223465546006,
      "loss": 5.3884,
      "mean_token_accuracy": 0.18430516868829727,
      "num_tokens": 18935859.0,
      "step": 8485
    },
    {
      "entropy": 5.540053749084473,
      "epoch": 0.9535575897119111,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004920117469307429,
      "loss": 5.5177,
      "mean_token_accuracy": 0.17664810568094252,
      "num_tokens": 18946939.0,
      "step": 8490
    },
    {
      "entropy": 5.587956333160401,
      "epoch": 0.9541191666198686,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004920011403971471,
      "loss": 5.4933,
      "mean_token_accuracy": 0.17564062625169755,
      "num_tokens": 18958067.0,
      "step": 8495
    },
    {
      "entropy": 5.549686574935913,
      "epoch": 0.9546807435278262,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004919905269541508,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18971737772226333,
      "num_tokens": 18967920.0,
      "step": 8500
    },
    {
      "entropy": 5.5899028301239015,
      "epoch": 0.9552423204357837,
      "grad_norm": 1.109375,
      "learning_rate": 0.000491979906602092,
      "loss": 5.4289,
      "mean_token_accuracy": 0.1890067756175995,
      "num_tokens": 18978244.0,
      "step": 8505
    },
    {
      "entropy": 5.550780010223389,
      "epoch": 0.9558038973437413,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004919692793413089,
      "loss": 5.383,
      "mean_token_accuracy": 0.1837337002158165,
      "num_tokens": 18988464.0,
      "step": 8510
    },
    {
      "entropy": 5.4218223094940186,
      "epoch": 0.9563654742516988,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004919586451721396,
      "loss": 5.3548,
      "mean_token_accuracy": 0.18404873907566072,
      "num_tokens": 18998318.0,
      "step": 8515
    },
    {
      "entropy": 5.404018306732178,
      "epoch": 0.9569270511596563,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004919480040949229,
      "loss": 5.3307,
      "mean_token_accuracy": 0.19233920872211457,
      "num_tokens": 19009802.0,
      "step": 8520
    },
    {
      "entropy": 5.656398439407349,
      "epoch": 0.9574886280676138,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004919373561099977,
      "loss": 5.5406,
      "mean_token_accuracy": 0.1823028177022934,
      "num_tokens": 19020174.0,
      "step": 8525
    },
    {
      "entropy": 5.59047417640686,
      "epoch": 0.9580502049755714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000491926701217703,
      "loss": 5.4738,
      "mean_token_accuracy": 0.17923413962125778,
      "num_tokens": 19031505.0,
      "step": 8530
    },
    {
      "entropy": 5.517679071426391,
      "epoch": 0.9586117818835289,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004919160394183779,
      "loss": 5.3809,
      "mean_token_accuracy": 0.18628357201814652,
      "num_tokens": 19042052.0,
      "step": 8535
    },
    {
      "entropy": 5.694993543624878,
      "epoch": 0.9591733587914865,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000491905370712362,
      "loss": 5.5881,
      "mean_token_accuracy": 0.17614448070526123,
      "num_tokens": 19052050.0,
      "step": 8540
    },
    {
      "entropy": 5.620108222961425,
      "epoch": 0.959734935699444,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004918946950999949,
      "loss": 5.4872,
      "mean_token_accuracy": 0.18063030987977982,
      "num_tokens": 19062053.0,
      "step": 8545
    },
    {
      "entropy": 5.479205656051636,
      "epoch": 0.9602965126074016,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004918840125816167,
      "loss": 5.4437,
      "mean_token_accuracy": 0.18385787606239318,
      "num_tokens": 19073095.0,
      "step": 8550
    },
    {
      "entropy": 5.556223583221436,
      "epoch": 0.9608580895153591,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004918733231575673,
      "loss": 5.4838,
      "mean_token_accuracy": 0.18409732580184937,
      "num_tokens": 19084521.0,
      "step": 8555
    },
    {
      "entropy": 5.590706491470337,
      "epoch": 0.9614196664233167,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004918626268281874,
      "loss": 5.4795,
      "mean_token_accuracy": 0.18052347451448442,
      "num_tokens": 19096372.0,
      "step": 8560
    },
    {
      "entropy": 5.652284622192383,
      "epoch": 0.9619812433312742,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004918519235938171,
      "loss": 5.5808,
      "mean_token_accuracy": 0.17533139437437056,
      "num_tokens": 19108535.0,
      "step": 8565
    },
    {
      "entropy": 5.634532070159912,
      "epoch": 0.9625428202392318,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004918412134547976,
      "loss": 5.4455,
      "mean_token_accuracy": 0.18690463453531264,
      "num_tokens": 19119480.0,
      "step": 8570
    },
    {
      "entropy": 5.522903394699097,
      "epoch": 0.9631043971471893,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004918304964114696,
      "loss": 5.4405,
      "mean_token_accuracy": 0.18453758656978608,
      "num_tokens": 19131168.0,
      "step": 8575
    },
    {
      "entropy": 5.567117691040039,
      "epoch": 0.9636659740551469,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004918197724641747,
      "loss": 5.5818,
      "mean_token_accuracy": 0.17886256873607637,
      "num_tokens": 19142454.0,
      "step": 8580
    },
    {
      "entropy": 5.613681936264038,
      "epoch": 0.9642275509631044,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000491809041613254,
      "loss": 5.4433,
      "mean_token_accuracy": 0.18341798186302186,
      "num_tokens": 19154682.0,
      "step": 8585
    },
    {
      "entropy": 5.4962602138519285,
      "epoch": 0.964789127871062,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004917983038590495,
      "loss": 5.3728,
      "mean_token_accuracy": 0.1835367724299431,
      "num_tokens": 19165190.0,
      "step": 8590
    },
    {
      "entropy": 5.613046789169312,
      "epoch": 0.9653507047790195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004917875592019028,
      "loss": 5.5105,
      "mean_token_accuracy": 0.18841075152158737,
      "num_tokens": 19176075.0,
      "step": 8595
    },
    {
      "entropy": 5.588446235656738,
      "epoch": 0.9659122816869771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004917768076421561,
      "loss": 5.5057,
      "mean_token_accuracy": 0.18028940707445146,
      "num_tokens": 19188450.0,
      "step": 8600
    },
    {
      "entropy": 5.416703701019287,
      "epoch": 0.9664738585949346,
      "grad_norm": 1.0,
      "learning_rate": 0.0004917660491801518,
      "loss": 5.3764,
      "mean_token_accuracy": 0.19176003336906433,
      "num_tokens": 19201254.0,
      "step": 8605
    },
    {
      "entropy": 5.478834772109986,
      "epoch": 0.9670354355028922,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004917552838162324,
      "loss": 5.4513,
      "mean_token_accuracy": 0.18401146531105042,
      "num_tokens": 19214062.0,
      "step": 8610
    },
    {
      "entropy": 5.524210739135742,
      "epoch": 0.9675970124108496,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004917445115507408,
      "loss": 5.3551,
      "mean_token_accuracy": 0.19021136909723282,
      "num_tokens": 19224844.0,
      "step": 8615
    },
    {
      "entropy": 5.542030286788941,
      "epoch": 0.9681585893188072,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004917337323840198,
      "loss": 5.5333,
      "mean_token_accuracy": 0.17929414361715318,
      "num_tokens": 19237347.0,
      "step": 8620
    },
    {
      "entropy": 5.600201034545899,
      "epoch": 0.9687201662267647,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004917229463164128,
      "loss": 5.4053,
      "mean_token_accuracy": 0.1853177845478058,
      "num_tokens": 19248743.0,
      "step": 8625
    },
    {
      "entropy": 5.538897275924683,
      "epoch": 0.9692817431347223,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004917121533482631,
      "loss": 5.3957,
      "mean_token_accuracy": 0.1866521954536438,
      "num_tokens": 19259904.0,
      "step": 8630
    },
    {
      "entropy": 5.433857393264771,
      "epoch": 0.9698433200426798,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004917013534799144,
      "loss": 5.3409,
      "mean_token_accuracy": 0.18779617547988892,
      "num_tokens": 19271388.0,
      "step": 8635
    },
    {
      "entropy": 5.528453731536866,
      "epoch": 0.9704048969506374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004916905467117104,
      "loss": 5.4779,
      "mean_token_accuracy": 0.1867511287331581,
      "num_tokens": 19282790.0,
      "step": 8640
    },
    {
      "entropy": 5.58134126663208,
      "epoch": 0.9709664738585949,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004916797330439955,
      "loss": 5.3818,
      "mean_token_accuracy": 0.1863004371523857,
      "num_tokens": 19293389.0,
      "step": 8645
    },
    {
      "entropy": 5.457135152816773,
      "epoch": 0.9715280507665525,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004916689124771138,
      "loss": 5.4519,
      "mean_token_accuracy": 0.18248879462480544,
      "num_tokens": 19304593.0,
      "step": 8650
    },
    {
      "entropy": 5.495477962493896,
      "epoch": 0.97208962767451,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004916580850114099,
      "loss": 5.3789,
      "mean_token_accuracy": 0.18114386200904847,
      "num_tokens": 19317388.0,
      "step": 8655
    },
    {
      "entropy": 5.583202219009399,
      "epoch": 0.9726512045824676,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004916472506472286,
      "loss": 5.4334,
      "mean_token_accuracy": 0.1869949772953987,
      "num_tokens": 19329228.0,
      "step": 8660
    },
    {
      "entropy": 5.5602278232574465,
      "epoch": 0.9732127814904251,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004916364093849147,
      "loss": 5.4054,
      "mean_token_accuracy": 0.1909436896443367,
      "num_tokens": 19339421.0,
      "step": 8665
    },
    {
      "entropy": 5.408206367492676,
      "epoch": 0.9737743583983827,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004916255612248136,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19228839129209518,
      "num_tokens": 19349618.0,
      "step": 8670
    },
    {
      "entropy": 5.375762796401977,
      "epoch": 0.9743359353063402,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004916147061672706,
      "loss": 5.3996,
      "mean_token_accuracy": 0.18693719059228897,
      "num_tokens": 19361324.0,
      "step": 8675
    },
    {
      "entropy": 5.607862186431885,
      "epoch": 0.9748975122142978,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004916038442126314,
      "loss": 5.4265,
      "mean_token_accuracy": 0.18402668535709382,
      "num_tokens": 19373631.0,
      "step": 8680
    },
    {
      "entropy": 5.447333478927613,
      "epoch": 0.9754590891222553,
      "grad_norm": 1.25,
      "learning_rate": 0.0004915929753612416,
      "loss": 5.3769,
      "mean_token_accuracy": 0.17710047960281372,
      "num_tokens": 19384682.0,
      "step": 8685
    },
    {
      "entropy": 5.544682550430298,
      "epoch": 0.9760206660302129,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004915820996134477,
      "loss": 5.4828,
      "mean_token_accuracy": 0.17632165998220445,
      "num_tokens": 19396344.0,
      "step": 8690
    },
    {
      "entropy": 5.599015569686889,
      "epoch": 0.9765822429381704,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004915712169695956,
      "loss": 5.3418,
      "mean_token_accuracy": 0.18669420331716538,
      "num_tokens": 19406921.0,
      "step": 8695
    },
    {
      "entropy": 5.471725845336914,
      "epoch": 0.977143819846128,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000491560327430032,
      "loss": 5.4649,
      "mean_token_accuracy": 0.18232473880052566,
      "num_tokens": 19416851.0,
      "step": 8700
    },
    {
      "entropy": 5.4985700130462645,
      "epoch": 0.9777053967540855,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004915494309951035,
      "loss": 5.4317,
      "mean_token_accuracy": 0.18158532083034515,
      "num_tokens": 19428180.0,
      "step": 8705
    },
    {
      "entropy": 5.663486528396606,
      "epoch": 0.9782669736620431,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004915385276651573,
      "loss": 5.479,
      "mean_token_accuracy": 0.18338771611452104,
      "num_tokens": 19439575.0,
      "step": 8710
    },
    {
      "entropy": 5.574627351760864,
      "epoch": 0.9788285505700005,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004915276174405403,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18291618674993515,
      "num_tokens": 19452074.0,
      "step": 8715
    },
    {
      "entropy": 5.454778337478638,
      "epoch": 0.9793901274779581,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004915167003215999,
      "loss": 5.3752,
      "mean_token_accuracy": 0.18522823452949524,
      "num_tokens": 19463085.0,
      "step": 8720
    },
    {
      "entropy": 5.651071643829345,
      "epoch": 0.9799517043859156,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004915057763086838,
      "loss": 5.5503,
      "mean_token_accuracy": 0.17118286192417145,
      "num_tokens": 19472726.0,
      "step": 8725
    },
    {
      "entropy": 5.4387369632720945,
      "epoch": 0.9805132812938732,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00049149484540214,
      "loss": 5.3765,
      "mean_token_accuracy": 0.18026935309171677,
      "num_tokens": 19482275.0,
      "step": 8730
    },
    {
      "entropy": 5.512332773208618,
      "epoch": 0.9810748582018307,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004914839076023161,
      "loss": 5.4713,
      "mean_token_accuracy": 0.18529911190271378,
      "num_tokens": 19492782.0,
      "step": 8735
    },
    {
      "entropy": 5.477008295059204,
      "epoch": 0.9816364351097883,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004914729629095609,
      "loss": 5.3551,
      "mean_token_accuracy": 0.1842048406600952,
      "num_tokens": 19505709.0,
      "step": 8740
    },
    {
      "entropy": 5.523062419891358,
      "epoch": 0.9821980120177458,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004914620113242224,
      "loss": 5.4241,
      "mean_token_accuracy": 0.18395508080720901,
      "num_tokens": 19516269.0,
      "step": 8745
    },
    {
      "entropy": 5.522498226165771,
      "epoch": 0.9827595889257034,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004914510528466495,
      "loss": 5.4216,
      "mean_token_accuracy": 0.1846705585718155,
      "num_tokens": 19526786.0,
      "step": 8750
    },
    {
      "entropy": 5.536771297454834,
      "epoch": 0.9833211658336609,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004914400874771913,
      "loss": 5.446,
      "mean_token_accuracy": 0.18200644552707673,
      "num_tokens": 19537865.0,
      "step": 8755
    },
    {
      "entropy": 5.4594237327575685,
      "epoch": 0.9838827427416185,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004914291152161967,
      "loss": 5.3812,
      "mean_token_accuracy": 0.19089686423540114,
      "num_tokens": 19548320.0,
      "step": 8760
    },
    {
      "entropy": 5.484426355361938,
      "epoch": 0.984444319649576,
      "grad_norm": 1.28125,
      "learning_rate": 0.000491418136064015,
      "loss": 5.3872,
      "mean_token_accuracy": 0.18776844888925553,
      "num_tokens": 19558856.0,
      "step": 8765
    },
    {
      "entropy": 5.499591302871704,
      "epoch": 0.9850058965575336,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000491407150020996,
      "loss": 5.4749,
      "mean_token_accuracy": 0.18665503710508347,
      "num_tokens": 19570792.0,
      "step": 8770
    },
    {
      "entropy": 5.54670147895813,
      "epoch": 0.9855674734654911,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004913961570874895,
      "loss": 5.3903,
      "mean_token_accuracy": 0.19058968722820283,
      "num_tokens": 19581982.0,
      "step": 8775
    },
    {
      "entropy": 5.506307983398438,
      "epoch": 0.9861290503734487,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004913851572638453,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18622274994850158,
      "num_tokens": 19592882.0,
      "step": 8780
    },
    {
      "entropy": 5.60463809967041,
      "epoch": 0.9866906272814062,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004913741505504138,
      "loss": 5.5573,
      "mean_token_accuracy": 0.18271916806697847,
      "num_tokens": 19605967.0,
      "step": 8785
    },
    {
      "entropy": 5.59593358039856,
      "epoch": 0.9872522041893638,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004913631369475454,
      "loss": 5.3979,
      "mean_token_accuracy": 0.18678078055381775,
      "num_tokens": 19616494.0,
      "step": 8790
    },
    {
      "entropy": 5.5236421585083,
      "epoch": 0.9878137810973213,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004913521164555909,
      "loss": 5.519,
      "mean_token_accuracy": 0.1811422035098076,
      "num_tokens": 19627013.0,
      "step": 8795
    },
    {
      "entropy": 5.517453908920288,
      "epoch": 0.9883753580052789,
      "grad_norm": 1.109375,
      "learning_rate": 0.000491341089074901,
      "loss": 5.4132,
      "mean_token_accuracy": 0.18959666192531585,
      "num_tokens": 19639099.0,
      "step": 8800
    },
    {
      "entropy": 5.465067911148071,
      "epoch": 0.9889369349132364,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000491330054805827,
      "loss": 5.3814,
      "mean_token_accuracy": 0.18871617764234544,
      "num_tokens": 19650650.0,
      "step": 8805
    },
    {
      "entropy": 5.557087993621826,
      "epoch": 0.9894985118211939,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004913190136487201,
      "loss": 5.4169,
      "mean_token_accuracy": 0.1847272425889969,
      "num_tokens": 19662869.0,
      "step": 8810
    },
    {
      "entropy": 5.520923185348511,
      "epoch": 0.9900600887291514,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004913079656039318,
      "loss": 5.4235,
      "mean_token_accuracy": 0.18705991208553313,
      "num_tokens": 19673232.0,
      "step": 8815
    },
    {
      "entropy": 5.501662254333496,
      "epoch": 0.990621665637109,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004912969106718142,
      "loss": 5.4434,
      "mean_token_accuracy": 0.18457782119512559,
      "num_tokens": 19685069.0,
      "step": 8820
    },
    {
      "entropy": 5.602150630950928,
      "epoch": 0.9911832425450665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000491285848852719,
      "loss": 5.5332,
      "mean_token_accuracy": 0.1780464842915535,
      "num_tokens": 19697018.0,
      "step": 8825
    },
    {
      "entropy": 5.547738885879516,
      "epoch": 0.9917448194530241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004912747801469984,
      "loss": 5.395,
      "mean_token_accuracy": 0.1881812572479248,
      "num_tokens": 19707922.0,
      "step": 8830
    },
    {
      "entropy": 5.418731737136841,
      "epoch": 0.9923063963609816,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000491263704555005,
      "loss": 5.3322,
      "mean_token_accuracy": 0.18567853271961213,
      "num_tokens": 19718693.0,
      "step": 8835
    },
    {
      "entropy": 5.5052399158477785,
      "epoch": 0.9928679732689392,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004912526220770913,
      "loss": 5.4055,
      "mean_token_accuracy": 0.18160765320062638,
      "num_tokens": 19730846.0,
      "step": 8840
    },
    {
      "entropy": 5.508100175857544,
      "epoch": 0.9934295501768967,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004912415327136103,
      "loss": 5.4222,
      "mean_token_accuracy": 0.18601943999528886,
      "num_tokens": 19742099.0,
      "step": 8845
    },
    {
      "entropy": 5.559634733200073,
      "epoch": 0.9939911270848543,
      "grad_norm": 1.109375,
      "learning_rate": 0.000491230436464915,
      "loss": 5.4315,
      "mean_token_accuracy": 0.184655299782753,
      "num_tokens": 19753216.0,
      "step": 8850
    },
    {
      "entropy": 5.549345874786377,
      "epoch": 0.9945527039928118,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004912193333313589,
      "loss": 5.3639,
      "mean_token_accuracy": 0.18711148649454118,
      "num_tokens": 19765316.0,
      "step": 8855
    },
    {
      "entropy": 5.501159429550171,
      "epoch": 0.9951142809007694,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004912082233132951,
      "loss": 5.3983,
      "mean_token_accuracy": 0.18220563977956772,
      "num_tokens": 19776208.0,
      "step": 8860
    },
    {
      "entropy": 5.555660915374756,
      "epoch": 0.9956758578087269,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004911971064110778,
      "loss": 5.4362,
      "mean_token_accuracy": 0.18303729295730592,
      "num_tokens": 19788065.0,
      "step": 8865
    },
    {
      "entropy": 5.489232730865479,
      "epoch": 0.9962374347166845,
      "grad_norm": 1.125,
      "learning_rate": 0.0004911859826250607,
      "loss": 5.3615,
      "mean_token_accuracy": 0.182569220662117,
      "num_tokens": 19799388.0,
      "step": 8870
    },
    {
      "entropy": 5.489706516265869,
      "epoch": 0.996799011624642,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004911748519555981,
      "loss": 5.4569,
      "mean_token_accuracy": 0.18266495019197465,
      "num_tokens": 19809933.0,
      "step": 8875
    },
    {
      "entropy": 5.696295070648193,
      "epoch": 0.9973605885325996,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004911637144030444,
      "loss": 5.5335,
      "mean_token_accuracy": 0.1796165332198143,
      "num_tokens": 19821566.0,
      "step": 8880
    },
    {
      "entropy": 5.631623840332031,
      "epoch": 0.9979221654405571,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004911525699677541,
      "loss": 5.4976,
      "mean_token_accuracy": 0.18242679089307784,
      "num_tokens": 19833727.0,
      "step": 8885
    },
    {
      "entropy": 5.554462242126465,
      "epoch": 0.9984837423485147,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004911414186500822,
      "loss": 5.5088,
      "mean_token_accuracy": 0.1791031152009964,
      "num_tokens": 19845176.0,
      "step": 8890
    },
    {
      "entropy": 5.506851100921631,
      "epoch": 0.9990453192564722,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004911302604503837,
      "loss": 5.4025,
      "mean_token_accuracy": 0.17646405696868897,
      "num_tokens": 19857528.0,
      "step": 8895
    },
    {
      "entropy": 5.525443267822266,
      "epoch": 0.9996068961644298,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004911190953690139,
      "loss": 5.3334,
      "mean_token_accuracy": 0.18568519949913026,
      "num_tokens": 19868232.0,
      "step": 8900
    },
    {
      "entropy": 5.478812747531467,
      "epoch": 1.0001123153815914,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004911079234063284,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18990937537617153,
      "num_tokens": 19877881.0,
      "step": 8905
    },
    {
      "entropy": 5.42882513999939,
      "epoch": 1.000673892289549,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004910967445626826,
      "loss": 5.3676,
      "mean_token_accuracy": 0.19181329309940337,
      "num_tokens": 19888398.0,
      "step": 8910
    },
    {
      "entropy": 5.564544439315796,
      "epoch": 1.0012354691975065,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004910855588384327,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1882049635052681,
      "num_tokens": 19899661.0,
      "step": 8915
    },
    {
      "entropy": 5.534649467468261,
      "epoch": 1.0017970461054642,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004910743662339348,
      "loss": 5.3912,
      "mean_token_accuracy": 0.1847395807504654,
      "num_tokens": 19911073.0,
      "step": 8920
    },
    {
      "entropy": 5.431390953063965,
      "epoch": 1.0023586230134216,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004910631667495452,
      "loss": 5.3141,
      "mean_token_accuracy": 0.19309445321559907,
      "num_tokens": 19922567.0,
      "step": 8925
    },
    {
      "entropy": 5.493463850021362,
      "epoch": 1.0029201999213793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910519603856207,
      "loss": 5.4199,
      "mean_token_accuracy": 0.18911998122930526,
      "num_tokens": 19934547.0,
      "step": 8930
    },
    {
      "entropy": 5.501732873916626,
      "epoch": 1.0034817768293367,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004910407471425179,
      "loss": 5.4282,
      "mean_token_accuracy": 0.1858392558991909,
      "num_tokens": 19947659.0,
      "step": 8935
    },
    {
      "entropy": 5.559143018722534,
      "epoch": 1.0040433537372944,
      "grad_norm": 1.1875,
      "learning_rate": 0.000491029527020594,
      "loss": 5.3322,
      "mean_token_accuracy": 0.1836379736661911,
      "num_tokens": 19958150.0,
      "step": 8940
    },
    {
      "entropy": 5.444316291809082,
      "epoch": 1.0046049306452518,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004910183000202061,
      "loss": 5.3126,
      "mean_token_accuracy": 0.19097789227962494,
      "num_tokens": 19968173.0,
      "step": 8945
    },
    {
      "entropy": 5.489464473724365,
      "epoch": 1.0051665075532095,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004910070661417119,
      "loss": 5.3637,
      "mean_token_accuracy": 0.18406429141759872,
      "num_tokens": 19978543.0,
      "step": 8950
    },
    {
      "entropy": 5.551176500320435,
      "epoch": 1.005728084461167,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909958253854687,
      "loss": 5.3219,
      "mean_token_accuracy": 0.1828606143593788,
      "num_tokens": 19989586.0,
      "step": 8955
    },
    {
      "entropy": 5.460301876068115,
      "epoch": 1.0062896613691246,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909845777518347,
      "loss": 5.3388,
      "mean_token_accuracy": 0.18899851590394973,
      "num_tokens": 20000524.0,
      "step": 8960
    },
    {
      "entropy": 5.527743482589722,
      "epoch": 1.006851238277082,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909733232411681,
      "loss": 5.4334,
      "mean_token_accuracy": 0.18191745579242707,
      "num_tokens": 20012935.0,
      "step": 8965
    },
    {
      "entropy": 5.538965511322021,
      "epoch": 1.0074128151850397,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004909620618538271,
      "loss": 5.3959,
      "mean_token_accuracy": 0.18554447144269942,
      "num_tokens": 20024435.0,
      "step": 8970
    },
    {
      "entropy": 5.549155759811401,
      "epoch": 1.007974392092997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004909507935901702,
      "loss": 5.408,
      "mean_token_accuracy": 0.18195135593414308,
      "num_tokens": 20036017.0,
      "step": 8975
    },
    {
      "entropy": 5.507486057281494,
      "epoch": 1.0085359690009548,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004909395184505564,
      "loss": 5.4001,
      "mean_token_accuracy": 0.18807803690433503,
      "num_tokens": 20047216.0,
      "step": 8980
    },
    {
      "entropy": 5.480719900131225,
      "epoch": 1.0090975459089122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004909282364353445,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18794850558042525,
      "num_tokens": 20057439.0,
      "step": 8985
    },
    {
      "entropy": 5.462898302078247,
      "epoch": 1.0096591228168699,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004909169475448938,
      "loss": 5.4052,
      "mean_token_accuracy": 0.18469297289848327,
      "num_tokens": 20069602.0,
      "step": 8990
    },
    {
      "entropy": 5.4774504661560055,
      "epoch": 1.0102206997248273,
      "grad_norm": 1.125,
      "learning_rate": 0.0004909056517795638,
      "loss": 5.3546,
      "mean_token_accuracy": 0.1900850549340248,
      "num_tokens": 20081366.0,
      "step": 8995
    },
    {
      "entropy": 5.517918586730957,
      "epoch": 1.010782276632785,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004908943491397141,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18946565985679625,
      "num_tokens": 20091384.0,
      "step": 9000
    },
    {
      "epoch": 1.010782276632785,
      "eval_entropy": 5.365454944224895,
      "eval_loss": 5.472287654876709,
      "eval_mean_token_accuracy": 0.1891292570620537,
      "eval_num_tokens": 20091384.0,
      "eval_runtime": 24.0414,
      "eval_samples_per_second": 1289.983,
      "eval_steps_per_second": 161.263,
      "step": 9000
    },
    {
      "entropy": 5.547128057479858,
      "epoch": 1.0113438535407424,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004908830396257045,
      "loss": 5.3861,
      "mean_token_accuracy": 0.1802237570285797,
      "num_tokens": 20102953.0,
      "step": 9005
    },
    {
      "entropy": 5.492046546936035,
      "epoch": 1.0119054304486998,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004908717232378953,
      "loss": 5.3341,
      "mean_token_accuracy": 0.18801592588424682,
      "num_tokens": 20113672.0,
      "step": 9010
    },
    {
      "entropy": 5.560129833221436,
      "epoch": 1.0124670073566575,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004908603999766467,
      "loss": 5.3656,
      "mean_token_accuracy": 0.18842772394418716,
      "num_tokens": 20125143.0,
      "step": 9015
    },
    {
      "entropy": 5.499776458740234,
      "epoch": 1.013028584264615,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004908490698423191,
      "loss": 5.3607,
      "mean_token_accuracy": 0.1915813073515892,
      "num_tokens": 20135489.0,
      "step": 9020
    },
    {
      "entropy": 5.436945962905884,
      "epoch": 1.0135901611725726,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004908377328352736,
      "loss": 5.3514,
      "mean_token_accuracy": 0.19145939350128174,
      "num_tokens": 20146768.0,
      "step": 9025
    },
    {
      "entropy": 5.3235231876373295,
      "epoch": 1.01415173808053,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004908263889558709,
      "loss": 5.184,
      "mean_token_accuracy": 0.1914340004324913,
      "num_tokens": 20156894.0,
      "step": 9030
    },
    {
      "entropy": 5.523900127410888,
      "epoch": 1.0147133149884877,
      "grad_norm": 1.5234375,
      "learning_rate": 0.0004908150382044722,
      "loss": 5.3948,
      "mean_token_accuracy": 0.19234025925397874,
      "num_tokens": 20168174.0,
      "step": 9035
    },
    {
      "entropy": 5.582646703720092,
      "epoch": 1.0152748918964452,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000490803680581439,
      "loss": 5.4649,
      "mean_token_accuracy": 0.18047902435064317,
      "num_tokens": 20179962.0,
      "step": 9040
    },
    {
      "entropy": 5.5096814155578615,
      "epoch": 1.0158364688044028,
      "grad_norm": 1.234375,
      "learning_rate": 0.000490792316087133,
      "loss": 5.3742,
      "mean_token_accuracy": 0.18952901810407638,
      "num_tokens": 20189842.0,
      "step": 9045
    },
    {
      "entropy": 5.4693807601928714,
      "epoch": 1.0163980457123603,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004907809447219158,
      "loss": 5.2566,
      "mean_token_accuracy": 0.18900258094072342,
      "num_tokens": 20200471.0,
      "step": 9050
    },
    {
      "entropy": 5.486717605590821,
      "epoch": 1.016959622620318,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004907695664861498,
      "loss": 5.339,
      "mean_token_accuracy": 0.18232023268938063,
      "num_tokens": 20212145.0,
      "step": 9055
    },
    {
      "entropy": 5.49371109008789,
      "epoch": 1.0175211995282754,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004907581813801971,
      "loss": 5.3423,
      "mean_token_accuracy": 0.19111180752515794,
      "num_tokens": 20222454.0,
      "step": 9060
    },
    {
      "entropy": 5.599603319168091,
      "epoch": 1.018082776436233,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004907467894044202,
      "loss": 5.46,
      "mean_token_accuracy": 0.18083517849445344,
      "num_tokens": 20233402.0,
      "step": 9065
    },
    {
      "entropy": 5.455314874649048,
      "epoch": 1.0186443533441905,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004907353905591817,
      "loss": 5.2682,
      "mean_token_accuracy": 0.19526829421520234,
      "num_tokens": 20244102.0,
      "step": 9070
    },
    {
      "entropy": 5.567287111282349,
      "epoch": 1.0192059302521481,
      "grad_norm": 1.125,
      "learning_rate": 0.000490723984844845,
      "loss": 5.4841,
      "mean_token_accuracy": 0.1795387774705887,
      "num_tokens": 20255003.0,
      "step": 9075
    },
    {
      "entropy": 5.553279209136963,
      "epoch": 1.0197675071601056,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004907125722617727,
      "loss": 5.3195,
      "mean_token_accuracy": 0.18895499557256698,
      "num_tokens": 20265211.0,
      "step": 9080
    },
    {
      "entropy": 5.472066879272461,
      "epoch": 1.0203290840680632,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004907011528103285,
      "loss": 5.2629,
      "mean_token_accuracy": 0.1926484301686287,
      "num_tokens": 20276552.0,
      "step": 9085
    },
    {
      "entropy": 5.496870470046997,
      "epoch": 1.0208906609760207,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004906897264908759,
      "loss": 5.4119,
      "mean_token_accuracy": 0.18400470912456512,
      "num_tokens": 20286805.0,
      "step": 9090
    },
    {
      "entropy": 5.505714368820191,
      "epoch": 1.0214522378839783,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004906782933037789,
      "loss": 5.3561,
      "mean_token_accuracy": 0.18586109280586244,
      "num_tokens": 20298536.0,
      "step": 9095
    },
    {
      "entropy": 5.512578296661377,
      "epoch": 1.0220138147919358,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004906668532494012,
      "loss": 5.3448,
      "mean_token_accuracy": 0.1839442014694214,
      "num_tokens": 20309842.0,
      "step": 9100
    },
    {
      "entropy": 5.477510786056518,
      "epoch": 1.0225753916998932,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004906554063281074,
      "loss": 5.2738,
      "mean_token_accuracy": 0.19146917015314102,
      "num_tokens": 20320413.0,
      "step": 9105
    },
    {
      "entropy": 5.467430496215821,
      "epoch": 1.0231369686078509,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004906439525402618,
      "loss": 5.3433,
      "mean_token_accuracy": 0.1845814421772957,
      "num_tokens": 20331812.0,
      "step": 9110
    },
    {
      "entropy": 5.489290714263916,
      "epoch": 1.0236985455158083,
      "grad_norm": 1.40625,
      "learning_rate": 0.0004906324918862291,
      "loss": 5.275,
      "mean_token_accuracy": 0.1928041696548462,
      "num_tokens": 20342148.0,
      "step": 9115
    },
    {
      "entropy": 5.512719917297363,
      "epoch": 1.024260122423766,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004906210243663742,
      "loss": 5.3828,
      "mean_token_accuracy": 0.18620077669620513,
      "num_tokens": 20353843.0,
      "step": 9120
    },
    {
      "entropy": 5.5219286441802975,
      "epoch": 1.0248216993317234,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004906095499810622,
      "loss": 5.3763,
      "mean_token_accuracy": 0.18759856075048448,
      "num_tokens": 20364846.0,
      "step": 9125
    },
    {
      "entropy": 5.528637361526489,
      "epoch": 1.025383276239681,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004905980687306586,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19188352078199386,
      "num_tokens": 20376874.0,
      "step": 9130
    },
    {
      "entropy": 5.440936708450318,
      "epoch": 1.0259448531476385,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004905865806155288,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18841730058193207,
      "num_tokens": 20388189.0,
      "step": 9135
    },
    {
      "entropy": 5.48921046257019,
      "epoch": 1.0265064300555962,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004905750856360387,
      "loss": 5.3677,
      "mean_token_accuracy": 0.19041854590177537,
      "num_tokens": 20399155.0,
      "step": 9140
    },
    {
      "entropy": 5.4925659656524655,
      "epoch": 1.0270680069635536,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004905635837925543,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19229500144720077,
      "num_tokens": 20410569.0,
      "step": 9145
    },
    {
      "entropy": 5.413026285171509,
      "epoch": 1.0276295838715113,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004905520750854418,
      "loss": 5.2957,
      "mean_token_accuracy": 0.1931578978896141,
      "num_tokens": 20420438.0,
      "step": 9150
    },
    {
      "entropy": 5.490417051315307,
      "epoch": 1.0281911607794687,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004905405595150674,
      "loss": 5.3492,
      "mean_token_accuracy": 0.1866835668683052,
      "num_tokens": 20431874.0,
      "step": 9155
    },
    {
      "entropy": 5.467247390747071,
      "epoch": 1.0287527376874264,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004905290370817982,
      "loss": 5.339,
      "mean_token_accuracy": 0.18550890684127808,
      "num_tokens": 20442792.0,
      "step": 9160
    },
    {
      "entropy": 5.518709945678711,
      "epoch": 1.0293143145953838,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000490517507786001,
      "loss": 5.3419,
      "mean_token_accuracy": 0.18542043417692183,
      "num_tokens": 20454410.0,
      "step": 9165
    },
    {
      "entropy": 5.614403772354126,
      "epoch": 1.0298758915033415,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004905059716280425,
      "loss": 5.4844,
      "mean_token_accuracy": 0.1769074648618698,
      "num_tokens": 20467072.0,
      "step": 9170
    },
    {
      "entropy": 5.502153110504151,
      "epoch": 1.030437468411299,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904944286082904,
      "loss": 5.4196,
      "mean_token_accuracy": 0.18582927733659743,
      "num_tokens": 20477697.0,
      "step": 9175
    },
    {
      "entropy": 5.480784749984741,
      "epoch": 1.0309990453192566,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004904828787271121,
      "loss": 5.2549,
      "mean_token_accuracy": 0.19651424139738083,
      "num_tokens": 20489024.0,
      "step": 9180
    },
    {
      "entropy": 5.450620079040528,
      "epoch": 1.031560622227214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004904713219848753,
      "loss": 5.341,
      "mean_token_accuracy": 0.1878539279103279,
      "num_tokens": 20499928.0,
      "step": 9185
    },
    {
      "entropy": 5.4454082489013675,
      "epoch": 1.0321221991351717,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004904597583819481,
      "loss": 5.3882,
      "mean_token_accuracy": 0.18157906383275985,
      "num_tokens": 20511085.0,
      "step": 9190
    },
    {
      "entropy": 5.522955465316772,
      "epoch": 1.032683776043129,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004904481879186987,
      "loss": 5.2497,
      "mean_token_accuracy": 0.18759061843156816,
      "num_tokens": 20522164.0,
      "step": 9195
    },
    {
      "entropy": 5.492053413391114,
      "epoch": 1.0332453529510865,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004904366105954954,
      "loss": 5.4161,
      "mean_token_accuracy": 0.1851205736398697,
      "num_tokens": 20534187.0,
      "step": 9200
    },
    {
      "entropy": 5.556139183044434,
      "epoch": 1.0338069298590442,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004904250264127067,
      "loss": 5.375,
      "mean_token_accuracy": 0.1928747445344925,
      "num_tokens": 20545819.0,
      "step": 9205
    },
    {
      "entropy": 5.561243534088135,
      "epoch": 1.0343685067670016,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004904134353707018,
      "loss": 5.4056,
      "mean_token_accuracy": 0.18094461262226105,
      "num_tokens": 20556803.0,
      "step": 9210
    },
    {
      "entropy": 5.547152280807495,
      "epoch": 1.0349300836749593,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004904018374698496,
      "loss": 5.3899,
      "mean_token_accuracy": 0.1852301314473152,
      "num_tokens": 20567366.0,
      "step": 9215
    },
    {
      "entropy": 5.480530738830566,
      "epoch": 1.0354916605829168,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903902327105191,
      "loss": 5.4126,
      "mean_token_accuracy": 0.18345397412776948,
      "num_tokens": 20579182.0,
      "step": 9220
    },
    {
      "entropy": 5.451757335662842,
      "epoch": 1.0360532374908744,
      "grad_norm": 1.25,
      "learning_rate": 0.0004903786210930803,
      "loss": 5.2818,
      "mean_token_accuracy": 0.18823508769273758,
      "num_tokens": 20590346.0,
      "step": 9225
    },
    {
      "entropy": 5.508099794387817,
      "epoch": 1.0366148143988319,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903670026179025,
      "loss": 5.3794,
      "mean_token_accuracy": 0.1897455409169197,
      "num_tokens": 20601279.0,
      "step": 9230
    },
    {
      "entropy": 5.5144219398498535,
      "epoch": 1.0371763913067895,
      "grad_norm": 1.234375,
      "learning_rate": 0.000490355377285356,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18815052807331084,
      "num_tokens": 20612163.0,
      "step": 9235
    },
    {
      "entropy": 5.510475730895996,
      "epoch": 1.037737968214747,
      "grad_norm": 1.125,
      "learning_rate": 0.0004903437450958106,
      "loss": 5.2656,
      "mean_token_accuracy": 0.18739271759986878,
      "num_tokens": 20622587.0,
      "step": 9240
    },
    {
      "entropy": 5.487977075576782,
      "epoch": 1.0382995451227046,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004903321060496369,
      "loss": 5.4228,
      "mean_token_accuracy": 0.17784843146800994,
      "num_tokens": 20633341.0,
      "step": 9245
    },
    {
      "entropy": 5.483494901657105,
      "epoch": 1.038861122030662,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004903204601472054,
      "loss": 5.3016,
      "mean_token_accuracy": 0.19420408755540847,
      "num_tokens": 20644582.0,
      "step": 9250
    },
    {
      "entropy": 5.449039554595947,
      "epoch": 1.0394226989386197,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004903088073888869,
      "loss": 5.2786,
      "mean_token_accuracy": 0.1968168333172798,
      "num_tokens": 20654936.0,
      "step": 9255
    },
    {
      "entropy": 5.42897047996521,
      "epoch": 1.0399842758465772,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004902971477750526,
      "loss": 5.2896,
      "mean_token_accuracy": 0.18713279962539672,
      "num_tokens": 20666348.0,
      "step": 9260
    },
    {
      "entropy": 5.61735463142395,
      "epoch": 1.0405458527545348,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004902854813060736,
      "loss": 5.4802,
      "mean_token_accuracy": 0.1854814499616623,
      "num_tokens": 20677482.0,
      "step": 9265
    },
    {
      "entropy": 5.592570543289185,
      "epoch": 1.0411074296624923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004902738079823213,
      "loss": 5.3807,
      "mean_token_accuracy": 0.19114326536655427,
      "num_tokens": 20688369.0,
      "step": 9270
    },
    {
      "entropy": 5.421113681793213,
      "epoch": 1.04166900657045,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004902621278041674,
      "loss": 5.3905,
      "mean_token_accuracy": 0.17911438643932343,
      "num_tokens": 20698933.0,
      "step": 9275
    },
    {
      "entropy": 5.4475332736969,
      "epoch": 1.0422305834784074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004902504407719839,
      "loss": 5.2643,
      "mean_token_accuracy": 0.1880852222442627,
      "num_tokens": 20710192.0,
      "step": 9280
    },
    {
      "entropy": 5.416199159622193,
      "epoch": 1.042792160386365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004902387468861429,
      "loss": 5.3283,
      "mean_token_accuracy": 0.18455059081315994,
      "num_tokens": 20721408.0,
      "step": 9285
    },
    {
      "entropy": 5.407592058181763,
      "epoch": 1.0433537372943225,
      "grad_norm": 1.125,
      "learning_rate": 0.0004902270461470169,
      "loss": 5.2331,
      "mean_token_accuracy": 0.18812310099601745,
      "num_tokens": 20732155.0,
      "step": 9290
    },
    {
      "entropy": 5.494250774383545,
      "epoch": 1.04391531420228,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000490215338554978,
      "loss": 5.228,
      "mean_token_accuracy": 0.1975995719432831,
      "num_tokens": 20741819.0,
      "step": 9295
    },
    {
      "entropy": 5.500752258300781,
      "epoch": 1.0444768911102376,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004902036241103994,
      "loss": 5.3533,
      "mean_token_accuracy": 0.19089614152908324,
      "num_tokens": 20754931.0,
      "step": 9300
    },
    {
      "entropy": 5.4902674674987795,
      "epoch": 1.045038468018195,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004901919028136539,
      "loss": 5.4354,
      "mean_token_accuracy": 0.18038577884435653,
      "num_tokens": 20766533.0,
      "step": 9305
    },
    {
      "entropy": 5.43985505104065,
      "epoch": 1.0456000449261527,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004901801746651147,
      "loss": 5.3096,
      "mean_token_accuracy": 0.19194923788309098,
      "num_tokens": 20777055.0,
      "step": 9310
    },
    {
      "entropy": 5.436862325668335,
      "epoch": 1.04616162183411,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004901684396651554,
      "loss": 5.2481,
      "mean_token_accuracy": 0.18928302675485612,
      "num_tokens": 20786871.0,
      "step": 9315
    },
    {
      "entropy": 5.561196994781494,
      "epoch": 1.0467231987420678,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004901566978141494,
      "loss": 5.4035,
      "mean_token_accuracy": 0.18465593904256822,
      "num_tokens": 20798546.0,
      "step": 9320
    },
    {
      "entropy": 5.486161231994629,
      "epoch": 1.0472847756500252,
      "grad_norm": 1.25,
      "learning_rate": 0.0004901449491124708,
      "loss": 5.3348,
      "mean_token_accuracy": 0.18827358335256578,
      "num_tokens": 20809657.0,
      "step": 9325
    },
    {
      "entropy": 5.466598749160767,
      "epoch": 1.0478463525579829,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004901331935604936,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18431787341833114,
      "num_tokens": 20819662.0,
      "step": 9330
    },
    {
      "entropy": 5.385705375671387,
      "epoch": 1.0484079294659403,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000490121431158592,
      "loss": 5.2207,
      "mean_token_accuracy": 0.19403200447559357,
      "num_tokens": 20830554.0,
      "step": 9335
    },
    {
      "entropy": 5.4792992115020756,
      "epoch": 1.048969506373898,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004901096619071407,
      "loss": 5.3867,
      "mean_token_accuracy": 0.18285879194736482,
      "num_tokens": 20842003.0,
      "step": 9340
    },
    {
      "entropy": 5.4880547523498535,
      "epoch": 1.0495310832818554,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900978858065142,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18953716903924941,
      "num_tokens": 20853453.0,
      "step": 9345
    },
    {
      "entropy": 5.604633808135986,
      "epoch": 1.050092660189813,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004900861028570877,
      "loss": 5.4264,
      "mean_token_accuracy": 0.18585819154977798,
      "num_tokens": 20867296.0,
      "step": 9350
    },
    {
      "entropy": 5.434928846359253,
      "epoch": 1.0506542370977705,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900743130592362,
      "loss": 5.332,
      "mean_token_accuracy": 0.19233575463294983,
      "num_tokens": 20878951.0,
      "step": 9355
    },
    {
      "entropy": 5.529877424240112,
      "epoch": 1.0512158140057282,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004900625164133351,
      "loss": 5.3913,
      "mean_token_accuracy": 0.18940540999174119,
      "num_tokens": 20891103.0,
      "step": 9360
    },
    {
      "entropy": 5.588580799102783,
      "epoch": 1.0517773909136856,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900507129197601,
      "loss": 5.4786,
      "mean_token_accuracy": 0.18177726119756699,
      "num_tokens": 20903570.0,
      "step": 9365
    },
    {
      "entropy": 5.563451719284058,
      "epoch": 1.0523389678216433,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004900389025788871,
      "loss": 5.4287,
      "mean_token_accuracy": 0.1831771895289421,
      "num_tokens": 20914156.0,
      "step": 9370
    },
    {
      "entropy": 5.415740489959717,
      "epoch": 1.0529005447296007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000490027085391092,
      "loss": 5.3435,
      "mean_token_accuracy": 0.18684058338403703,
      "num_tokens": 20925192.0,
      "step": 9375
    },
    {
      "entropy": 5.539825057983398,
      "epoch": 1.0534621216375584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000490015261356751,
      "loss": 5.2245,
      "mean_token_accuracy": 0.20045977979898452,
      "num_tokens": 20935281.0,
      "step": 9380
    },
    {
      "entropy": 5.5713512897491455,
      "epoch": 1.0540236985455158,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004900034304762408,
      "loss": 5.433,
      "mean_token_accuracy": 0.1860588848590851,
      "num_tokens": 20947397.0,
      "step": 9385
    },
    {
      "entropy": 5.404055833816528,
      "epoch": 1.0545852754534732,
      "grad_norm": 1.171875,
      "learning_rate": 0.000489991592749938,
      "loss": 5.3063,
      "mean_token_accuracy": 0.19088430106639862,
      "num_tokens": 20958473.0,
      "step": 9390
    },
    {
      "entropy": 5.593134450912475,
      "epoch": 1.055146852361431,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004899797481782193,
      "loss": 5.4018,
      "mean_token_accuracy": 0.17994294911623002,
      "num_tokens": 20970112.0,
      "step": 9395
    },
    {
      "entropy": 5.478477859497071,
      "epoch": 1.0557084292693883,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004899678967614623,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1890983387827873,
      "num_tokens": 20980762.0,
      "step": 9400
    },
    {
      "entropy": 5.411729145050049,
      "epoch": 1.056270006177346,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004899560385000439,
      "loss": 5.2223,
      "mean_token_accuracy": 0.19373668730258942,
      "num_tokens": 20991663.0,
      "step": 9405
    },
    {
      "entropy": 5.424403285980224,
      "epoch": 1.0568315830853034,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000489944173394342,
      "loss": 5.2334,
      "mean_token_accuracy": 0.19383057802915574,
      "num_tokens": 21002121.0,
      "step": 9410
    },
    {
      "entropy": 5.421369171142578,
      "epoch": 1.0573931599932611,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004899323014447342,
      "loss": 5.3056,
      "mean_token_accuracy": 0.18037335127592086,
      "num_tokens": 21013791.0,
      "step": 9415
    },
    {
      "entropy": 5.408607339859008,
      "epoch": 1.0579547369012186,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004899204226515985,
      "loss": 5.2632,
      "mean_token_accuracy": 0.1933366373181343,
      "num_tokens": 21022977.0,
      "step": 9420
    },
    {
      "entropy": 5.5042637348175045,
      "epoch": 1.0585163138091762,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004899085370153133,
      "loss": 5.4211,
      "mean_token_accuracy": 0.1889128491282463,
      "num_tokens": 21036372.0,
      "step": 9425
    },
    {
      "entropy": 5.686546850204468,
      "epoch": 1.0590778907171337,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004898966445362569,
      "loss": 5.4883,
      "mean_token_accuracy": 0.17733839601278306,
      "num_tokens": 21047984.0,
      "step": 9430
    },
    {
      "entropy": 5.5658105373382565,
      "epoch": 1.0596394676250913,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004898847452148079,
      "loss": 5.3628,
      "mean_token_accuracy": 0.18377949744462968,
      "num_tokens": 21060127.0,
      "step": 9435
    },
    {
      "entropy": 5.464265060424805,
      "epoch": 1.0602010445330488,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004898728390513454,
      "loss": 5.3286,
      "mean_token_accuracy": 0.1852984070777893,
      "num_tokens": 21070659.0,
      "step": 9440
    },
    {
      "entropy": 5.449292993545532,
      "epoch": 1.0607626214410064,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004898609260462483,
      "loss": 5.2649,
      "mean_token_accuracy": 0.19679632931947708,
      "num_tokens": 21081667.0,
      "step": 9445
    },
    {
      "entropy": 5.464795923233032,
      "epoch": 1.0613241983489639,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004898490061998961,
      "loss": 5.3127,
      "mean_token_accuracy": 0.19050173312425614,
      "num_tokens": 21094275.0,
      "step": 9450
    },
    {
      "entropy": 5.498070907592774,
      "epoch": 1.0618857752569215,
      "grad_norm": 1.125,
      "learning_rate": 0.0004898370795126681,
      "loss": 5.3735,
      "mean_token_accuracy": 0.18140986412763596,
      "num_tokens": 21106360.0,
      "step": 9455
    },
    {
      "entropy": 5.5575744152069095,
      "epoch": 1.062447352164879,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004898251459849441,
      "loss": 5.3303,
      "mean_token_accuracy": 0.18390751928091048,
      "num_tokens": 21117647.0,
      "step": 9460
    },
    {
      "entropy": 5.3772101402282715,
      "epoch": 1.0630089290728366,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004898132056171041,
      "loss": 5.1869,
      "mean_token_accuracy": 0.19775499850511552,
      "num_tokens": 21129599.0,
      "step": 9465
    },
    {
      "entropy": 5.450941371917724,
      "epoch": 1.063570505980794,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004898012584095285,
      "loss": 5.3136,
      "mean_token_accuracy": 0.18599545657634736,
      "num_tokens": 21140573.0,
      "step": 9470
    },
    {
      "entropy": 5.537936067581176,
      "epoch": 1.0641320828887517,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004897893043625975,
      "loss": 5.3998,
      "mean_token_accuracy": 0.18946179896593093,
      "num_tokens": 21151899.0,
      "step": 9475
    },
    {
      "entropy": 5.607831811904907,
      "epoch": 1.0646936597967092,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004897773434766917,
      "loss": 5.4523,
      "mean_token_accuracy": 0.18118599504232408,
      "num_tokens": 21162748.0,
      "step": 9480
    },
    {
      "entropy": 5.517045688629151,
      "epoch": 1.0652552367046666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000489765375752192,
      "loss": 5.3594,
      "mean_token_accuracy": 0.18705749064683913,
      "num_tokens": 21172962.0,
      "step": 9485
    },
    {
      "entropy": 5.458545970916748,
      "epoch": 1.0658168136126243,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004897534011894793,
      "loss": 5.3303,
      "mean_token_accuracy": 0.19111815690994263,
      "num_tokens": 21183805.0,
      "step": 9490
    },
    {
      "entropy": 5.386651802062988,
      "epoch": 1.0663783905205817,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004897414197889352,
      "loss": 5.2611,
      "mean_token_accuracy": 0.1960390642285347,
      "num_tokens": 21195129.0,
      "step": 9495
    },
    {
      "entropy": 5.430661201477051,
      "epoch": 1.0669399674285394,
      "grad_norm": 1.15625,
      "learning_rate": 0.000489729431550941,
      "loss": 5.3708,
      "mean_token_accuracy": 0.18504551351070403,
      "num_tokens": 21206300.0,
      "step": 9500
    },
    {
      "entropy": 5.498693799972534,
      "epoch": 1.0675015443364968,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004897174364758783,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18857437670230864,
      "num_tokens": 21217137.0,
      "step": 9505
    },
    {
      "entropy": 5.44688606262207,
      "epoch": 1.0680631212444545,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004897054345641291,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19169701188802718,
      "num_tokens": 21228399.0,
      "step": 9510
    },
    {
      "entropy": 5.482944536209106,
      "epoch": 1.068624698152412,
      "grad_norm": 1.375,
      "learning_rate": 0.0004896934258160757,
      "loss": 5.3934,
      "mean_token_accuracy": 0.18246855437755585,
      "num_tokens": 21239396.0,
      "step": 9515
    },
    {
      "entropy": 5.484139537811279,
      "epoch": 1.0691862750603696,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004896814102321004,
      "loss": 5.2914,
      "mean_token_accuracy": 0.18956653028726578,
      "num_tokens": 21249966.0,
      "step": 9520
    },
    {
      "entropy": 5.495353746414184,
      "epoch": 1.069747851968327,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004896693878125857,
      "loss": 5.319,
      "mean_token_accuracy": 0.1876276433467865,
      "num_tokens": 21260157.0,
      "step": 9525
    },
    {
      "entropy": 5.525558805465698,
      "epoch": 1.0703094288762847,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004896573585579144,
      "loss": 5.377,
      "mean_token_accuracy": 0.19186344593763352,
      "num_tokens": 21270738.0,
      "step": 9530
    },
    {
      "entropy": 5.421734237670899,
      "epoch": 1.070871005784242,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004896453224684695,
      "loss": 5.3227,
      "mean_token_accuracy": 0.18809787034988404,
      "num_tokens": 21282469.0,
      "step": 9535
    },
    {
      "entropy": 5.446729040145874,
      "epoch": 1.0714325826921998,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004896332795446345,
      "loss": 5.2767,
      "mean_token_accuracy": 0.1856149896979332,
      "num_tokens": 21293474.0,
      "step": 9540
    },
    {
      "entropy": 5.521066331863404,
      "epoch": 1.0719941596001572,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004896212297867924,
      "loss": 5.3985,
      "mean_token_accuracy": 0.18117642253637314,
      "num_tokens": 21304904.0,
      "step": 9545
    },
    {
      "entropy": 5.517581748962402,
      "epoch": 1.0725557365081149,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004896091731953271,
      "loss": 5.3413,
      "mean_token_accuracy": 0.18659606724977493,
      "num_tokens": 21315407.0,
      "step": 9550
    },
    {
      "entropy": 5.445900201797485,
      "epoch": 1.0731173134160723,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004895971097706227,
      "loss": 5.2617,
      "mean_token_accuracy": 0.19540172666311265,
      "num_tokens": 21326496.0,
      "step": 9555
    },
    {
      "entropy": 5.4842877864837645,
      "epoch": 1.07367889032403,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004895850395130629,
      "loss": 5.3689,
      "mean_token_accuracy": 0.18484977632761002,
      "num_tokens": 21339449.0,
      "step": 9560
    },
    {
      "entropy": 5.494334268569946,
      "epoch": 1.0742404672319874,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004895729624230323,
      "loss": 5.4279,
      "mean_token_accuracy": 0.17960025370121002,
      "num_tokens": 21350431.0,
      "step": 9565
    },
    {
      "entropy": 5.535207462310791,
      "epoch": 1.074802044139945,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004895608785009153,
      "loss": 5.3195,
      "mean_token_accuracy": 0.18789668381214142,
      "num_tokens": 21361168.0,
      "step": 9570
    },
    {
      "entropy": 5.461625766754151,
      "epoch": 1.0753636210479025,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004895487877470969,
      "loss": 5.3054,
      "mean_token_accuracy": 0.1900552123785019,
      "num_tokens": 21372635.0,
      "step": 9575
    },
    {
      "entropy": 5.361543989181518,
      "epoch": 1.07592519795586,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004895366901619617,
      "loss": 5.3069,
      "mean_token_accuracy": 0.18846821635961533,
      "num_tokens": 21384051.0,
      "step": 9580
    },
    {
      "entropy": 5.4686877727508545,
      "epoch": 1.0764867748638176,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004895245857458951,
      "loss": 5.3445,
      "mean_token_accuracy": 0.17978210151195526,
      "num_tokens": 21397496.0,
      "step": 9585
    },
    {
      "entropy": 5.500164365768432,
      "epoch": 1.077048351771775,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004895124744992825,
      "loss": 5.4007,
      "mean_token_accuracy": 0.18141803443431853,
      "num_tokens": 21409349.0,
      "step": 9590
    },
    {
      "entropy": 5.4704530239105225,
      "epoch": 1.0776099286797327,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004895003564225095,
      "loss": 5.3592,
      "mean_token_accuracy": 0.1870335593819618,
      "num_tokens": 21421843.0,
      "step": 9595
    },
    {
      "entropy": 5.448120498657227,
      "epoch": 1.0781715055876901,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004894882315159621,
      "loss": 5.2391,
      "mean_token_accuracy": 0.1946958541870117,
      "num_tokens": 21433230.0,
      "step": 9600
    },
    {
      "entropy": 5.505858850479126,
      "epoch": 1.0787330824956478,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000489476099780026,
      "loss": 5.3593,
      "mean_token_accuracy": 0.18117243945598602,
      "num_tokens": 21444295.0,
      "step": 9605
    },
    {
      "entropy": 5.519126939773559,
      "epoch": 1.0792946594036052,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004894639612150879,
      "loss": 5.3079,
      "mean_token_accuracy": 0.18445271998643875,
      "num_tokens": 21455721.0,
      "step": 9610
    },
    {
      "entropy": 5.47895188331604,
      "epoch": 1.079856236311563,
      "grad_norm": 1.5390625,
      "learning_rate": 0.0004894518158215339,
      "loss": 5.3263,
      "mean_token_accuracy": 0.18795359879732132,
      "num_tokens": 21466514.0,
      "step": 9615
    },
    {
      "entropy": 5.493203258514404,
      "epoch": 1.0804178132195204,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000489439663599751,
      "loss": 5.3484,
      "mean_token_accuracy": 0.18723692893981933,
      "num_tokens": 21478782.0,
      "step": 9620
    },
    {
      "entropy": 5.519286251068115,
      "epoch": 1.080979390127478,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000489427504550126,
      "loss": 5.3381,
      "mean_token_accuracy": 0.1937120258808136,
      "num_tokens": 21490428.0,
      "step": 9625
    },
    {
      "entropy": 5.336224174499511,
      "epoch": 1.0815409670354355,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004894153386730462,
      "loss": 5.1984,
      "mean_token_accuracy": 0.18784916549921035,
      "num_tokens": 21501530.0,
      "step": 9630
    },
    {
      "entropy": 5.460989904403687,
      "epoch": 1.0821025439433931,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004894031659688988,
      "loss": 5.2732,
      "mean_token_accuracy": 0.19398806244134903,
      "num_tokens": 21512122.0,
      "step": 9635
    },
    {
      "entropy": 5.379255104064941,
      "epoch": 1.0826641208513506,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004893909864380714,
      "loss": 5.2042,
      "mean_token_accuracy": 0.19041338860988616,
      "num_tokens": 21521896.0,
      "step": 9640
    },
    {
      "entropy": 5.447192287445068,
      "epoch": 1.0832256977593082,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000489378800080952,
      "loss": 5.3813,
      "mean_token_accuracy": 0.18846023082733154,
      "num_tokens": 21533584.0,
      "step": 9645
    },
    {
      "entropy": 5.442208385467529,
      "epoch": 1.0837872746672657,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004893666068979283,
      "loss": 5.3461,
      "mean_token_accuracy": 0.1912103682756424,
      "num_tokens": 21545432.0,
      "step": 9650
    },
    {
      "entropy": 5.462935447692871,
      "epoch": 1.0843488515752233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004893544068893887,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18368449807167053,
      "num_tokens": 21556547.0,
      "step": 9655
    },
    {
      "entropy": 5.513771343231201,
      "epoch": 1.0849104284831808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004893422000557217,
      "loss": 5.4091,
      "mean_token_accuracy": 0.18440658003091812,
      "num_tokens": 21567354.0,
      "step": 9660
    },
    {
      "entropy": 5.529688119888306,
      "epoch": 1.0854720053911384,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004893299863973159,
      "loss": 5.2884,
      "mean_token_accuracy": 0.18864646852016448,
      "num_tokens": 21578694.0,
      "step": 9665
    },
    {
      "entropy": 5.380402565002441,
      "epoch": 1.0860335822990959,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004893177659145601,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19846600741147996,
      "num_tokens": 21589365.0,
      "step": 9670
    },
    {
      "entropy": 5.474283170700073,
      "epoch": 1.0865951592070533,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004893055386078438,
      "loss": 5.2917,
      "mean_token_accuracy": 0.18696294575929642,
      "num_tokens": 21600766.0,
      "step": 9675
    },
    {
      "entropy": 5.497973012924194,
      "epoch": 1.087156736115011,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004892933044775558,
      "loss": 5.4869,
      "mean_token_accuracy": 0.17887595295906067,
      "num_tokens": 21613969.0,
      "step": 9680
    },
    {
      "entropy": 5.466034364700318,
      "epoch": 1.0877183130229684,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004892810635240859,
      "loss": 5.3093,
      "mean_token_accuracy": 0.19140152633190155,
      "num_tokens": 21626302.0,
      "step": 9685
    },
    {
      "entropy": 5.545321846008301,
      "epoch": 1.088279889930926,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004892688157478238,
      "loss": 5.375,
      "mean_token_accuracy": 0.18986247926950456,
      "num_tokens": 21637058.0,
      "step": 9690
    },
    {
      "entropy": 5.506683731079102,
      "epoch": 1.0888414668388835,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004892565611491597,
      "loss": 5.3865,
      "mean_token_accuracy": 0.1832375004887581,
      "num_tokens": 21648729.0,
      "step": 9695
    },
    {
      "entropy": 5.364797878265381,
      "epoch": 1.0894030437468412,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004892442997284834,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19554288685321808,
      "num_tokens": 21659961.0,
      "step": 9700
    },
    {
      "entropy": 5.438081502914429,
      "epoch": 1.0899646206547986,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004892320314861855,
      "loss": 5.3155,
      "mean_token_accuracy": 0.19172561466693877,
      "num_tokens": 21670577.0,
      "step": 9705
    },
    {
      "entropy": 5.438214874267578,
      "epoch": 1.0905261975627563,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004892197564226568,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19490559995174409,
      "num_tokens": 21679882.0,
      "step": 9710
    },
    {
      "entropy": 5.516940402984619,
      "epoch": 1.0910877744707137,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004892074745382878,
      "loss": 5.3991,
      "mean_token_accuracy": 0.1930496111512184,
      "num_tokens": 21690595.0,
      "step": 9715
    },
    {
      "entropy": 5.44785852432251,
      "epoch": 1.0916493513786714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004891951858334698,
      "loss": 5.2536,
      "mean_token_accuracy": 0.1920822814106941,
      "num_tokens": 21701251.0,
      "step": 9720
    },
    {
      "entropy": 5.543320083618164,
      "epoch": 1.0922109282866288,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004891828903085939,
      "loss": 5.4376,
      "mean_token_accuracy": 0.18532729893922806,
      "num_tokens": 21713286.0,
      "step": 9725
    },
    {
      "entropy": 5.532457065582276,
      "epoch": 1.0927725051945865,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004891705879640518,
      "loss": 5.4779,
      "mean_token_accuracy": 0.18655193150043486,
      "num_tokens": 21725508.0,
      "step": 9730
    },
    {
      "entropy": 5.463955593109131,
      "epoch": 1.093334082102544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004891582788002351,
      "loss": 5.2824,
      "mean_token_accuracy": 0.18588511496782303,
      "num_tokens": 21736467.0,
      "step": 9735
    },
    {
      "entropy": 5.525961256027221,
      "epoch": 1.0938956590105016,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004891459628175358,
      "loss": 5.3851,
      "mean_token_accuracy": 0.18890249878168106,
      "num_tokens": 21746114.0,
      "step": 9740
    },
    {
      "entropy": 5.418683290481567,
      "epoch": 1.094457235918459,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000489133640016346,
      "loss": 5.3852,
      "mean_token_accuracy": 0.184807588160038,
      "num_tokens": 21757145.0,
      "step": 9745
    },
    {
      "entropy": 5.508657789230346,
      "epoch": 1.0950188128264167,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004891213103970579,
      "loss": 5.3535,
      "mean_token_accuracy": 0.1814421758055687,
      "num_tokens": 21768294.0,
      "step": 9750
    },
    {
      "entropy": 5.551876974105835,
      "epoch": 1.095580389734374,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004891089739600643,
      "loss": 5.3598,
      "mean_token_accuracy": 0.18730462193489075,
      "num_tokens": 21779667.0,
      "step": 9755
    },
    {
      "entropy": 5.417598676681519,
      "epoch": 1.0961419666423318,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004890966307057581,
      "loss": 5.2194,
      "mean_token_accuracy": 0.19345995038747787,
      "num_tokens": 21789635.0,
      "step": 9760
    },
    {
      "entropy": 5.351821994781494,
      "epoch": 1.0967035435502892,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000489084280634532,
      "loss": 5.1999,
      "mean_token_accuracy": 0.19870178401470184,
      "num_tokens": 21799462.0,
      "step": 9765
    },
    {
      "entropy": 5.369420671463013,
      "epoch": 1.0972651204582466,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004890719237467794,
      "loss": 5.2762,
      "mean_token_accuracy": 0.19289530813694,
      "num_tokens": 21811520.0,
      "step": 9770
    },
    {
      "entropy": 5.561643457412719,
      "epoch": 1.0978266973662043,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004890595600428937,
      "loss": 5.345,
      "mean_token_accuracy": 0.189879010617733,
      "num_tokens": 21821997.0,
      "step": 9775
    },
    {
      "entropy": 5.492782306671143,
      "epoch": 1.0983882742741617,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004890471895232684,
      "loss": 5.4036,
      "mean_token_accuracy": 0.18264955133199692,
      "num_tokens": 21833353.0,
      "step": 9780
    },
    {
      "entropy": 5.42255539894104,
      "epoch": 1.0989498511821194,
      "grad_norm": 1.125,
      "learning_rate": 0.0004890348121882978,
      "loss": 5.3194,
      "mean_token_accuracy": 0.19271555542945862,
      "num_tokens": 21843691.0,
      "step": 9785
    },
    {
      "entropy": 5.423749876022339,
      "epoch": 1.0995114280900768,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004890224280383758,
      "loss": 5.2945,
      "mean_token_accuracy": 0.18863110840320588,
      "num_tokens": 21853771.0,
      "step": 9790
    },
    {
      "entropy": 5.375594663619995,
      "epoch": 1.1000730049980345,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004890100370738966,
      "loss": 5.1814,
      "mean_token_accuracy": 0.19738784283399582,
      "num_tokens": 21863424.0,
      "step": 9795
    },
    {
      "entropy": 5.4654093265533445,
      "epoch": 1.100634581905992,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004889976392952548,
      "loss": 5.3052,
      "mean_token_accuracy": 0.1905913159251213,
      "num_tokens": 21875407.0,
      "step": 9800
    },
    {
      "entropy": 5.480332612991333,
      "epoch": 1.1011961588139496,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004889852347028452,
      "loss": 5.3738,
      "mean_token_accuracy": 0.18767334520816803,
      "num_tokens": 21887081.0,
      "step": 9805
    },
    {
      "entropy": 5.400891351699829,
      "epoch": 1.101757735721907,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004889728232970626,
      "loss": 5.2789,
      "mean_token_accuracy": 0.18497878611087798,
      "num_tokens": 21897858.0,
      "step": 9810
    },
    {
      "entropy": 5.458182239532471,
      "epoch": 1.1023193126298647,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004889604050783025,
      "loss": 5.3206,
      "mean_token_accuracy": 0.1876216262578964,
      "num_tokens": 21907985.0,
      "step": 9815
    },
    {
      "entropy": 5.462566804885864,
      "epoch": 1.1028808895378222,
      "grad_norm": 1.125,
      "learning_rate": 0.00048894798004696,
      "loss": 5.3697,
      "mean_token_accuracy": 0.181930510699749,
      "num_tokens": 21918726.0,
      "step": 9820
    },
    {
      "entropy": 5.445682525634766,
      "epoch": 1.1034424664457798,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004889355482034309,
      "loss": 5.2899,
      "mean_token_accuracy": 0.1885381817817688,
      "num_tokens": 21929027.0,
      "step": 9825
    },
    {
      "entropy": 5.423603439331055,
      "epoch": 1.1040040433537373,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004889231095481111,
      "loss": 5.218,
      "mean_token_accuracy": 0.19823161363601685,
      "num_tokens": 21938356.0,
      "step": 9830
    },
    {
      "entropy": 5.434048748016357,
      "epoch": 1.104565620261695,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004889106640813964,
      "loss": 5.2498,
      "mean_token_accuracy": 0.1899213507771492,
      "num_tokens": 21947889.0,
      "step": 9835
    },
    {
      "entropy": 5.502470302581787,
      "epoch": 1.1051271971696524,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004888982118036831,
      "loss": 5.3469,
      "mean_token_accuracy": 0.1923964411020279,
      "num_tokens": 21958409.0,
      "step": 9840
    },
    {
      "entropy": 5.412892150878906,
      "epoch": 1.10568877407761,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004888857527153679,
      "loss": 5.313,
      "mean_token_accuracy": 0.18687284737825394,
      "num_tokens": 21970120.0,
      "step": 9845
    },
    {
      "entropy": 5.466385173797607,
      "epoch": 1.1062503509855675,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004888732868168473,
      "loss": 5.3068,
      "mean_token_accuracy": 0.19212121665477752,
      "num_tokens": 21981499.0,
      "step": 9850
    },
    {
      "entropy": 5.392094850540161,
      "epoch": 1.1068119278935251,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004888608141085183,
      "loss": 5.2051,
      "mean_token_accuracy": 0.1995815321803093,
      "num_tokens": 21991263.0,
      "step": 9855
    },
    {
      "entropy": 5.373749208450318,
      "epoch": 1.1073735048014826,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004888483345907781,
      "loss": 5.1921,
      "mean_token_accuracy": 0.1903995916247368,
      "num_tokens": 22001582.0,
      "step": 9860
    },
    {
      "entropy": 5.443917083740234,
      "epoch": 1.10793508170944,
      "grad_norm": 1.203125,
      "learning_rate": 0.000488835848264024,
      "loss": 5.2795,
      "mean_token_accuracy": 0.201464182138443,
      "num_tokens": 22012447.0,
      "step": 9865
    },
    {
      "entropy": 5.458194971084595,
      "epoch": 1.1084966586173977,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004888233551286534,
      "loss": 5.3733,
      "mean_token_accuracy": 0.18749553859233856,
      "num_tokens": 22024865.0,
      "step": 9870
    },
    {
      "entropy": 5.370743751525879,
      "epoch": 1.109058235525355,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004888108551850644,
      "loss": 5.2987,
      "mean_token_accuracy": 0.19044186919927597,
      "num_tokens": 22036298.0,
      "step": 9875
    },
    {
      "entropy": 5.561146926879883,
      "epoch": 1.1096198124333128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004887983484336548,
      "loss": 5.2975,
      "mean_token_accuracy": 0.19128030389547349,
      "num_tokens": 22047499.0,
      "step": 9880
    },
    {
      "entropy": 5.456936359405518,
      "epoch": 1.1101813893412702,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004887858348748229,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18805715441703796,
      "num_tokens": 22057612.0,
      "step": 9885
    },
    {
      "entropy": 5.418582105636597,
      "epoch": 1.1107429662492279,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000488773314508967,
      "loss": 5.3718,
      "mean_token_accuracy": 0.18790108859539031,
      "num_tokens": 22068566.0,
      "step": 9890
    },
    {
      "entropy": 5.4749613285064695,
      "epoch": 1.1113045431571853,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004887607873364859,
      "loss": 5.2504,
      "mean_token_accuracy": 0.1891116678714752,
      "num_tokens": 22079821.0,
      "step": 9895
    },
    {
      "entropy": 5.445761823654175,
      "epoch": 1.111866120065143,
      "grad_norm": 1.125,
      "learning_rate": 0.0004887482533577786,
      "loss": 5.3182,
      "mean_token_accuracy": 0.18613671511411667,
      "num_tokens": 22090143.0,
      "step": 9900
    },
    {
      "entropy": 5.430360984802246,
      "epoch": 1.1124276969731004,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004887357125732438,
      "loss": 5.3635,
      "mean_token_accuracy": 0.18811374604701997,
      "num_tokens": 22101601.0,
      "step": 9905
    },
    {
      "entropy": 5.492471408843994,
      "epoch": 1.112989273881058,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000488723164983281,
      "loss": 5.2749,
      "mean_token_accuracy": 0.19400020837783813,
      "num_tokens": 22113102.0,
      "step": 9910
    },
    {
      "entropy": 5.506822776794434,
      "epoch": 1.1135508507890155,
      "grad_norm": 1.3125,
      "learning_rate": 0.0004887106105882898,
      "loss": 5.4031,
      "mean_token_accuracy": 0.1856501504778862,
      "num_tokens": 22124650.0,
      "step": 9915
    },
    {
      "entropy": 5.482801198959351,
      "epoch": 1.1141124276969732,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004886980493886699,
      "loss": 5.3414,
      "mean_token_accuracy": 0.19093808829784392,
      "num_tokens": 22136637.0,
      "step": 9920
    },
    {
      "entropy": 5.414367532730102,
      "epoch": 1.1146740046049306,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004886854813848212,
      "loss": 5.2537,
      "mean_token_accuracy": 0.19132547676563263,
      "num_tokens": 22149318.0,
      "step": 9925
    },
    {
      "entropy": 5.375391817092895,
      "epoch": 1.1152355815128883,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004886729065771439,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19446331560611724,
      "num_tokens": 22159410.0,
      "step": 9930
    },
    {
      "entropy": 5.417624998092651,
      "epoch": 1.1157971584208457,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004886603249660384,
      "loss": 5.2535,
      "mean_token_accuracy": 0.1905391827225685,
      "num_tokens": 22170017.0,
      "step": 9935
    },
    {
      "entropy": 5.399700403213501,
      "epoch": 1.1163587353288034,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004886477365519052,
      "loss": 5.251,
      "mean_token_accuracy": 0.19104736298322678,
      "num_tokens": 22179961.0,
      "step": 9940
    },
    {
      "entropy": 5.471824359893799,
      "epoch": 1.1169203122367608,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004886351413351453,
      "loss": 5.3761,
      "mean_token_accuracy": 0.18171337395906448,
      "num_tokens": 22191415.0,
      "step": 9945
    },
    {
      "entropy": 5.524193572998047,
      "epoch": 1.1174818891447185,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004886225393161596,
      "loss": 5.3189,
      "mean_token_accuracy": 0.20112415999174119,
      "num_tokens": 22205354.0,
      "step": 9950
    },
    {
      "entropy": 5.553970861434936,
      "epoch": 1.118043466052676,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004886099304953495,
      "loss": 5.3749,
      "mean_token_accuracy": 0.18762922286987305,
      "num_tokens": 22216989.0,
      "step": 9955
    },
    {
      "entropy": 5.54153618812561,
      "epoch": 1.1186050429606333,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004885973148731162,
      "loss": 5.3836,
      "mean_token_accuracy": 0.18514294773340226,
      "num_tokens": 22228421.0,
      "step": 9960
    },
    {
      "entropy": 5.465285587310791,
      "epoch": 1.119166619868591,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004885846924498617,
      "loss": 5.2455,
      "mean_token_accuracy": 0.1924235299229622,
      "num_tokens": 22240621.0,
      "step": 9965
    },
    {
      "entropy": 5.386264324188232,
      "epoch": 1.1197281967765484,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004885720632259876,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19798070043325425,
      "num_tokens": 22250936.0,
      "step": 9970
    },
    {
      "entropy": 5.4189681053161625,
      "epoch": 1.120289773684506,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004885594272018963,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18961561918258668,
      "num_tokens": 22261766.0,
      "step": 9975
    },
    {
      "entropy": 5.460068082809448,
      "epoch": 1.1208513505924635,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048854678437799,
      "loss": 5.2812,
      "mean_token_accuracy": 0.18682191669940948,
      "num_tokens": 22271831.0,
      "step": 9980
    },
    {
      "entropy": 5.36069803237915,
      "epoch": 1.1214129275004212,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004885341347546713,
      "loss": 5.3049,
      "mean_token_accuracy": 0.18514818996191024,
      "num_tokens": 22283756.0,
      "step": 9985
    },
    {
      "entropy": 5.358317184448242,
      "epoch": 1.1219745044083786,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004885214783323428,
      "loss": 5.2559,
      "mean_token_accuracy": 0.1932868927717209,
      "num_tokens": 22294576.0,
      "step": 9990
    },
    {
      "entropy": 5.436130475997925,
      "epoch": 1.1225360813163363,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004885088151114076,
      "loss": 5.3336,
      "mean_token_accuracy": 0.1859567016363144,
      "num_tokens": 22305003.0,
      "step": 9995
    },
    {
      "entropy": 5.417113876342773,
      "epoch": 1.1230976582242937,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004884961450922691,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19139058142900467,
      "num_tokens": 22316505.0,
      "step": 10000
    },
    {
      "entropy": 5.467714834213257,
      "epoch": 1.1236592351322514,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004884834682753305,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18907896876335145,
      "num_tokens": 22327853.0,
      "step": 10005
    },
    {
      "entropy": 5.460165548324585,
      "epoch": 1.1242208120402089,
      "grad_norm": 1.4609375,
      "learning_rate": 0.0004884707846609953,
      "loss": 5.2663,
      "mean_token_accuracy": 0.1911141648888588,
      "num_tokens": 22337386.0,
      "step": 10010
    },
    {
      "entropy": 5.475612831115723,
      "epoch": 1.1247823889481665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004884580942496677,
      "loss": 5.4458,
      "mean_token_accuracy": 0.18234493285417558,
      "num_tokens": 22348742.0,
      "step": 10015
    },
    {
      "entropy": 5.512842607498169,
      "epoch": 1.125343965856124,
      "grad_norm": 1.125,
      "learning_rate": 0.0004884453970417515,
      "loss": 5.3589,
      "mean_token_accuracy": 0.18709605783224106,
      "num_tokens": 22360083.0,
      "step": 10020
    },
    {
      "entropy": 5.430652379989624,
      "epoch": 1.1259055427640816,
      "grad_norm": 1.2109375,
      "learning_rate": 0.000488432693037651,
      "loss": 5.3102,
      "mean_token_accuracy": 0.18195385485887527,
      "num_tokens": 22370982.0,
      "step": 10025
    },
    {
      "entropy": 5.418444204330444,
      "epoch": 1.126467119672039,
      "grad_norm": 1.15625,
      "learning_rate": 0.000488419982237771,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19589971303939818,
      "num_tokens": 22382383.0,
      "step": 10030
    },
    {
      "entropy": 5.474942779541015,
      "epoch": 1.1270286965799967,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004884072646425157,
      "loss": 5.2734,
      "mean_token_accuracy": 0.18627331554889678,
      "num_tokens": 22393994.0,
      "step": 10035
    },
    {
      "entropy": 5.344572257995606,
      "epoch": 1.1275902734879542,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004883945402522907,
      "loss": 5.2063,
      "mean_token_accuracy": 0.19449977427721024,
      "num_tokens": 22405354.0,
      "step": 10040
    },
    {
      "entropy": 5.353824520111084,
      "epoch": 1.1281518503959118,
      "grad_norm": 1.125,
      "learning_rate": 0.0004883818090675005,
      "loss": 5.2457,
      "mean_token_accuracy": 0.1900654375553131,
      "num_tokens": 22416708.0,
      "step": 10045
    },
    {
      "entropy": 5.408300590515137,
      "epoch": 1.1287134273038693,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004883690710885507,
      "loss": 5.2579,
      "mean_token_accuracy": 0.1918439030647278,
      "num_tokens": 22427724.0,
      "step": 10050
    },
    {
      "entropy": 5.471399211883545,
      "epoch": 1.1292750042118267,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883563263158471,
      "loss": 5.3445,
      "mean_token_accuracy": 0.18918549716472627,
      "num_tokens": 22437823.0,
      "step": 10055
    },
    {
      "entropy": 5.453328466415405,
      "epoch": 1.1298365811197844,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004883435747497952,
      "loss": 5.3751,
      "mean_token_accuracy": 0.18997286558151244,
      "num_tokens": 22450155.0,
      "step": 10060
    },
    {
      "entropy": 5.490304660797119,
      "epoch": 1.130398158027742,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004883308163908012,
      "loss": 5.2901,
      "mean_token_accuracy": 0.1922532871365547,
      "num_tokens": 22460484.0,
      "step": 10065
    },
    {
      "entropy": 5.380366659164428,
      "epoch": 1.1309597349356995,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004883180512392714,
      "loss": 5.3082,
      "mean_token_accuracy": 0.18927801996469498,
      "num_tokens": 22472063.0,
      "step": 10070
    },
    {
      "entropy": 5.419078254699707,
      "epoch": 1.131521311843657,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004883052792956119,
      "loss": 5.3094,
      "mean_token_accuracy": 0.19651554077863692,
      "num_tokens": 22483669.0,
      "step": 10075
    },
    {
      "entropy": 5.523752450942993,
      "epoch": 1.1320828887516146,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882925005602296,
      "loss": 5.3802,
      "mean_token_accuracy": 0.188030843436718,
      "num_tokens": 22495505.0,
      "step": 10080
    },
    {
      "entropy": 5.415947437286377,
      "epoch": 1.132644465659572,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004882797150335314,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1881848931312561,
      "num_tokens": 22505260.0,
      "step": 10085
    },
    {
      "entropy": 5.4666749954223635,
      "epoch": 1.1332060425675297,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048826692271592434,
      "loss": 5.3886,
      "mean_token_accuracy": 0.18599129170179368,
      "num_tokens": 22516362.0,
      "step": 10090
    },
    {
      "entropy": 5.556431722640991,
      "epoch": 1.133767619475487,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882541236078158,
      "loss": 5.3421,
      "mean_token_accuracy": 0.18657373636960983,
      "num_tokens": 22527760.0,
      "step": 10095
    },
    {
      "entropy": 5.635705757141113,
      "epoch": 1.1343291963834448,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048824131770961316,
      "loss": 5.4176,
      "mean_token_accuracy": 0.18186973929405212,
      "num_tokens": 22540601.0,
      "step": 10100
    },
    {
      "entropy": 5.3831298828125,
      "epoch": 1.1348907732914022,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048822850502172437,
      "loss": 5.3446,
      "mean_token_accuracy": 0.18705189526081084,
      "num_tokens": 22552469.0,
      "step": 10105
    },
    {
      "entropy": 5.38705587387085,
      "epoch": 1.1354523501993599,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004882156855445571,
      "loss": 5.2892,
      "mean_token_accuracy": 0.1914632275700569,
      "num_tokens": 22563552.0,
      "step": 10110
    },
    {
      "entropy": 5.4868463516235355,
      "epoch": 1.1360139271073173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004882028592785199,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19123449176549911,
      "num_tokens": 22575619.0,
      "step": 10115
    },
    {
      "entropy": 5.408901166915894,
      "epoch": 1.136575504015275,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048819002622402095,
      "loss": 5.2491,
      "mean_token_accuracy": 0.18964056223630904,
      "num_tokens": 22585854.0,
      "step": 10120
    },
    {
      "entropy": 5.456530618667602,
      "epoch": 1.1371370809232324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004881771863814688,
      "loss": 5.2682,
      "mean_token_accuracy": 0.1985445573925972,
      "num_tokens": 22595899.0,
      "step": 10125
    },
    {
      "entropy": 5.425875425338745,
      "epoch": 1.13769865783119,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048816433975127244,
      "loss": 5.2433,
      "mean_token_accuracy": 0.19127049595117568,
      "num_tokens": 22607517.0,
      "step": 10130
    },
    {
      "entropy": 5.445045423507691,
      "epoch": 1.1382602347391475,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004881514863338407,
      "loss": 5.3526,
      "mean_token_accuracy": 0.18836387246847153,
      "num_tokens": 22618531.0,
      "step": 10135
    },
    {
      "entropy": 5.326637220382691,
      "epoch": 1.1388218116471052,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048813862612958307,
      "loss": 5.2068,
      "mean_token_accuracy": 0.19102227687835693,
      "num_tokens": 22629193.0,
      "step": 10140
    },
    {
      "entropy": 5.518526268005371,
      "epoch": 1.1393833885550626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004881257591389089,
      "loss": 5.3212,
      "mean_token_accuracy": 0.1933045968413353,
      "num_tokens": 22639794.0,
      "step": 10145
    },
    {
      "entropy": 5.482657623291016,
      "epoch": 1.13994496546302,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048811288536222794,
      "loss": 5.2667,
      "mean_token_accuracy": 0.1996556416153908,
      "num_tokens": 22650061.0,
      "step": 10150
    },
    {
      "entropy": 5.412937450408935,
      "epoch": 1.1405065423709777,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048810000479995,
      "loss": 5.3522,
      "mean_token_accuracy": 0.18283225744962692,
      "num_tokens": 22661399.0,
      "step": 10155
    },
    {
      "entropy": 5.5768355369567875,
      "epoch": 1.1410681192789354,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048808711745248524,
      "loss": 5.446,
      "mean_token_accuracy": 0.1815304145216942,
      "num_tokens": 22672445.0,
      "step": 10160
    },
    {
      "entropy": 5.483533716201782,
      "epoch": 1.1416296961868928,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048807422332024404,
      "loss": 5.2568,
      "mean_token_accuracy": 0.19154933393001555,
      "num_tokens": 22682545.0,
      "step": 10165
    },
    {
      "entropy": 5.4239561557769775,
      "epoch": 1.1421912730948502,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004880613224036369,
      "loss": 5.3176,
      "mean_token_accuracy": 0.19063657522201538,
      "num_tokens": 22693885.0,
      "step": 10170
    },
    {
      "entropy": 5.402264738082886,
      "epoch": 1.142752850002808,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048804841470307466,
      "loss": 5.2469,
      "mean_token_accuracy": 0.18972668498754502,
      "num_tokens": 22703841.0,
      "step": 10175
    },
    {
      "entropy": 5.4588727951049805,
      "epoch": 1.1433144269107653,
      "grad_norm": 1.015625,
      "learning_rate": 0.00048803550021896824,
      "loss": 5.4018,
      "mean_token_accuracy": 0.1856292799115181,
      "num_tokens": 22714142.0,
      "step": 10180
    },
    {
      "entropy": 5.481220769882202,
      "epoch": 1.143876003818723,
      "grad_norm": 1.25,
      "learning_rate": 0.0004880225789517289,
      "loss": 5.2769,
      "mean_token_accuracy": 0.1871703326702118,
      "num_tokens": 22725056.0,
      "step": 10185
    },
    {
      "entropy": 5.407083368301391,
      "epoch": 1.1444375807266804,
      "grad_norm": 1.234375,
      "learning_rate": 0.000488009650901768,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19286762326955795,
      "num_tokens": 22735664.0,
      "step": 10190
    },
    {
      "entropy": 5.385028409957886,
      "epoch": 1.144999157634638,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048799671606949724,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18895055651664733,
      "num_tokens": 22748071.0,
      "step": 10195
    },
    {
      "entropy": 5.489776229858398,
      "epoch": 1.1455607345425955,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048798377445532837,
      "loss": 5.3386,
      "mean_token_accuracy": 0.185627843439579,
      "num_tokens": 22759477.0,
      "step": 10200
    },
    {
      "entropy": 5.407359600067139,
      "epoch": 1.1461223114505532,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004879708260596736,
      "loss": 5.2113,
      "mean_token_accuracy": 0.1960008829832077,
      "num_tokens": 22769955.0,
      "step": 10205
    },
    {
      "entropy": 5.334399747848511,
      "epoch": 1.1466838883585107,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004879578708829451,
      "loss": 5.3072,
      "mean_token_accuracy": 0.18502966314554214,
      "num_tokens": 22781216.0,
      "step": 10210
    },
    {
      "entropy": 5.42430100440979,
      "epoch": 1.1472454652664683,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004879449089255553,
      "loss": 5.2468,
      "mean_token_accuracy": 0.19306482821702958,
      "num_tokens": 22792581.0,
      "step": 10215
    },
    {
      "entropy": 5.387201738357544,
      "epoch": 1.1478070421744258,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004879319401879171,
      "loss": 5.2138,
      "mean_token_accuracy": 0.1944917306303978,
      "num_tokens": 22803305.0,
      "step": 10220
    },
    {
      "entropy": 5.455933570861816,
      "epoch": 1.1483686190823834,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048791896467044335,
      "loss": 5.3479,
      "mean_token_accuracy": 0.19006223380565643,
      "num_tokens": 22815558.0,
      "step": 10225
    },
    {
      "entropy": 5.405339765548706,
      "epoch": 1.1489301959903409,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048790598237354716,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19270478785037995,
      "num_tokens": 22825428.0,
      "step": 10230
    },
    {
      "entropy": 5.488089990615845,
      "epoch": 1.1494917728982985,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004878929932976419,
      "loss": 5.3194,
      "mean_token_accuracy": 0.18923746645450593,
      "num_tokens": 22836294.0,
      "step": 10235
    },
    {
      "entropy": 5.450779771804809,
      "epoch": 1.150053349806256,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004878799974431412,
      "loss": 5.3953,
      "mean_token_accuracy": 0.18792014420032502,
      "num_tokens": 22847835.0,
      "step": 10240
    },
    {
      "entropy": 5.496893215179443,
      "epoch": 1.1506149267142134,
      "grad_norm": 1.4921875,
      "learning_rate": 0.0004878669948104588,
      "loss": 5.3336,
      "mean_token_accuracy": 0.18722350001335145,
      "num_tokens": 22859726.0,
      "step": 10245
    },
    {
      "entropy": 5.42360110282898,
      "epoch": 1.151176503622171,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048785398540000867,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1876784473657608,
      "num_tokens": 22871250.0,
      "step": 10250
    },
    {
      "entropy": 5.481786108016967,
      "epoch": 1.1517380805301287,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048784096921220514,
      "loss": 5.3554,
      "mean_token_accuracy": 0.1848970040678978,
      "num_tokens": 22882034.0,
      "step": 10255
    },
    {
      "entropy": 5.520159578323364,
      "epoch": 1.1522996574380862,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004878279462474627,
      "loss": 5.4237,
      "mean_token_accuracy": 0.18318284600973128,
      "num_tokens": 22894606.0,
      "step": 10260
    },
    {
      "entropy": 5.392535734176636,
      "epoch": 1.1528612343460436,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048781491650619574,
      "loss": 5.2736,
      "mean_token_accuracy": 0.19084737896919252,
      "num_tokens": 22905246.0,
      "step": 10265
    },
    {
      "entropy": 5.420025253295899,
      "epoch": 1.1534228112540013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048780187998881944,
      "loss": 5.2171,
      "mean_token_accuracy": 0.1891181707382202,
      "num_tokens": 22916202.0,
      "step": 10270
    },
    {
      "entropy": 5.427329444885254,
      "epoch": 1.1539843881619587,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004877888366957486,
      "loss": 5.2108,
      "mean_token_accuracy": 0.1901298314332962,
      "num_tokens": 22926451.0,
      "step": 10275
    },
    {
      "entropy": 5.45597357749939,
      "epoch": 1.1545459650699164,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004877757866273988,
      "loss": 5.3092,
      "mean_token_accuracy": 0.19215435832738875,
      "num_tokens": 22936618.0,
      "step": 10280
    },
    {
      "entropy": 5.513113403320313,
      "epoch": 1.1551075419778738,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004877627297841853,
      "loss": 5.3455,
      "mean_token_accuracy": 0.1840074509382248,
      "num_tokens": 22948278.0,
      "step": 10285
    },
    {
      "entropy": 5.4155669689178465,
      "epoch": 1.1556691188858315,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004877496661665241,
      "loss": 5.2333,
      "mean_token_accuracy": 0.19042912721633912,
      "num_tokens": 22959928.0,
      "step": 10290
    },
    {
      "entropy": 5.383219242095947,
      "epoch": 1.156230695793789,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000487736595774831,
      "loss": 5.2733,
      "mean_token_accuracy": 0.1911442294716835,
      "num_tokens": 22971050.0,
      "step": 10295
    },
    {
      "entropy": 5.384141302108764,
      "epoch": 1.1567922727017466,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048772351860952215,
      "loss": 5.2174,
      "mean_token_accuracy": 0.18788133114576339,
      "num_tokens": 22981471.0,
      "step": 10300
    },
    {
      "entropy": 5.533522033691407,
      "epoch": 1.157353849609704,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004877104346710139,
      "loss": 5.4395,
      "mean_token_accuracy": 0.19253017902374267,
      "num_tokens": 22992526.0,
      "step": 10305
    },
    {
      "entropy": 5.394940519332886,
      "epoch": 1.1579154265176617,
      "grad_norm": 1.09375,
      "learning_rate": 0.000487697343959723,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18800359815359116,
      "num_tokens": 23003408.0,
      "step": 10310
    },
    {
      "entropy": 5.43931999206543,
      "epoch": 1.158477003425619,
      "grad_norm": 1.125,
      "learning_rate": 0.0004876842464760661,
      "loss": 5.2245,
      "mean_token_accuracy": 0.19599764943122863,
      "num_tokens": 23013406.0,
      "step": 10315
    },
    {
      "entropy": 5.3559839725494385,
      "epoch": 1.1590385803335768,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048767114222046036,
      "loss": 5.2855,
      "mean_token_accuracy": 0.18987657725811005,
      "num_tokens": 23024960.0,
      "step": 10320
    },
    {
      "entropy": 5.4738908290863035,
      "epoch": 1.1596001572415342,
      "grad_norm": 1.15625,
      "learning_rate": 0.000487658031193323,
      "loss": 5.3127,
      "mean_token_accuracy": 0.18507126122713088,
      "num_tokens": 23037514.0,
      "step": 10325
    },
    {
      "entropy": 5.485889482498169,
      "epoch": 1.1601617341494919,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004876449133950714,
      "loss": 5.4131,
      "mean_token_accuracy": 0.18047918528318405,
      "num_tokens": 23048460.0,
      "step": 10330
    },
    {
      "entropy": 5.494572591781616,
      "epoch": 1.1607233110574493,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004876317888261233,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19138725996017455,
      "num_tokens": 23058866.0,
      "step": 10335
    },
    {
      "entropy": 5.407448577880859,
      "epoch": 1.1612848879654067,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048761865748689657,
      "loss": 5.2924,
      "mean_token_accuracy": 0.18648287653923035,
      "num_tokens": 23069097.0,
      "step": 10340
    },
    {
      "entropy": 5.517587041854858,
      "epoch": 1.1618464648733644,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048760551937780927,
      "loss": 5.2815,
      "mean_token_accuracy": 0.19022362530231476,
      "num_tokens": 23079553.0,
      "step": 10345
    },
    {
      "entropy": 5.435378313064575,
      "epoch": 1.162408041781322,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048759237449927986,
      "loss": 5.2878,
      "mean_token_accuracy": 0.18987303376197814,
      "num_tokens": 23090554.0,
      "step": 10350
    },
    {
      "entropy": 5.368752288818359,
      "epoch": 1.1629696186892795,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048757922285172674,
      "loss": 5.2004,
      "mean_token_accuracy": 0.1975506916642189,
      "num_tokens": 23101740.0,
      "step": 10355
    },
    {
      "entropy": 5.4575437068939205,
      "epoch": 1.163531195597237,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048756606443556874,
      "loss": 5.2802,
      "mean_token_accuracy": 0.19502927213907242,
      "num_tokens": 23112842.0,
      "step": 10360
    },
    {
      "entropy": 5.3834569454193115,
      "epoch": 1.1640927725051946,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048755289925122474,
      "loss": 5.302,
      "mean_token_accuracy": 0.18870871812105178,
      "num_tokens": 23123822.0,
      "step": 10365
    },
    {
      "entropy": 5.49096269607544,
      "epoch": 1.164654349413152,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004875397272991141,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1867375984787941,
      "num_tokens": 23134045.0,
      "step": 10370
    },
    {
      "entropy": 5.378085708618164,
      "epoch": 1.1652159263211097,
      "grad_norm": 1.1875,
      "learning_rate": 0.000487526548579656,
      "loss": 5.3334,
      "mean_token_accuracy": 0.1819787546992302,
      "num_tokens": 23144200.0,
      "step": 10375
    },
    {
      "entropy": 5.445252132415772,
      "epoch": 1.1657775032290671,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048751336309327026,
      "loss": 5.2569,
      "mean_token_accuracy": 0.19609665125608444,
      "num_tokens": 23155671.0,
      "step": 10380
    },
    {
      "entropy": 5.561450242996216,
      "epoch": 1.1663390801370248,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004875001708403766,
      "loss": 5.4035,
      "mean_token_accuracy": 0.18700147718191146,
      "num_tokens": 23166644.0,
      "step": 10385
    },
    {
      "entropy": 5.449707078933716,
      "epoch": 1.1669006570449822,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048748697182139503,
      "loss": 5.2835,
      "mean_token_accuracy": 0.1888731300830841,
      "num_tokens": 23177074.0,
      "step": 10390
    },
    {
      "entropy": 5.49114236831665,
      "epoch": 1.16746223395294,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004874737660367459,
      "loss": 5.4005,
      "mean_token_accuracy": 0.17823949903249742,
      "num_tokens": 23188724.0,
      "step": 10395
    },
    {
      "entropy": 5.44239993095398,
      "epoch": 1.1680238108608973,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048746055348684967,
      "loss": 5.3206,
      "mean_token_accuracy": 0.19123686254024505,
      "num_tokens": 23198459.0,
      "step": 10400
    },
    {
      "entropy": 5.512176275253296,
      "epoch": 1.168585387768855,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000487447334172127,
      "loss": 5.3234,
      "mean_token_accuracy": 0.18930345624685288,
      "num_tokens": 23208991.0,
      "step": 10405
    },
    {
      "entropy": 5.439350032806397,
      "epoch": 1.1691469646768125,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048743410809299877,
      "loss": 5.255,
      "mean_token_accuracy": 0.1920282632112503,
      "num_tokens": 23219309.0,
      "step": 10410
    },
    {
      "entropy": 5.447716093063354,
      "epoch": 1.1697085415847701,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004874208752498862,
      "loss": 5.3533,
      "mean_token_accuracy": 0.18778776079416276,
      "num_tokens": 23230551.0,
      "step": 10415
    },
    {
      "entropy": 5.527687406539917,
      "epoch": 1.1702701184927276,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004874076356432106,
      "loss": 5.4158,
      "mean_token_accuracy": 0.18665536791086196,
      "num_tokens": 23242799.0,
      "step": 10420
    },
    {
      "entropy": 5.558752107620239,
      "epoch": 1.1708316954006852,
      "grad_norm": 1.25,
      "learning_rate": 0.00048739438927339346,
      "loss": 5.4046,
      "mean_token_accuracy": 0.18513983488082886,
      "num_tokens": 23255112.0,
      "step": 10425
    },
    {
      "entropy": 5.405725288391113,
      "epoch": 1.1713932723086427,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048738113614085656,
      "loss": 5.3149,
      "mean_token_accuracy": 0.19400553703308104,
      "num_tokens": 23264800.0,
      "step": 10430
    },
    {
      "entropy": 5.412166881561279,
      "epoch": 1.1719548492166,
      "grad_norm": 1.046875,
      "learning_rate": 0.000487367876246022,
      "loss": 5.247,
      "mean_token_accuracy": 0.18509465754032134,
      "num_tokens": 23276672.0,
      "step": 10435
    },
    {
      "entropy": 5.457706880569458,
      "epoch": 1.1725164261245578,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004873546095893118,
      "loss": 5.2838,
      "mean_token_accuracy": 0.1924625962972641,
      "num_tokens": 23288282.0,
      "step": 10440
    },
    {
      "entropy": 5.464623498916626,
      "epoch": 1.1730780030325154,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004873413361711485,
      "loss": 5.3043,
      "mean_token_accuracy": 0.1897026091814041,
      "num_tokens": 23299078.0,
      "step": 10445
    },
    {
      "entropy": 5.391138410568237,
      "epoch": 1.1736395799404729,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048732805599195465,
      "loss": 5.3017,
      "mean_token_accuracy": 0.1927384614944458,
      "num_tokens": 23310036.0,
      "step": 10450
    },
    {
      "entropy": 5.407145595550537,
      "epoch": 1.1742011568484303,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048731476905215326,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19730805307626725,
      "num_tokens": 23321029.0,
      "step": 10455
    },
    {
      "entropy": 5.4644828796386715,
      "epoch": 1.174762733756388,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004873014753521673,
      "loss": 5.3347,
      "mean_token_accuracy": 0.18478613793849946,
      "num_tokens": 23331584.0,
      "step": 10460
    },
    {
      "entropy": 5.5647083759307865,
      "epoch": 1.1753243106643454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004872881748924199,
      "loss": 5.4194,
      "mean_token_accuracy": 0.17783666253089905,
      "num_tokens": 23343171.0,
      "step": 10465
    },
    {
      "entropy": 5.532898759841919,
      "epoch": 1.175885887572303,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004872748676733347,
      "loss": 5.3291,
      "mean_token_accuracy": 0.18714045137166976,
      "num_tokens": 23353047.0,
      "step": 10470
    },
    {
      "entropy": 5.462832355499268,
      "epoch": 1.1764474644802605,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048726155369533547,
      "loss": 5.3235,
      "mean_token_accuracy": 0.18652098178863524,
      "num_tokens": 23364447.0,
      "step": 10475
    },
    {
      "entropy": 5.521936130523682,
      "epoch": 1.1770090413882182,
      "grad_norm": 1.21875,
      "learning_rate": 0.000487248232958846,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18660225123167037,
      "num_tokens": 23375616.0,
      "step": 10480
    },
    {
      "entropy": 5.507867956161499,
      "epoch": 1.1775706182961756,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004872349054642904,
      "loss": 5.4418,
      "mean_token_accuracy": 0.18429977297782899,
      "num_tokens": 23388176.0,
      "step": 10485
    },
    {
      "entropy": 5.459119129180908,
      "epoch": 1.1781321952041333,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004872215712120932,
      "loss": 5.36,
      "mean_token_accuracy": 0.19346140772104264,
      "num_tokens": 23400563.0,
      "step": 10490
    },
    {
      "entropy": 5.500574350357056,
      "epoch": 1.1786937721120907,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004872082302026788,
      "loss": 5.3197,
      "mean_token_accuracy": 0.1850305199623108,
      "num_tokens": 23410740.0,
      "step": 10495
    },
    {
      "entropy": 5.458720588684082,
      "epoch": 1.1792553490200484,
      "grad_norm": 1.125,
      "learning_rate": 0.0004871948824364721,
      "loss": 5.3944,
      "mean_token_accuracy": 0.1873966410756111,
      "num_tokens": 23423086.0,
      "step": 10500
    },
    {
      "entropy": 5.517767381668091,
      "epoch": 1.1798169259280058,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000487181527913898,
      "loss": 5.3933,
      "mean_token_accuracy": 0.18787606209516525,
      "num_tokens": 23435869.0,
      "step": 10505
    },
    {
      "entropy": 5.472037982940674,
      "epoch": 1.1803785028359635,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048716816663538177,
      "loss": 5.2688,
      "mean_token_accuracy": 0.20028235018253326,
      "num_tokens": 23446635.0,
      "step": 10510
    },
    {
      "entropy": 5.424922513961792,
      "epoch": 1.180940079743921,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004871547986013488,
      "loss": 5.2882,
      "mean_token_accuracy": 0.19051352739334107,
      "num_tokens": 23458214.0,
      "step": 10515
    },
    {
      "entropy": 5.467806768417359,
      "epoch": 1.1815016566518786,
      "grad_norm": 1.125,
      "learning_rate": 0.0004871414238122248,
      "loss": 5.319,
      "mean_token_accuracy": 0.1956918492913246,
      "num_tokens": 23469383.0,
      "step": 10520
    },
    {
      "entropy": 5.394396638870239,
      "epoch": 1.182063233559836,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004871280422684355,
      "loss": 5.2641,
      "mean_token_accuracy": 0.1961277022957802,
      "num_tokens": 23480841.0,
      "step": 10525
    },
    {
      "entropy": 5.397922849655151,
      "epoch": 1.1826248104677934,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048711465397040726,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19275450557470322,
      "num_tokens": 23490796.0,
      "step": 10530
    },
    {
      "entropy": 5.442110633850097,
      "epoch": 1.183186387375751,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048710125891856604,
      "loss": 5.2584,
      "mean_token_accuracy": 0.18879759460687637,
      "num_tokens": 23501759.0,
      "step": 10535
    },
    {
      "entropy": 5.489678812026978,
      "epoch": 1.1837479642837088,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048708785711333845,
      "loss": 5.3096,
      "mean_token_accuracy": 0.19035088866949082,
      "num_tokens": 23512209.0,
      "step": 10540
    },
    {
      "entropy": 5.489083290100098,
      "epoch": 1.1843095411916662,
      "grad_norm": 1.03125,
      "learning_rate": 0.00048707444855515124,
      "loss": 5.4,
      "mean_token_accuracy": 0.18393512666225434,
      "num_tokens": 23524316.0,
      "step": 10545
    },
    {
      "entropy": 5.45975170135498,
      "epoch": 1.1848711180996236,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004870610332444314,
      "loss": 5.3622,
      "mean_token_accuracy": 0.19088199287652968,
      "num_tokens": 23536199.0,
      "step": 10550
    },
    {
      "entropy": 5.490462684631348,
      "epoch": 1.1854326950075813,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004870476111816059,
      "loss": 5.3725,
      "mean_token_accuracy": 0.18413449823856354,
      "num_tokens": 23548034.0,
      "step": 10555
    },
    {
      "entropy": 5.480147743225098,
      "epoch": 1.1859942719155387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048703418236710234,
      "loss": 5.304,
      "mean_token_accuracy": 0.18967850059270858,
      "num_tokens": 23559280.0,
      "step": 10560
    },
    {
      "entropy": 5.45183572769165,
      "epoch": 1.1865558488234964,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004870207468013481,
      "loss": 5.299,
      "mean_token_accuracy": 0.18800936341285707,
      "num_tokens": 23569163.0,
      "step": 10565
    },
    {
      "entropy": 5.497716760635376,
      "epoch": 1.1871174257314538,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004870073044847711,
      "loss": 5.3638,
      "mean_token_accuracy": 0.190348057448864,
      "num_tokens": 23580526.0,
      "step": 10570
    },
    {
      "entropy": 5.39793291091919,
      "epoch": 1.1876790026394115,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004869938554177993,
      "loss": 5.2552,
      "mean_token_accuracy": 0.18700382262468337,
      "num_tokens": 23590552.0,
      "step": 10575
    },
    {
      "entropy": 5.319877290725708,
      "epoch": 1.188240579547369,
      "grad_norm": 1.25,
      "learning_rate": 0.00048698039960086084,
      "loss": 5.1425,
      "mean_token_accuracy": 0.1934616208076477,
      "num_tokens": 23601889.0,
      "step": 10580
    },
    {
      "entropy": 5.471102046966553,
      "epoch": 1.1888021564553266,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004869669370343843,
      "loss": 5.3973,
      "mean_token_accuracy": 0.1950755625963211,
      "num_tokens": 23613833.0,
      "step": 10585
    },
    {
      "entropy": 5.543602371215821,
      "epoch": 1.189363733363284,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004869534677187983,
      "loss": 5.3626,
      "mean_token_accuracy": 0.1897484079003334,
      "num_tokens": 23625404.0,
      "step": 10590
    },
    {
      "entropy": 5.398644065856933,
      "epoch": 1.1899253102712417,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004869399916545317,
      "loss": 5.231,
      "mean_token_accuracy": 0.19434747993946075,
      "num_tokens": 23635670.0,
      "step": 10595
    },
    {
      "entropy": 5.385092449188233,
      "epoch": 1.1904868871791991,
      "grad_norm": 1.125,
      "learning_rate": 0.0004869265088420135,
      "loss": 5.2273,
      "mean_token_accuracy": 0.19717648923397063,
      "num_tokens": 23646893.0,
      "step": 10600
    },
    {
      "entropy": 5.393044805526733,
      "epoch": 1.1910484640871568,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004869130192816731,
      "loss": 5.3483,
      "mean_token_accuracy": 0.1863220676779747,
      "num_tokens": 23657733.0,
      "step": 10605
    },
    {
      "entropy": 5.398750305175781,
      "epoch": 1.1916100409951143,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048689952297394,
      "loss": 5.1986,
      "mean_token_accuracy": 0.19344332069158554,
      "num_tokens": 23668100.0,
      "step": 10610
    },
    {
      "entropy": 5.525930118560791,
      "epoch": 1.192171617903072,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004868860199192438,
      "loss": 5.3295,
      "mean_token_accuracy": 0.18708044290542603,
      "num_tokens": 23678295.0,
      "step": 10615
    },
    {
      "entropy": 5.414384031295777,
      "epoch": 1.1927331948110294,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004868725101180146,
      "loss": 5.1878,
      "mean_token_accuracy": 0.1949734643101692,
      "num_tokens": 23689310.0,
      "step": 10620
    },
    {
      "entropy": 5.398082637786866,
      "epoch": 1.1932947717189868,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004868589935706826,
      "loss": 5.3447,
      "mean_token_accuracy": 0.18504620790481568,
      "num_tokens": 23699475.0,
      "step": 10625
    },
    {
      "entropy": 5.53235239982605,
      "epoch": 1.1938563486269445,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048684547027767793,
      "loss": 5.3378,
      "mean_token_accuracy": 0.1882213145494461,
      "num_tokens": 23710186.0,
      "step": 10630
    },
    {
      "entropy": 5.3258768081665036,
      "epoch": 1.1944179255349021,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004868319402394314,
      "loss": 5.1655,
      "mean_token_accuracy": 0.19515339732170106,
      "num_tokens": 23720449.0,
      "step": 10635
    },
    {
      "entropy": 5.502135848999023,
      "epoch": 1.1949795024428596,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048681840345637375,
      "loss": 5.4174,
      "mean_token_accuracy": 0.18592655062675476,
      "num_tokens": 23732295.0,
      "step": 10640
    },
    {
      "entropy": 5.448860836029053,
      "epoch": 1.195541079350817,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048680485992893597,
      "loss": 5.3326,
      "mean_token_accuracy": 0.18566824644804,
      "num_tokens": 23743616.0,
      "step": 10645
    },
    {
      "entropy": 5.516355848312378,
      "epoch": 1.1961026562587747,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048679130965754926,
      "loss": 5.3476,
      "mean_token_accuracy": 0.18490096032619477,
      "num_tokens": 23755288.0,
      "step": 10650
    },
    {
      "entropy": 5.4344583511352536,
      "epoch": 1.196664233166732,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004867777526426451,
      "loss": 5.2086,
      "mean_token_accuracy": 0.19361613541841508,
      "num_tokens": 23765916.0,
      "step": 10655
    },
    {
      "entropy": 5.320996284484863,
      "epoch": 1.1972258100746898,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048676418888465525,
      "loss": 5.1861,
      "mean_token_accuracy": 0.19435495883226395,
      "num_tokens": 23778891.0,
      "step": 10660
    },
    {
      "entropy": 5.454932785034179,
      "epoch": 1.1977873869826472,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004867506183840114,
      "loss": 5.3229,
      "mean_token_accuracy": 0.19375743865966796,
      "num_tokens": 23790402.0,
      "step": 10665
    },
    {
      "entropy": 5.494229364395141,
      "epoch": 1.1983489638906049,
      "grad_norm": 1.546875,
      "learning_rate": 0.0004867370411411459,
      "loss": 5.2689,
      "mean_token_accuracy": 0.19692872166633607,
      "num_tokens": 23801928.0,
      "step": 10670
    },
    {
      "entropy": 5.523255634307861,
      "epoch": 1.1989105407985623,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048672345715649066,
      "loss": 5.3559,
      "mean_token_accuracy": 0.17944910526275634,
      "num_tokens": 23812302.0,
      "step": 10675
    },
    {
      "entropy": 5.468920755386352,
      "epoch": 1.19947211770652,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004867098664304785,
      "loss": 5.3232,
      "mean_token_accuracy": 0.19833283722400666,
      "num_tokens": 23824033.0,
      "step": 10680
    },
    {
      "entropy": 5.422686243057251,
      "epoch": 1.2000336946144774,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004866962689635422,
      "loss": 5.2695,
      "mean_token_accuracy": 0.19360737949609758,
      "num_tokens": 23834907.0,
      "step": 10685
    },
    {
      "entropy": 5.511562442779541,
      "epoch": 1.200595271522435,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048668266475611447,
      "loss": 5.4555,
      "mean_token_accuracy": 0.19213903397321702,
      "num_tokens": 23846418.0,
      "step": 10690
    },
    {
      "entropy": 5.568163251876831,
      "epoch": 1.2011568484303925,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004866690538086286,
      "loss": 5.3861,
      "mean_token_accuracy": 0.18392764925956726,
      "num_tokens": 23858423.0,
      "step": 10695
    },
    {
      "entropy": 5.495675897598266,
      "epoch": 1.2017184253383502,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048665543612151803,
      "loss": 5.2856,
      "mean_token_accuracy": 0.1922287255525589,
      "num_tokens": 23868340.0,
      "step": 10700
    },
    {
      "entropy": 5.404595184326172,
      "epoch": 1.2022800022463076,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048664181169521625,
      "loss": 5.2339,
      "mean_token_accuracy": 0.1947616904973984,
      "num_tokens": 23878752.0,
      "step": 10705
    },
    {
      "entropy": 5.4817146301269535,
      "epoch": 1.2028415791542653,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004866281805301571,
      "loss": 5.2756,
      "mean_token_accuracy": 0.19479703307151794,
      "num_tokens": 23889374.0,
      "step": 10710
    },
    {
      "entropy": 5.493065309524536,
      "epoch": 1.2034031560622227,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004866145426267746,
      "loss": 5.3337,
      "mean_token_accuracy": 0.19232070744037627,
      "num_tokens": 23901651.0,
      "step": 10715
    },
    {
      "entropy": 5.371082305908203,
      "epoch": 1.2039647329701804,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000486600897985503,
      "loss": 5.2648,
      "mean_token_accuracy": 0.19026408046483995,
      "num_tokens": 23911818.0,
      "step": 10720
    },
    {
      "entropy": 5.480171823501587,
      "epoch": 1.2045263098781378,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048658724660677674,
      "loss": 5.3014,
      "mean_token_accuracy": 0.19395873546600342,
      "num_tokens": 23923840.0,
      "step": 10725
    },
    {
      "entropy": 5.406844663619995,
      "epoch": 1.2050878867860955,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004865735884910305,
      "loss": 5.2053,
      "mean_token_accuracy": 0.19550656527280807,
      "num_tokens": 23935406.0,
      "step": 10730
    },
    {
      "entropy": 5.451653146743775,
      "epoch": 1.205649463694053,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004865599236386991,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19326698035001755,
      "num_tokens": 23946350.0,
      "step": 10735
    },
    {
      "entropy": 5.312716627120972,
      "epoch": 1.2062110406020103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004865462520502177,
      "loss": 5.2493,
      "mean_token_accuracy": 0.1963890239596367,
      "num_tokens": 23956315.0,
      "step": 10740
    },
    {
      "entropy": 5.40750732421875,
      "epoch": 1.206772617509968,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004865325737260216,
      "loss": 5.1635,
      "mean_token_accuracy": 0.1909998580813408,
      "num_tokens": 23966649.0,
      "step": 10745
    },
    {
      "entropy": 5.550435543060303,
      "epoch": 1.2073341944179254,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048651888866654624,
      "loss": 5.4358,
      "mean_token_accuracy": 0.18272399306297302,
      "num_tokens": 23977674.0,
      "step": 10750
    },
    {
      "entropy": 5.456652879714966,
      "epoch": 1.207895771325883,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004865051968722275,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1892181560397148,
      "num_tokens": 23987985.0,
      "step": 10755
    },
    {
      "entropy": 5.491566848754883,
      "epoch": 1.2084573482338405,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048649149834350126,
      "loss": 5.4234,
      "mean_token_accuracy": 0.18354978859424592,
      "num_tokens": 23998730.0,
      "step": 10760
    },
    {
      "entropy": 5.420482730865478,
      "epoch": 1.2090189251417982,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004864777930808037,
      "loss": 5.2479,
      "mean_token_accuracy": 0.18344485014677048,
      "num_tokens": 24008743.0,
      "step": 10765
    },
    {
      "entropy": 5.450106430053711,
      "epoch": 1.2095805020497556,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004864640810845711,
      "loss": 5.3022,
      "mean_token_accuracy": 0.19231080114841462,
      "num_tokens": 24019805.0,
      "step": 10770
    },
    {
      "entropy": 5.46649923324585,
      "epoch": 1.2101420789577133,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004864503623552402,
      "loss": 5.2927,
      "mean_token_accuracy": 0.18949525952339172,
      "num_tokens": 24031437.0,
      "step": 10775
    },
    {
      "entropy": 5.474605321884155,
      "epoch": 1.2107036558656707,
      "grad_norm": 1.125,
      "learning_rate": 0.0004864366368932478,
      "loss": 5.4003,
      "mean_token_accuracy": 0.18210068792104722,
      "num_tokens": 24043922.0,
      "step": 10780
    },
    {
      "entropy": 5.468896150588989,
      "epoch": 1.2112652327736284,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048642290469903083,
      "loss": 5.222,
      "mean_token_accuracy": 0.19848127365112306,
      "num_tokens": 24054755.0,
      "step": 10785
    },
    {
      "entropy": 5.407658338546753,
      "epoch": 1.2118268096815858,
      "grad_norm": 1.046875,
      "learning_rate": 0.00048640916577302664,
      "loss": 5.2372,
      "mean_token_accuracy": 0.18253839612007142,
      "num_tokens": 24067347.0,
      "step": 10790
    },
    {
      "entropy": 5.496824598312378,
      "epoch": 1.2123883865895435,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048639542011567243,
      "loss": 5.2989,
      "mean_token_accuracy": 0.1892787605524063,
      "num_tokens": 24078938.0,
      "step": 10795
    },
    {
      "entropy": 5.5106813430786135,
      "epoch": 1.212949963497501,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004863816677274062,
      "loss": 5.3106,
      "mean_token_accuracy": 0.18729422241449356,
      "num_tokens": 24090275.0,
      "step": 10800
    },
    {
      "entropy": 5.416562557220459,
      "epoch": 1.2135115404054586,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048636790860866576,
      "loss": 5.3268,
      "mean_token_accuracy": 0.193812794983387,
      "num_tokens": 24101518.0,
      "step": 10805
    },
    {
      "entropy": 5.473560333251953,
      "epoch": 1.214073117313416,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004863541427598889,
      "loss": 5.4127,
      "mean_token_accuracy": 0.18143971264362335,
      "num_tokens": 24113320.0,
      "step": 10810
    },
    {
      "entropy": 5.41842131614685,
      "epoch": 1.2146346942213737,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004863403701815144,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19541742652654648,
      "num_tokens": 24124842.0,
      "step": 10815
    },
    {
      "entropy": 5.506654262542725,
      "epoch": 1.2151962711293312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004863265908739805,
      "loss": 5.3807,
      "mean_token_accuracy": 0.18791324347257615,
      "num_tokens": 24136391.0,
      "step": 10820
    },
    {
      "entropy": 5.521567010879517,
      "epoch": 1.2157578480372888,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004863128048377259,
      "loss": 5.3488,
      "mean_token_accuracy": 0.1879984974861145,
      "num_tokens": 24147194.0,
      "step": 10825
    },
    {
      "entropy": 5.4770787239074705,
      "epoch": 1.2163194249452463,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004862990120731896,
      "loss": 5.4148,
      "mean_token_accuracy": 0.182609261572361,
      "num_tokens": 24158170.0,
      "step": 10830
    },
    {
      "entropy": 5.461014938354492,
      "epoch": 1.2168810018532037,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048628521258081096,
      "loss": 5.2462,
      "mean_token_accuracy": 0.18959937393665313,
      "num_tokens": 24168671.0,
      "step": 10835
    },
    {
      "entropy": 5.439473819732666,
      "epoch": 1.2174425787611614,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048627140636102906,
      "loss": 5.2876,
      "mean_token_accuracy": 0.1922362893819809,
      "num_tokens": 24179898.0,
      "step": 10840
    },
    {
      "entropy": 5.365127849578857,
      "epoch": 1.2180041556691188,
      "grad_norm": 1.125,
      "learning_rate": 0.0004862575934142837,
      "loss": 5.275,
      "mean_token_accuracy": 0.19409412443637847,
      "num_tokens": 24190622.0,
      "step": 10845
    },
    {
      "entropy": 5.624100637435913,
      "epoch": 1.2185657325770765,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004862437737410147,
      "loss": 5.4444,
      "mean_token_accuracy": 0.18157976120710373,
      "num_tokens": 24202526.0,
      "step": 10850
    },
    {
      "entropy": 5.418727684020996,
      "epoch": 1.219127309485034,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004862299473416619,
      "loss": 5.2707,
      "mean_token_accuracy": 0.19151708483695984,
      "num_tokens": 24213186.0,
      "step": 10855
    },
    {
      "entropy": 5.381842422485351,
      "epoch": 1.2196888863929916,
      "grad_norm": 1.125,
      "learning_rate": 0.00048621611421666573,
      "loss": 5.2918,
      "mean_token_accuracy": 0.19073220640420913,
      "num_tokens": 24225490.0,
      "step": 10860
    },
    {
      "entropy": 5.466461229324341,
      "epoch": 1.220250463300949,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004862022743664665,
      "loss": 5.2689,
      "mean_token_accuracy": 0.1927856519818306,
      "num_tokens": 24236653.0,
      "step": 10865
    },
    {
      "entropy": 5.4559417247772215,
      "epoch": 1.2208120402089067,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00048618842779150494,
      "loss": 5.2481,
      "mean_token_accuracy": 0.1990346670150757,
      "num_tokens": 24247843.0,
      "step": 10870
    },
    {
      "entropy": 5.397953939437866,
      "epoch": 1.221373617116864,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048617457449222205,
      "loss": 5.3187,
      "mean_token_accuracy": 0.19366680234670638,
      "num_tokens": 24258332.0,
      "step": 10875
    },
    {
      "entropy": 5.520880270004272,
      "epoch": 1.2219351940248218,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004861607144690587,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19346436113119125,
      "num_tokens": 24268685.0,
      "step": 10880
    },
    {
      "entropy": 5.50979061126709,
      "epoch": 1.2224967709327792,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004861468477224563,
      "loss": 5.3293,
      "mean_token_accuracy": 0.18430805057287217,
      "num_tokens": 24280660.0,
      "step": 10885
    },
    {
      "entropy": 5.4520392417907715,
      "epoch": 1.2230583478407369,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048613297425285635,
      "loss": 5.2736,
      "mean_token_accuracy": 0.19194200187921523,
      "num_tokens": 24291649.0,
      "step": 10890
    },
    {
      "entropy": 5.431670236587524,
      "epoch": 1.2236199247486943,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004861190940607007,
      "loss": 5.2791,
      "mean_token_accuracy": 0.1915912300348282,
      "num_tokens": 24303108.0,
      "step": 10895
    },
    {
      "entropy": 5.392286491394043,
      "epoch": 1.224181501656652,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004861052071464312,
      "loss": 5.3088,
      "mean_token_accuracy": 0.18251196593046187,
      "num_tokens": 24315939.0,
      "step": 10900
    },
    {
      "entropy": 5.430204057693482,
      "epoch": 1.2247430785646094,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048609131351048994,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1899013489484787,
      "num_tokens": 24326916.0,
      "step": 10905
    },
    {
      "entropy": 5.472850608825683,
      "epoch": 1.225304655472567,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004860774131533195,
      "loss": 5.3397,
      "mean_token_accuracy": 0.1885451301932335,
      "num_tokens": 24337790.0,
      "step": 10910
    },
    {
      "entropy": 5.390128707885742,
      "epoch": 1.2258662323805245,
      "grad_norm": 1.25,
      "learning_rate": 0.00048606350607536235,
      "loss": 5.1527,
      "mean_token_accuracy": 0.2028306245803833,
      "num_tokens": 24347930.0,
      "step": 10915
    },
    {
      "entropy": 5.4161817073822025,
      "epoch": 1.2264278092884822,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004860495922770612,
      "loss": 5.2619,
      "mean_token_accuracy": 0.19034540355205537,
      "num_tokens": 24359213.0,
      "step": 10920
    },
    {
      "entropy": 5.368238496780395,
      "epoch": 1.2269893861964396,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048603567175885926,
      "loss": 5.2314,
      "mean_token_accuracy": 0.18936524093151091,
      "num_tokens": 24371000.0,
      "step": 10925
    },
    {
      "entropy": 5.4050709247589115,
      "epoch": 1.227550963104397,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004860217445211997,
      "loss": 5.1722,
      "mean_token_accuracy": 0.1944772258400917,
      "num_tokens": 24381301.0,
      "step": 10930
    },
    {
      "entropy": 5.506228065490722,
      "epoch": 1.2281125400123547,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004860078105645259,
      "loss": 5.4127,
      "mean_token_accuracy": 0.17683492451906205,
      "num_tokens": 24391808.0,
      "step": 10935
    },
    {
      "entropy": 5.517259931564331,
      "epoch": 1.2286741169203121,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004859938698892816,
      "loss": 5.5139,
      "mean_token_accuracy": 0.18332257568836213,
      "num_tokens": 24404220.0,
      "step": 10940
    },
    {
      "entropy": 5.501758766174317,
      "epoch": 1.2292356938282698,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048597992249591065,
      "loss": 5.3143,
      "mean_token_accuracy": 0.19299614131450654,
      "num_tokens": 24414906.0,
      "step": 10945
    },
    {
      "entropy": 5.453164577484131,
      "epoch": 1.2297972707362272,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048596596838485704,
      "loss": 5.2962,
      "mean_token_accuracy": 0.18702104091644287,
      "num_tokens": 24426469.0,
      "step": 10950
    },
    {
      "entropy": 5.343540000915527,
      "epoch": 1.230358847644185,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004859520075565654,
      "loss": 5.2776,
      "mean_token_accuracy": 0.18950754553079605,
      "num_tokens": 24437683.0,
      "step": 10955
    },
    {
      "entropy": 5.381892347335816,
      "epoch": 1.2309204245521423,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048593804001147986,
      "loss": 5.1715,
      "mean_token_accuracy": 0.19221088886260987,
      "num_tokens": 24449038.0,
      "step": 10960
    },
    {
      "entropy": 5.4757568359375,
      "epoch": 1.2314820014601,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004859240657500453,
      "loss": 5.3034,
      "mean_token_accuracy": 0.19335017949342728,
      "num_tokens": 24460228.0,
      "step": 10965
    },
    {
      "entropy": 5.43294243812561,
      "epoch": 1.2320435783680574,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048591008477270675,
      "loss": 5.3549,
      "mean_token_accuracy": 0.18759188950061798,
      "num_tokens": 24471222.0,
      "step": 10970
    },
    {
      "entropy": 5.41448221206665,
      "epoch": 1.232605155276015,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048589609707990925,
      "loss": 5.3026,
      "mean_token_accuracy": 0.18707839548587799,
      "num_tokens": 24482127.0,
      "step": 10975
    },
    {
      "entropy": 5.510404109954834,
      "epoch": 1.2331667321839725,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048588210267209823,
      "loss": 5.2695,
      "mean_token_accuracy": 0.19078285843133927,
      "num_tokens": 24492963.0,
      "step": 10980
    },
    {
      "entropy": 5.516666746139526,
      "epoch": 1.2337283090919302,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048586810154971935,
      "loss": 5.3064,
      "mean_token_accuracy": 0.19331692159175873,
      "num_tokens": 24502885.0,
      "step": 10985
    },
    {
      "entropy": 5.5171865940094,
      "epoch": 1.2342898859998876,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048585409371321826,
      "loss": 5.4155,
      "mean_token_accuracy": 0.18873243778944016,
      "num_tokens": 24514990.0,
      "step": 10990
    },
    {
      "entropy": 5.502025985717774,
      "epoch": 1.2348514629078453,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048584007916304106,
      "loss": 5.3234,
      "mean_token_accuracy": 0.18487985283136368,
      "num_tokens": 24525555.0,
      "step": 10995
    },
    {
      "entropy": 5.515440940856934,
      "epoch": 1.2354130398158027,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048582605789963395,
      "loss": 5.3591,
      "mean_token_accuracy": 0.1826901316642761,
      "num_tokens": 24536350.0,
      "step": 11000
    },
    {
      "entropy": 5.553280448913574,
      "epoch": 1.2359746167237604,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004858120299234434,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18822159618139267,
      "num_tokens": 24547756.0,
      "step": 11005
    },
    {
      "entropy": 5.380481958389282,
      "epoch": 1.2365361936317179,
      "grad_norm": 1.0,
      "learning_rate": 0.00048579799523491607,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19720670729875564,
      "num_tokens": 24559897.0,
      "step": 11010
    },
    {
      "entropy": 5.415151214599609,
      "epoch": 1.2370977705396755,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004857839538344988,
      "loss": 5.272,
      "mean_token_accuracy": 0.18695087879896163,
      "num_tokens": 24570168.0,
      "step": 11015
    },
    {
      "entropy": 5.48857741355896,
      "epoch": 1.237659347447633,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048576990572263875,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19748227000236512,
      "num_tokens": 24580918.0,
      "step": 11020
    },
    {
      "entropy": 5.41733341217041,
      "epoch": 1.2382209243555904,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004857558508997831,
      "loss": 5.2592,
      "mean_token_accuracy": 0.18681925237178804,
      "num_tokens": 24592378.0,
      "step": 11025
    },
    {
      "entropy": 5.5000134944915775,
      "epoch": 1.238782501263548,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004857417893663794,
      "loss": 5.3514,
      "mean_token_accuracy": 0.19073790311813354,
      "num_tokens": 24603666.0,
      "step": 11030
    },
    {
      "entropy": 5.44213171005249,
      "epoch": 1.2393440781715055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004857277211228754,
      "loss": 5.3064,
      "mean_token_accuracy": 0.18842377364635468,
      "num_tokens": 24615299.0,
      "step": 11035
    },
    {
      "entropy": 5.424975919723511,
      "epoch": 1.2399056550794632,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004857136461697191,
      "loss": 5.2718,
      "mean_token_accuracy": 0.1951408326625824,
      "num_tokens": 24626138.0,
      "step": 11040
    },
    {
      "entropy": 5.401485586166382,
      "epoch": 1.2404672319874206,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048569956450735853,
      "loss": 5.2521,
      "mean_token_accuracy": 0.1900665804743767,
      "num_tokens": 24636720.0,
      "step": 11045
    },
    {
      "entropy": 5.491612339019776,
      "epoch": 1.2410288088953783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048568547613624206,
      "loss": 5.2763,
      "mean_token_accuracy": 0.18914327621459961,
      "num_tokens": 24648147.0,
      "step": 11050
    },
    {
      "entropy": 5.410216045379639,
      "epoch": 1.2415903858033357,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048567138105681833,
      "loss": 5.3188,
      "mean_token_accuracy": 0.18939395099878312,
      "num_tokens": 24659825.0,
      "step": 11055
    },
    {
      "entropy": 5.416470289230347,
      "epoch": 1.2421519627112934,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048565727926953617,
      "loss": 5.2747,
      "mean_token_accuracy": 0.19203497022390364,
      "num_tokens": 24671583.0,
      "step": 11060
    },
    {
      "entropy": 5.3895632266998295,
      "epoch": 1.2427135396192508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004856431707748445,
      "loss": 5.2171,
      "mean_token_accuracy": 0.19326931089162827,
      "num_tokens": 24682034.0,
      "step": 11065
    },
    {
      "entropy": 5.446405649185181,
      "epoch": 1.2432751165272085,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004856290555731926,
      "loss": 5.2969,
      "mean_token_accuracy": 0.1903885543346405,
      "num_tokens": 24693624.0,
      "step": 11070
    },
    {
      "entropy": 5.522940254211425,
      "epoch": 1.243836693435166,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004856149336650299,
      "loss": 5.3042,
      "mean_token_accuracy": 0.1911284103989601,
      "num_tokens": 24705340.0,
      "step": 11075
    },
    {
      "entropy": 5.507666826248169,
      "epoch": 1.2443982703431236,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004856008050508059,
      "loss": 5.3778,
      "mean_token_accuracy": 0.1818677768111229,
      "num_tokens": 24717115.0,
      "step": 11080
    },
    {
      "entropy": 5.440750980377198,
      "epoch": 1.244959847251081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004855866697309707,
      "loss": 5.2029,
      "mean_token_accuracy": 0.19436603486537934,
      "num_tokens": 24727768.0,
      "step": 11085
    },
    {
      "entropy": 5.478287649154663,
      "epoch": 1.2455214241590387,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004855725277059742,
      "loss": 5.3659,
      "mean_token_accuracy": 0.18264859467744826,
      "num_tokens": 24740054.0,
      "step": 11090
    },
    {
      "entropy": 5.515737247467041,
      "epoch": 1.246083001066996,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004855583789762668,
      "loss": 5.247,
      "mean_token_accuracy": 0.19351306259632112,
      "num_tokens": 24750658.0,
      "step": 11095
    },
    {
      "entropy": 5.548893022537231,
      "epoch": 1.2466445779749538,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000485544223542299,
      "loss": 5.4536,
      "mean_token_accuracy": 0.1879055991768837,
      "num_tokens": 24762730.0,
      "step": 11100
    },
    {
      "entropy": 5.341693353652954,
      "epoch": 1.2472061548829112,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004855300614045214,
      "loss": 5.1932,
      "mean_token_accuracy": 0.19531592577695847,
      "num_tokens": 24773865.0,
      "step": 11105
    },
    {
      "entropy": 5.392433977127075,
      "epoch": 1.2477677317908689,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048551589256338497,
      "loss": 5.295,
      "mean_token_accuracy": 0.19436483234167098,
      "num_tokens": 24783030.0,
      "step": 11110
    },
    {
      "entropy": 5.425662326812744,
      "epoch": 1.2483293086988263,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048550171701934087,
      "loss": 5.2417,
      "mean_token_accuracy": 0.19768193662166594,
      "num_tokens": 24793177.0,
      "step": 11115
    },
    {
      "entropy": 5.517284297943116,
      "epoch": 1.2488908856067837,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048548753477284053,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19410386085510253,
      "num_tokens": 24804796.0,
      "step": 11120
    },
    {
      "entropy": 5.5148420333862305,
      "epoch": 1.2494524625147414,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004854733458243354,
      "loss": 5.3328,
      "mean_token_accuracy": 0.18776785880327224,
      "num_tokens": 24816806.0,
      "step": 11125
    },
    {
      "entropy": 5.403828048706055,
      "epoch": 1.250014039422699,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004854591501742773,
      "loss": 5.2887,
      "mean_token_accuracy": 0.19395187199115754,
      "num_tokens": 24826888.0,
      "step": 11130
    },
    {
      "entropy": 5.454483222961426,
      "epoch": 1.2505756163306565,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048544494782311825,
      "loss": 5.2379,
      "mean_token_accuracy": 0.1941351369023323,
      "num_tokens": 24837731.0,
      "step": 11135
    },
    {
      "entropy": 5.410605335235596,
      "epoch": 1.251137193238614,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048543073877131037,
      "loss": 5.2613,
      "mean_token_accuracy": 0.19203507006168366,
      "num_tokens": 24847900.0,
      "step": 11140
    },
    {
      "entropy": 5.433638858795166,
      "epoch": 1.2516987701465716,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048541652301930627,
      "loss": 5.3377,
      "mean_token_accuracy": 0.18702833503484725,
      "num_tokens": 24858878.0,
      "step": 11145
    },
    {
      "entropy": 5.465994501113892,
      "epoch": 1.252260347054529,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004854023005675585,
      "loss": 5.2876,
      "mean_token_accuracy": 0.1894189640879631,
      "num_tokens": 24870051.0,
      "step": 11150
    },
    {
      "entropy": 5.445054960250855,
      "epoch": 1.2528219239624867,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004853880714165197,
      "loss": 5.3806,
      "mean_token_accuracy": 0.18749876469373702,
      "num_tokens": 24881671.0,
      "step": 11155
    },
    {
      "entropy": 5.498525667190552,
      "epoch": 1.2533835008704441,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048537383556664315,
      "loss": 5.2288,
      "mean_token_accuracy": 0.18579935133457184,
      "num_tokens": 24892944.0,
      "step": 11160
    },
    {
      "entropy": 5.455405616760254,
      "epoch": 1.2539450777784018,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004853595930183822,
      "loss": 5.3277,
      "mean_token_accuracy": 0.18537677079439163,
      "num_tokens": 24903346.0,
      "step": 11165
    },
    {
      "entropy": 5.435509586334229,
      "epoch": 1.2545066546863592,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004853453437721901,
      "loss": 5.309,
      "mean_token_accuracy": 0.1889562875032425,
      "num_tokens": 24913904.0,
      "step": 11170
    },
    {
      "entropy": 5.456475830078125,
      "epoch": 1.255068231594317,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004853310878285208,
      "loss": 5.2536,
      "mean_token_accuracy": 0.19424795508384704,
      "num_tokens": 24924635.0,
      "step": 11175
    },
    {
      "entropy": 5.493177223205566,
      "epoch": 1.2556298085022743,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048531682518782797,
      "loss": 5.4469,
      "mean_token_accuracy": 0.18727472871541978,
      "num_tokens": 24935466.0,
      "step": 11180
    },
    {
      "entropy": 5.536214828491211,
      "epoch": 1.256191385410232,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048530255585056594,
      "loss": 5.3958,
      "mean_token_accuracy": 0.1839703157544136,
      "num_tokens": 24948348.0,
      "step": 11185
    },
    {
      "entropy": 5.469631385803223,
      "epoch": 1.2567529623181894,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004852882798171889,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1951408088207245,
      "num_tokens": 24959242.0,
      "step": 11190
    },
    {
      "entropy": 5.371219491958618,
      "epoch": 1.2573145392261469,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004852739970881515,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19454684853553772,
      "num_tokens": 24969855.0,
      "step": 11195
    },
    {
      "entropy": 5.419975805282593,
      "epoch": 1.2578761161341045,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048525970766390847,
      "loss": 5.2366,
      "mean_token_accuracy": 0.19612980335950853,
      "num_tokens": 24980378.0,
      "step": 11200
    },
    {
      "entropy": 5.46827802658081,
      "epoch": 1.2584376930420622,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004852454115449148,
      "loss": 5.3111,
      "mean_token_accuracy": 0.18501074463129044,
      "num_tokens": 24990646.0,
      "step": 11205
    },
    {
      "entropy": 5.570227384567261,
      "epoch": 1.2589992699500197,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004852311087316257,
      "loss": 5.4081,
      "mean_token_accuracy": 0.18361775428056717,
      "num_tokens": 25002644.0,
      "step": 11210
    },
    {
      "entropy": 5.498043537139893,
      "epoch": 1.259560846857977,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004852167992244966,
      "loss": 5.3182,
      "mean_token_accuracy": 0.1940513327717781,
      "num_tokens": 25012635.0,
      "step": 11215
    },
    {
      "entropy": 5.470759153366089,
      "epoch": 1.2601224237659348,
      "grad_norm": 1.3046875,
      "learning_rate": 0.000485202483023983,
      "loss": 5.4338,
      "mean_token_accuracy": 0.181339068710804,
      "num_tokens": 25024289.0,
      "step": 11220
    },
    {
      "entropy": 5.475177049636841,
      "epoch": 1.2606840006738924,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048518816013054083,
      "loss": 5.1931,
      "mean_token_accuracy": 0.19016578942537307,
      "num_tokens": 25035316.0,
      "step": 11225
    },
    {
      "entropy": 5.424777173995972,
      "epoch": 1.2612455775818499,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00048517383054462617,
      "loss": 5.2482,
      "mean_token_accuracy": 0.19294827282428742,
      "num_tokens": 25046011.0,
      "step": 11230
    },
    {
      "entropy": 5.449552965164185,
      "epoch": 1.2618071544898073,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004851594942666952,
      "loss": 5.2448,
      "mean_token_accuracy": 0.1918046534061432,
      "num_tokens": 25057402.0,
      "step": 11235
    },
    {
      "entropy": 5.4338935852050785,
      "epoch": 1.262368731397765,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048514515129720445,
      "loss": 5.2696,
      "mean_token_accuracy": 0.19318236261606217,
      "num_tokens": 25068287.0,
      "step": 11240
    },
    {
      "entropy": 5.389315986633301,
      "epoch": 1.2629303083057224,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004851308016366105,
      "loss": 5.22,
      "mean_token_accuracy": 0.1940531313419342,
      "num_tokens": 25079534.0,
      "step": 11245
    },
    {
      "entropy": 5.3505539894104,
      "epoch": 1.26349188521368,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048511644528537043,
      "loss": 5.2093,
      "mean_token_accuracy": 0.19275834262371064,
      "num_tokens": 25090825.0,
      "step": 11250
    },
    {
      "entropy": 5.479386663436889,
      "epoch": 1.2640534621216375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004851020822439412,
      "loss": 5.3138,
      "mean_token_accuracy": 0.18911905139684676,
      "num_tokens": 25101859.0,
      "step": 11255
    },
    {
      "entropy": 5.416166830062866,
      "epoch": 1.2646150390295952,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048508771251278015,
      "loss": 5.2307,
      "mean_token_accuracy": 0.19599294364452363,
      "num_tokens": 25111416.0,
      "step": 11260
    },
    {
      "entropy": 5.470286464691162,
      "epoch": 1.2651766159375526,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004850733360923449,
      "loss": 5.33,
      "mean_token_accuracy": 0.19135213196277617,
      "num_tokens": 25122129.0,
      "step": 11265
    },
    {
      "entropy": 5.438811779022217,
      "epoch": 1.2657381928455103,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004850589529830931,
      "loss": 5.2997,
      "mean_token_accuracy": 0.19144551306962967,
      "num_tokens": 25132429.0,
      "step": 11270
    },
    {
      "entropy": 5.376098918914795,
      "epoch": 1.2662997697534677,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048504456318548285,
      "loss": 5.3517,
      "mean_token_accuracy": 0.1917685702443123,
      "num_tokens": 25144242.0,
      "step": 11275
    },
    {
      "entropy": 5.559010601043701,
      "epoch": 1.2668613466614254,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004850301666999722,
      "loss": 5.3544,
      "mean_token_accuracy": 0.1884535312652588,
      "num_tokens": 25155332.0,
      "step": 11280
    },
    {
      "entropy": 5.49990930557251,
      "epoch": 1.2674229235693828,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004850157635270196,
      "loss": 5.2626,
      "mean_token_accuracy": 0.19197598099708557,
      "num_tokens": 25166024.0,
      "step": 11285
    },
    {
      "entropy": 5.3888874530792235,
      "epoch": 1.2679845004773402,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004850013536670836,
      "loss": 5.2165,
      "mean_token_accuracy": 0.19812757670879363,
      "num_tokens": 25176697.0,
      "step": 11290
    },
    {
      "entropy": 5.489449119567871,
      "epoch": 1.268546077385298,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004849869371206231,
      "loss": 5.3293,
      "mean_token_accuracy": 0.192319555580616,
      "num_tokens": 25187421.0,
      "step": 11295
    },
    {
      "entropy": 5.465096426010132,
      "epoch": 1.2691076542932556,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048497251388809697,
      "loss": 5.321,
      "mean_token_accuracy": 0.19427925646305083,
      "num_tokens": 25198851.0,
      "step": 11300
    },
    {
      "entropy": 5.511981678009033,
      "epoch": 1.269669231201213,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004849580839699647,
      "loss": 5.3864,
      "mean_token_accuracy": 0.18640688955783843,
      "num_tokens": 25209984.0,
      "step": 11305
    },
    {
      "entropy": 5.46703724861145,
      "epoch": 1.2702308081091704,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048494364736668554,
      "loss": 5.264,
      "mean_token_accuracy": 0.1923437312245369,
      "num_tokens": 25220672.0,
      "step": 11310
    },
    {
      "entropy": 5.41923041343689,
      "epoch": 1.270792385017128,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004849292040787192,
      "loss": 5.2915,
      "mean_token_accuracy": 0.1876407966017723,
      "num_tokens": 25231906.0,
      "step": 11315
    },
    {
      "entropy": 5.443008232116699,
      "epoch": 1.2713539619250858,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004849147541065256,
      "loss": 5.3203,
      "mean_token_accuracy": 0.184516441822052,
      "num_tokens": 25242902.0,
      "step": 11320
    },
    {
      "entropy": 5.478665542602539,
      "epoch": 1.2719155388330432,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004849002974505649,
      "loss": 5.3616,
      "mean_token_accuracy": 0.18804416209459304,
      "num_tokens": 25254363.0,
      "step": 11325
    },
    {
      "entropy": 5.451558780670166,
      "epoch": 1.2724771157410006,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00048488583411129723,
      "loss": 5.3122,
      "mean_token_accuracy": 0.1893569067120552,
      "num_tokens": 25267108.0,
      "step": 11330
    },
    {
      "entropy": 5.389564561843872,
      "epoch": 1.2730386926489583,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004848713640891832,
      "loss": 5.2664,
      "mean_token_accuracy": 0.19266522377729417,
      "num_tokens": 25278583.0,
      "step": 11335
    },
    {
      "entropy": 5.549525308609009,
      "epoch": 1.2736002695569157,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004848568873846835,
      "loss": 5.3379,
      "mean_token_accuracy": 0.18999812602996827,
      "num_tokens": 25289813.0,
      "step": 11340
    },
    {
      "entropy": 5.468724489212036,
      "epoch": 1.2741618464648734,
      "grad_norm": 1.4140625,
      "learning_rate": 0.00048484240399825925,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19058207869529725,
      "num_tokens": 25301003.0,
      "step": 11345
    },
    {
      "entropy": 5.374887895584107,
      "epoch": 1.2747234233728308,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048482791393037134,
      "loss": 5.2332,
      "mean_token_accuracy": 0.19559966176748275,
      "num_tokens": 25312535.0,
      "step": 11350
    },
    {
      "entropy": 5.47691707611084,
      "epoch": 1.2752850002807885,
      "grad_norm": 1.0625,
      "learning_rate": 0.00048481341718148137,
      "loss": 5.2668,
      "mean_token_accuracy": 0.1896713063120842,
      "num_tokens": 25323333.0,
      "step": 11355
    },
    {
      "entropy": 5.533271026611328,
      "epoch": 1.275846577188746,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048479891375205074,
      "loss": 5.3775,
      "mean_token_accuracy": 0.18619518280029296,
      "num_tokens": 25334865.0,
      "step": 11360
    },
    {
      "entropy": 5.480346918106079,
      "epoch": 1.2764081540967036,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048478440364254136,
      "loss": 5.3528,
      "mean_token_accuracy": 0.19026066064834596,
      "num_tokens": 25346672.0,
      "step": 11365
    },
    {
      "entropy": 5.407110071182251,
      "epoch": 1.276969731004661,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004847698868534152,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1946812555193901,
      "num_tokens": 25358702.0,
      "step": 11370
    },
    {
      "entropy": 5.494549655914307,
      "epoch": 1.2775313079126187,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004847553633851345,
      "loss": 5.2705,
      "mean_token_accuracy": 0.190895876288414,
      "num_tokens": 25369813.0,
      "step": 11375
    },
    {
      "entropy": 5.510781240463257,
      "epoch": 1.2780928848205761,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004847408332381616,
      "loss": 5.3344,
      "mean_token_accuracy": 0.1878958985209465,
      "num_tokens": 25381042.0,
      "step": 11380
    },
    {
      "entropy": 5.42286901473999,
      "epoch": 1.2786544617285336,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048472629641295933,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18533823490142823,
      "num_tokens": 25391286.0,
      "step": 11385
    },
    {
      "entropy": 5.438542604446411,
      "epoch": 1.2792160386364912,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004847117529099904,
      "loss": 5.2648,
      "mean_token_accuracy": 0.19312452524900436,
      "num_tokens": 25402335.0,
      "step": 11390
    },
    {
      "entropy": 5.442463350296021,
      "epoch": 1.279777615544449,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048469720272971797,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19763229191303253,
      "num_tokens": 25413484.0,
      "step": 11395
    },
    {
      "entropy": 5.477785062789917,
      "epoch": 1.2803391924524063,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004846826458726052,
      "loss": 5.3238,
      "mean_token_accuracy": 0.19005614817142485,
      "num_tokens": 25424276.0,
      "step": 11400
    },
    {
      "entropy": 5.288488245010376,
      "epoch": 1.2809007693603638,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048466808233911567,
      "loss": 5.205,
      "mean_token_accuracy": 0.19397176951169967,
      "num_tokens": 25435178.0,
      "step": 11405
    },
    {
      "entropy": 5.336793279647827,
      "epoch": 1.2814623462683215,
      "grad_norm": 1.125,
      "learning_rate": 0.00048465351212971313,
      "loss": 5.1652,
      "mean_token_accuracy": 0.1950131192803383,
      "num_tokens": 25445792.0,
      "step": 11410
    },
    {
      "entropy": 5.438464117050171,
      "epoch": 1.2820239231762791,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004846389352448614,
      "loss": 5.2531,
      "mean_token_accuracy": 0.1899751305580139,
      "num_tokens": 25456598.0,
      "step": 11415
    },
    {
      "entropy": 5.452799940109253,
      "epoch": 1.2825855000842366,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004846243516850247,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19124604463577272,
      "num_tokens": 25468311.0,
      "step": 11420
    },
    {
      "entropy": 5.533212280273437,
      "epoch": 1.283147076992194,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004846097614506673,
      "loss": 5.3393,
      "mean_token_accuracy": 0.19007845371961593,
      "num_tokens": 25480680.0,
      "step": 11425
    },
    {
      "entropy": 5.530454778671265,
      "epoch": 1.2837086539001517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048459516454225384,
      "loss": 5.3193,
      "mean_token_accuracy": 0.19120112657546998,
      "num_tokens": 25492255.0,
      "step": 11430
    },
    {
      "entropy": 5.395530557632446,
      "epoch": 1.284270230808109,
      "grad_norm": 1.234375,
      "learning_rate": 0.000484580560960249,
      "loss": 5.3356,
      "mean_token_accuracy": 0.18894716203212739,
      "num_tokens": 25504530.0,
      "step": 11435
    },
    {
      "entropy": 5.369412565231324,
      "epoch": 1.2848318077160668,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004845659507051178,
      "loss": 5.2935,
      "mean_token_accuracy": 0.1962587207555771,
      "num_tokens": 25515273.0,
      "step": 11440
    },
    {
      "entropy": 5.51286473274231,
      "epoch": 1.2853933846240242,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004845513337773255,
      "loss": 5.4064,
      "mean_token_accuracy": 0.1862334504723549,
      "num_tokens": 25526050.0,
      "step": 11445
    },
    {
      "entropy": 5.539785432815552,
      "epoch": 1.2859549615319819,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048453671017733755,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1879033401608467,
      "num_tokens": 25538372.0,
      "step": 11450
    },
    {
      "entropy": 5.5917744636535645,
      "epoch": 1.2865165384399393,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004845220799056194,
      "loss": 5.4223,
      "mean_token_accuracy": 0.1843128889799118,
      "num_tokens": 25549636.0,
      "step": 11455
    },
    {
      "entropy": 5.459404993057251,
      "epoch": 1.287078115347897,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048450744296263685,
      "loss": 5.3434,
      "mean_token_accuracy": 0.18785598874092102,
      "num_tokens": 25560850.0,
      "step": 11460
    },
    {
      "entropy": 5.446890068054199,
      "epoch": 1.2876396922558544,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048449279934885625,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19257668554782867,
      "num_tokens": 25572433.0,
      "step": 11465
    },
    {
      "entropy": 5.53641529083252,
      "epoch": 1.288201269163812,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004844781490647435,
      "loss": 5.3339,
      "mean_token_accuracy": 0.1898763120174408,
      "num_tokens": 25582372.0,
      "step": 11470
    },
    {
      "entropy": 5.356485843658447,
      "epoch": 1.2887628460717695,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004844634921107654,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19469079077243806,
      "num_tokens": 25592570.0,
      "step": 11475
    },
    {
      "entropy": 5.4044684886932375,
      "epoch": 1.289324422979727,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048444882848738835,
      "loss": 5.3016,
      "mean_token_accuracy": 0.18627564907073973,
      "num_tokens": 25603573.0,
      "step": 11480
    },
    {
      "entropy": 5.486279726028442,
      "epoch": 1.2898859998876846,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048443415819507944,
      "loss": 5.3217,
      "mean_token_accuracy": 0.1926312506198883,
      "num_tokens": 25614233.0,
      "step": 11485
    },
    {
      "entropy": 5.454488325119018,
      "epoch": 1.2904475767956423,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004844194812343057,
      "loss": 5.2891,
      "mean_token_accuracy": 0.18609788566827773,
      "num_tokens": 25625077.0,
      "step": 11490
    },
    {
      "entropy": 5.39323673248291,
      "epoch": 1.2910091537035997,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004844047976055344,
      "loss": 5.3216,
      "mean_token_accuracy": 0.1889407902956009,
      "num_tokens": 25636693.0,
      "step": 11495
    },
    {
      "entropy": 5.352136850357056,
      "epoch": 1.2915707306115571,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048439010730923313,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19164122641086578,
      "num_tokens": 25647644.0,
      "step": 11500
    },
    {
      "entropy": 5.3950450897216795,
      "epoch": 1.2921323075195148,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048437541034586964,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19805077016353606,
      "num_tokens": 25659373.0,
      "step": 11505
    },
    {
      "entropy": 5.446753025054932,
      "epoch": 1.2926938844274725,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048436070671591195,
      "loss": 5.3383,
      "mean_token_accuracy": 0.19561638087034225,
      "num_tokens": 25670479.0,
      "step": 11510
    },
    {
      "entropy": 5.477274656295776,
      "epoch": 1.29325546133543,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048434599641982806,
      "loss": 5.3518,
      "mean_token_accuracy": 0.19136536568403245,
      "num_tokens": 25682748.0,
      "step": 11515
    },
    {
      "entropy": 5.426623821258545,
      "epoch": 1.2938170382433873,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048433127945808645,
      "loss": 5.1714,
      "mean_token_accuracy": 0.20042379051446915,
      "num_tokens": 25693170.0,
      "step": 11520
    },
    {
      "entropy": 5.3838776588439945,
      "epoch": 1.294378615151345,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004843165558311558,
      "loss": 5.3052,
      "mean_token_accuracy": 0.18993958532810212,
      "num_tokens": 25706009.0,
      "step": 11525
    },
    {
      "entropy": 5.488208913803101,
      "epoch": 1.2949401920593024,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004843018255395048,
      "loss": 5.3512,
      "mean_token_accuracy": 0.1883769527077675,
      "num_tokens": 25716036.0,
      "step": 11530
    },
    {
      "entropy": 5.4260131359100345,
      "epoch": 1.29550176896726,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048428708858360244,
      "loss": 5.1558,
      "mean_token_accuracy": 0.19588555097579957,
      "num_tokens": 25726516.0,
      "step": 11535
    },
    {
      "entropy": 5.419932651519775,
      "epoch": 1.2960633458752175,
      "grad_norm": 1.25,
      "learning_rate": 0.00048427234496391816,
      "loss": 5.3282,
      "mean_token_accuracy": 0.18386960476636888,
      "num_tokens": 25736340.0,
      "step": 11540
    },
    {
      "entropy": 5.337124300003052,
      "epoch": 1.2966249227831752,
      "grad_norm": 1.125,
      "learning_rate": 0.00048425759468092113,
      "loss": 5.278,
      "mean_token_accuracy": 0.18759150952100753,
      "num_tokens": 25746439.0,
      "step": 11545
    },
    {
      "entropy": 5.4072774887084964,
      "epoch": 1.2971864996911326,
      "grad_norm": 1.125,
      "learning_rate": 0.0004842428377350811,
      "loss": 5.192,
      "mean_token_accuracy": 0.19348155558109284,
      "num_tokens": 25757065.0,
      "step": 11550
    },
    {
      "entropy": 5.382005643844605,
      "epoch": 1.2977480765990903,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048422807412686796,
      "loss": 5.2586,
      "mean_token_accuracy": 0.1891964316368103,
      "num_tokens": 25768021.0,
      "step": 11555
    },
    {
      "entropy": 5.45659327507019,
      "epoch": 1.2983096535070477,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004842133038567519,
      "loss": 5.3606,
      "mean_token_accuracy": 0.1872963860630989,
      "num_tokens": 25780197.0,
      "step": 11560
    },
    {
      "entropy": 5.460580730438233,
      "epoch": 1.2988712304150054,
      "grad_norm": 1.171875,
      "learning_rate": 0.000484198526925203,
      "loss": 5.2651,
      "mean_token_accuracy": 0.1899239093065262,
      "num_tokens": 25790536.0,
      "step": 11565
    },
    {
      "entropy": 5.376824045181275,
      "epoch": 1.2994328073229628,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004841837433326919,
      "loss": 5.2366,
      "mean_token_accuracy": 0.1900440752506256,
      "num_tokens": 25802397.0,
      "step": 11570
    },
    {
      "entropy": 5.44475245475769,
      "epoch": 1.2999943842309205,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004841689530796893,
      "loss": 5.3695,
      "mean_token_accuracy": 0.18956518918275833,
      "num_tokens": 25813961.0,
      "step": 11575
    },
    {
      "entropy": 5.5441686630249025,
      "epoch": 1.300555961138878,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004841541561666661,
      "loss": 5.3781,
      "mean_token_accuracy": 0.18661829233169555,
      "num_tokens": 25825610.0,
      "step": 11580
    },
    {
      "entropy": 5.404580068588257,
      "epoch": 1.3011175380468356,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048413935259409343,
      "loss": 5.1963,
      "mean_token_accuracy": 0.19673963636159897,
      "num_tokens": 25836076.0,
      "step": 11585
    },
    {
      "entropy": 5.387002182006836,
      "epoch": 1.301679114954793,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048412454236244267,
      "loss": 5.3085,
      "mean_token_accuracy": 0.19685765653848647,
      "num_tokens": 25847341.0,
      "step": 11590
    },
    {
      "entropy": 5.374657535552979,
      "epoch": 1.3022406918627505,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048410972547218545,
      "loss": 5.2085,
      "mean_token_accuracy": 0.19608255475759506,
      "num_tokens": 25856912.0,
      "step": 11595
    },
    {
      "entropy": 5.408575534820557,
      "epoch": 1.3028022687707081,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048409490192379337,
      "loss": 5.2241,
      "mean_token_accuracy": 0.1889120191335678,
      "num_tokens": 25868370.0,
      "step": 11600
    },
    {
      "entropy": 5.37081298828125,
      "epoch": 1.3033638456786658,
      "grad_norm": 1.359375,
      "learning_rate": 0.0004840800717177386,
      "loss": 5.2623,
      "mean_token_accuracy": 0.18693817108869554,
      "num_tokens": 25879104.0,
      "step": 11605
    },
    {
      "entropy": 5.439136743545532,
      "epoch": 1.3039254225866233,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004840652348544931,
      "loss": 5.327,
      "mean_token_accuracy": 0.1828071340918541,
      "num_tokens": 25889744.0,
      "step": 11610
    },
    {
      "entropy": 5.55516505241394,
      "epoch": 1.3044869994945807,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004840503913345296,
      "loss": 5.4253,
      "mean_token_accuracy": 0.18115986436605452,
      "num_tokens": 25901615.0,
      "step": 11615
    },
    {
      "entropy": 5.477020883560181,
      "epoch": 1.3050485764025384,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004840355411583205,
      "loss": 5.3005,
      "mean_token_accuracy": 0.18793725967407227,
      "num_tokens": 25911418.0,
      "step": 11620
    },
    {
      "entropy": 5.484296083450317,
      "epoch": 1.3056101533104958,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048402068432633864,
      "loss": 5.3674,
      "mean_token_accuracy": 0.18996285647153854,
      "num_tokens": 25922141.0,
      "step": 11625
    },
    {
      "entropy": 5.511279201507568,
      "epoch": 1.3061717302184535,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004840058208390572,
      "loss": 5.3762,
      "mean_token_accuracy": 0.18528974801301956,
      "num_tokens": 25933897.0,
      "step": 11630
    },
    {
      "entropy": 5.4885965347290036,
      "epoch": 1.306733307126411,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004839909506969494,
      "loss": 5.3054,
      "mean_token_accuracy": 0.18785456866025924,
      "num_tokens": 25943960.0,
      "step": 11635
    },
    {
      "entropy": 5.530137395858764,
      "epoch": 1.3072948840343686,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004839760739004886,
      "loss": 5.3499,
      "mean_token_accuracy": 0.19060384184122087,
      "num_tokens": 25955104.0,
      "step": 11640
    },
    {
      "entropy": 5.4485923767089846,
      "epoch": 1.307856460942326,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048396119045014853,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1946641430258751,
      "num_tokens": 25965180.0,
      "step": 11645
    },
    {
      "entropy": 5.4380322933197025,
      "epoch": 1.3084180378502837,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004839463003464032,
      "loss": 5.2965,
      "mean_token_accuracy": 0.19169021844863893,
      "num_tokens": 25976600.0,
      "step": 11650
    },
    {
      "entropy": 5.390657043457031,
      "epoch": 1.308979614758241,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004839314035897265,
      "loss": 5.2748,
      "mean_token_accuracy": 0.18478263765573502,
      "num_tokens": 25987509.0,
      "step": 11655
    },
    {
      "entropy": 5.439964342117309,
      "epoch": 1.3095411916661988,
      "grad_norm": 1.2421875,
      "learning_rate": 0.000483916500180593,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19598240107297898,
      "num_tokens": 25997962.0,
      "step": 11660
    },
    {
      "entropy": 5.436206483840943,
      "epoch": 1.3101027685741562,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048390159011947706,
      "loss": 5.2645,
      "mean_token_accuracy": 0.19282422661781312,
      "num_tokens": 26008932.0,
      "step": 11665
    },
    {
      "entropy": 5.437794780731201,
      "epoch": 1.3106643454821139,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004838866734068535,
      "loss": 5.2903,
      "mean_token_accuracy": 0.18584103584289552,
      "num_tokens": 26020513.0,
      "step": 11670
    },
    {
      "entropy": 5.455857372283935,
      "epoch": 1.3112259223900713,
      "grad_norm": 1.125,
      "learning_rate": 0.00048387175004319723,
      "loss": 5.3335,
      "mean_token_accuracy": 0.18863563537597655,
      "num_tokens": 26032526.0,
      "step": 11675
    },
    {
      "entropy": 5.444508790969849,
      "epoch": 1.311787499298029,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004838568200289834,
      "loss": 5.2729,
      "mean_token_accuracy": 0.19432038217782974,
      "num_tokens": 26044216.0,
      "step": 11680
    },
    {
      "entropy": 5.393547534942627,
      "epoch": 1.3123490762059864,
      "grad_norm": 1.296875,
      "learning_rate": 0.0004838418833646875,
      "loss": 5.1935,
      "mean_token_accuracy": 0.1901652529835701,
      "num_tokens": 26054012.0,
      "step": 11685
    },
    {
      "entropy": 5.461218786239624,
      "epoch": 1.3129106531139438,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048382694005078485,
      "loss": 5.3303,
      "mean_token_accuracy": 0.1871440127491951,
      "num_tokens": 26065365.0,
      "step": 11690
    },
    {
      "entropy": 5.4768517971038815,
      "epoch": 1.3134722300219015,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048381199008775165,
      "loss": 5.3339,
      "mean_token_accuracy": 0.18522805273532866,
      "num_tokens": 26075964.0,
      "step": 11695
    },
    {
      "entropy": 5.412776088714599,
      "epoch": 1.3140338069298592,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004837970334760635,
      "loss": 5.2779,
      "mean_token_accuracy": 0.18798635601997377,
      "num_tokens": 26087572.0,
      "step": 11700
    },
    {
      "entropy": 5.421622371673584,
      "epoch": 1.3145953838378166,
      "grad_norm": 1.140625,
      "learning_rate": 0.000483782070216197,
      "loss": 5.1936,
      "mean_token_accuracy": 0.19675613939762115,
      "num_tokens": 26098612.0,
      "step": 11705
    },
    {
      "entropy": 5.479706764221191,
      "epoch": 1.315156960745774,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004837671003086283,
      "loss": 5.3656,
      "mean_token_accuracy": 0.1914112836122513,
      "num_tokens": 26109491.0,
      "step": 11710
    },
    {
      "entropy": 5.455551099777222,
      "epoch": 1.3157185376537317,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004837521237538342,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19193665087223052,
      "num_tokens": 26120544.0,
      "step": 11715
    },
    {
      "entropy": 5.508788585662842,
      "epoch": 1.3162801145616891,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004837371405522915,
      "loss": 5.3515,
      "mean_token_accuracy": 0.18545960187911986,
      "num_tokens": 26131549.0,
      "step": 11720
    },
    {
      "entropy": 5.428353118896484,
      "epoch": 1.3168416914696468,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048372215070447734,
      "loss": 5.336,
      "mean_token_accuracy": 0.1950080707669258,
      "num_tokens": 26142871.0,
      "step": 11725
    },
    {
      "entropy": 5.519115543365478,
      "epoch": 1.3174032683776042,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048370715421086884,
      "loss": 5.42,
      "mean_token_accuracy": 0.19015678614377976,
      "num_tokens": 26155293.0,
      "step": 11730
    },
    {
      "entropy": 5.54082498550415,
      "epoch": 1.317964845285562,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004836921510719438,
      "loss": 5.3436,
      "mean_token_accuracy": 0.19598201662302017,
      "num_tokens": 26167069.0,
      "step": 11735
    },
    {
      "entropy": 5.464824247360229,
      "epoch": 1.3185264221935193,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004836771412881795,
      "loss": 5.3459,
      "mean_token_accuracy": 0.18280351012945176,
      "num_tokens": 26179142.0,
      "step": 11740
    },
    {
      "entropy": 5.50031681060791,
      "epoch": 1.319087999101477,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004836621248600542,
      "loss": 5.2944,
      "mean_token_accuracy": 0.1861179530620575,
      "num_tokens": 26190059.0,
      "step": 11745
    },
    {
      "entropy": 5.504112815856933,
      "epoch": 1.3196495760094344,
      "grad_norm": 1.125,
      "learning_rate": 0.0004836471017880459,
      "loss": 5.3078,
      "mean_token_accuracy": 0.1901737257838249,
      "num_tokens": 26199859.0,
      "step": 11750
    },
    {
      "entropy": 5.452195405960083,
      "epoch": 1.320211152917392,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004836320720726329,
      "loss": 5.3366,
      "mean_token_accuracy": 0.18487878739833832,
      "num_tokens": 26211920.0,
      "step": 11755
    },
    {
      "entropy": 5.434342622756958,
      "epoch": 1.3207727298253495,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004836170357142939,
      "loss": 5.3025,
      "mean_token_accuracy": 0.18566835373640062,
      "num_tokens": 26222875.0,
      "step": 11760
    },
    {
      "entropy": 5.478261089324951,
      "epoch": 1.3213343067333072,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004836019927135076,
      "loss": 5.3302,
      "mean_token_accuracy": 0.18550052046775817,
      "num_tokens": 26233549.0,
      "step": 11765
    },
    {
      "entropy": 5.391707468032837,
      "epoch": 1.3218958836412646,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004835869430707529,
      "loss": 5.2128,
      "mean_token_accuracy": 0.19971633851528167,
      "num_tokens": 26243886.0,
      "step": 11770
    },
    {
      "entropy": 5.357774019241333,
      "epoch": 1.3224574605492223,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048357188678650905,
      "loss": 5.2424,
      "mean_token_accuracy": 0.20098111033439636,
      "num_tokens": 26254475.0,
      "step": 11775
    },
    {
      "entropy": 5.553868150711059,
      "epoch": 1.3230190374571797,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004835568238612553,
      "loss": 5.4169,
      "mean_token_accuracy": 0.18773356974124908,
      "num_tokens": 26265412.0,
      "step": 11780
    },
    {
      "entropy": 5.527062511444091,
      "epoch": 1.3235806143651372,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048354175429547155,
      "loss": 5.3614,
      "mean_token_accuracy": 0.18862174302339554,
      "num_tokens": 26276942.0,
      "step": 11785
    },
    {
      "entropy": 5.462158823013306,
      "epoch": 1.3241421912730948,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004835266780896374,
      "loss": 5.3094,
      "mean_token_accuracy": 0.19409841001033784,
      "num_tokens": 26288009.0,
      "step": 11790
    },
    {
      "entropy": 5.5329564094543455,
      "epoch": 1.3247037681810525,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004835115952442329,
      "loss": 5.442,
      "mean_token_accuracy": 0.18399617075920105,
      "num_tokens": 26301583.0,
      "step": 11795
    },
    {
      "entropy": 5.536028575897217,
      "epoch": 1.32526534508901,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048349650575973843,
      "loss": 5.3623,
      "mean_token_accuracy": 0.19186441153287886,
      "num_tokens": 26312922.0,
      "step": 11800
    },
    {
      "entropy": 5.480445623397827,
      "epoch": 1.3258269219969674,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048348140963663436,
      "loss": 5.3167,
      "mean_token_accuracy": 0.1927507370710373,
      "num_tokens": 26324108.0,
      "step": 11805
    },
    {
      "entropy": 5.4644311428070065,
      "epoch": 1.326388498904925,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004834663068754012,
      "loss": 5.3759,
      "mean_token_accuracy": 0.18691876381635666,
      "num_tokens": 26335669.0,
      "step": 11810
    },
    {
      "entropy": 5.3608105182647705,
      "epoch": 1.3269500758128827,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004834511974765202,
      "loss": 5.2411,
      "mean_token_accuracy": 0.1950603812932968,
      "num_tokens": 26346065.0,
      "step": 11815
    },
    {
      "entropy": 5.490773105621338,
      "epoch": 1.3275116527208402,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048343608144047194,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19345590472221375,
      "num_tokens": 26357567.0,
      "step": 11820
    },
    {
      "entropy": 5.467666339874268,
      "epoch": 1.3280732296287976,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048342095876773825,
      "loss": 5.2732,
      "mean_token_accuracy": 0.1917034789919853,
      "num_tokens": 26367805.0,
      "step": 11825
    },
    {
      "entropy": 5.358530426025391,
      "epoch": 1.3286348065367553,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004834058294588003,
      "loss": 5.1861,
      "mean_token_accuracy": 0.1910465106368065,
      "num_tokens": 26379343.0,
      "step": 11830
    },
    {
      "entropy": 5.5350868701934814,
      "epoch": 1.3291963834447127,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00048339069351413993,
      "loss": 5.4139,
      "mean_token_accuracy": 0.1839591458439827,
      "num_tokens": 26392039.0,
      "step": 11835
    },
    {
      "entropy": 5.385570287704468,
      "epoch": 1.3297579603526704,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483375550934239,
      "loss": 5.1721,
      "mean_token_accuracy": 0.2043902635574341,
      "num_tokens": 26403324.0,
      "step": 11840
    },
    {
      "entropy": 5.402548217773438,
      "epoch": 1.3303195372606278,
      "grad_norm": 1.4765625,
      "learning_rate": 0.0004833604017195798,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19430694878101348,
      "num_tokens": 26413435.0,
      "step": 11845
    },
    {
      "entropy": 5.44825291633606,
      "epoch": 1.3308811141685855,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004833452458706445,
      "loss": 5.3717,
      "mean_token_accuracy": 0.1835521325469017,
      "num_tokens": 26424254.0,
      "step": 11850
    },
    {
      "entropy": 5.527906084060669,
      "epoch": 1.331442691076543,
      "grad_norm": 1.125,
      "learning_rate": 0.0004833300833879159,
      "loss": 5.2986,
      "mean_token_accuracy": 0.1901722103357315,
      "num_tokens": 26435658.0,
      "step": 11855
    },
    {
      "entropy": 5.486177921295166,
      "epoch": 1.3320042679845006,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004833149142718765,
      "loss": 5.3319,
      "mean_token_accuracy": 0.1876470848917961,
      "num_tokens": 26447174.0,
      "step": 11860
    },
    {
      "entropy": 5.514771032333374,
      "epoch": 1.332565844892458,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004832997385230096,
      "loss": 5.4246,
      "mean_token_accuracy": 0.18689661622047424,
      "num_tokens": 26459303.0,
      "step": 11865
    },
    {
      "entropy": 5.484613656997681,
      "epoch": 1.3331274218004157,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004832845561417981,
      "loss": 5.2938,
      "mean_token_accuracy": 0.1870502322912216,
      "num_tokens": 26469981.0,
      "step": 11870
    },
    {
      "entropy": 5.414060592651367,
      "epoch": 1.333688998708373,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004832693671287256,
      "loss": 5.2973,
      "mean_token_accuracy": 0.19529629945755006,
      "num_tokens": 26481176.0,
      "step": 11875
    },
    {
      "entropy": 5.385348558425903,
      "epoch": 1.3342505756163305,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004832541714842757,
      "loss": 5.2466,
      "mean_token_accuracy": 0.19705854952335358,
      "num_tokens": 26492795.0,
      "step": 11880
    },
    {
      "entropy": 5.513296794891358,
      "epoch": 1.3348121525242882,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004832389692089321,
      "loss": 5.4,
      "mean_token_accuracy": 0.18487481623888016,
      "num_tokens": 26504731.0,
      "step": 11885
    },
    {
      "entropy": 5.530076169967652,
      "epoch": 1.3353737294322459,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048322376030317905,
      "loss": 5.2896,
      "mean_token_accuracy": 0.19095005542039872,
      "num_tokens": 26516816.0,
      "step": 11890
    },
    {
      "entropy": 5.463229370117188,
      "epoch": 1.3359353063402033,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004832085447675007,
      "loss": 5.3627,
      "mean_token_accuracy": 0.19170191884040833,
      "num_tokens": 26528861.0,
      "step": 11895
    },
    {
      "entropy": 5.519645738601684,
      "epoch": 1.3364968832481607,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048319332260238153,
      "loss": 5.3508,
      "mean_token_accuracy": 0.18895784467458726,
      "num_tokens": 26538816.0,
      "step": 11900
    },
    {
      "entropy": 5.5793710231781,
      "epoch": 1.3370584601561184,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00048317809380830624,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18716011494398116,
      "num_tokens": 26548959.0,
      "step": 11905
    },
    {
      "entropy": 5.377211284637451,
      "epoch": 1.337620037064076,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004831628583857597,
      "loss": 5.25,
      "mean_token_accuracy": 0.1925012707710266,
      "num_tokens": 26559189.0,
      "step": 11910
    },
    {
      "entropy": 5.327203035354614,
      "epoch": 1.3381816139720335,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048314761633522696,
      "loss": 5.2186,
      "mean_token_accuracy": 0.1972116231918335,
      "num_tokens": 26570066.0,
      "step": 11915
    },
    {
      "entropy": 5.472088718414307,
      "epoch": 1.338743190879991,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004831323676571934,
      "loss": 5.3735,
      "mean_token_accuracy": 0.1926550641655922,
      "num_tokens": 26581489.0,
      "step": 11920
    },
    {
      "entropy": 5.527829360961914,
      "epoch": 1.3393047677879486,
      "grad_norm": 1.25,
      "learning_rate": 0.0004831171123521446,
      "loss": 5.3055,
      "mean_token_accuracy": 0.1859520986676216,
      "num_tokens": 26592051.0,
      "step": 11925
    },
    {
      "entropy": 5.481430625915527,
      "epoch": 1.339866344695906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004831018504205662,
      "loss": 5.3014,
      "mean_token_accuracy": 0.19255633354187013,
      "num_tokens": 26602296.0,
      "step": 11930
    },
    {
      "entropy": 5.414361953735352,
      "epoch": 1.3404279216038637,
      "grad_norm": 1.125,
      "learning_rate": 0.00048308658186294414,
      "loss": 5.2833,
      "mean_token_accuracy": 0.19021888822317123,
      "num_tokens": 26613837.0,
      "step": 11935
    },
    {
      "entropy": 5.444925212860108,
      "epoch": 1.3409894985118211,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004830713066797646,
      "loss": 5.3497,
      "mean_token_accuracy": 0.18984478563070298,
      "num_tokens": 26624898.0,
      "step": 11940
    },
    {
      "entropy": 5.522456264495849,
      "epoch": 1.3415510754197788,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048305602487151396,
      "loss": 5.3278,
      "mean_token_accuracy": 0.19578097611665726,
      "num_tokens": 26636449.0,
      "step": 11945
    },
    {
      "entropy": 5.593957424163818,
      "epoch": 1.3421126523277362,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004830407364386789,
      "loss": 5.4146,
      "mean_token_accuracy": 0.1892138123512268,
      "num_tokens": 26649192.0,
      "step": 11950
    },
    {
      "entropy": 5.417656087875367,
      "epoch": 1.342674229235694,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000483025441381746,
      "loss": 5.2276,
      "mean_token_accuracy": 0.19503106623888017,
      "num_tokens": 26660005.0,
      "step": 11955
    },
    {
      "entropy": 5.545748949050903,
      "epoch": 1.3432358061436513,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048301013970120235,
      "loss": 5.3833,
      "mean_token_accuracy": 0.18726867586374282,
      "num_tokens": 26672165.0,
      "step": 11960
    },
    {
      "entropy": 5.432900714874267,
      "epoch": 1.343797383051609,
      "grad_norm": 1.34375,
      "learning_rate": 0.00048299483139753517,
      "loss": 5.3151,
      "mean_token_accuracy": 0.1915122762322426,
      "num_tokens": 26683135.0,
      "step": 11965
    },
    {
      "entropy": 5.404838228225708,
      "epoch": 1.3443589599595664,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004829795164712319,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19967983514070511,
      "num_tokens": 26693724.0,
      "step": 11970
    },
    {
      "entropy": 5.500458383560181,
      "epoch": 1.3449205368675239,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048296419492278023,
      "loss": 5.4076,
      "mean_token_accuracy": 0.18995761424303054,
      "num_tokens": 26705880.0,
      "step": 11975
    },
    {
      "entropy": 5.514515018463134,
      "epoch": 1.3454821137754815,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004829488667526679,
      "loss": 5.3753,
      "mean_token_accuracy": 0.18945090770721434,
      "num_tokens": 26716978.0,
      "step": 11980
    },
    {
      "entropy": 5.57459192276001,
      "epoch": 1.3460436906834392,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048293353196138294,
      "loss": 5.3143,
      "mean_token_accuracy": 0.20220770239830016,
      "num_tokens": 26728380.0,
      "step": 11985
    },
    {
      "entropy": 5.406123495101928,
      "epoch": 1.3466052675913966,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048291819054941375,
      "loss": 5.2769,
      "mean_token_accuracy": 0.19418565183877945,
      "num_tokens": 26739034.0,
      "step": 11990
    },
    {
      "entropy": 5.258612871170044,
      "epoch": 1.347166844499354,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004829028425172487,
      "loss": 5.0791,
      "mean_token_accuracy": 0.20205246955156325,
      "num_tokens": 26750123.0,
      "step": 11995
    },
    {
      "entropy": 5.372890615463257,
      "epoch": 1.3477284214073117,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048288748786537656,
      "loss": 5.2683,
      "mean_token_accuracy": 0.20076695084571838,
      "num_tokens": 26760411.0,
      "step": 12000
    },
    {
      "epoch": 1.3477284214073117,
      "eval_entropy": 5.272682303708332,
      "eval_loss": 5.398852825164795,
      "eval_mean_token_accuracy": 0.19458526289313915,
      "eval_num_tokens": 26760411.0,
      "eval_runtime": 24.0649,
      "eval_samples_per_second": 1288.724,
      "eval_steps_per_second": 161.106,
      "step": 12000
    },
    {
      "entropy": 5.460191154479981,
      "epoch": 1.3482899983152694,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004828721265942861,
      "loss": 5.3236,
      "mean_token_accuracy": 0.19536752700805665,
      "num_tokens": 26770821.0,
      "step": 12005
    },
    {
      "entropy": 5.488104820251465,
      "epoch": 1.3488515752232269,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048285675870446653,
      "loss": 5.3108,
      "mean_token_accuracy": 0.19492012560367583,
      "num_tokens": 26782099.0,
      "step": 12010
    },
    {
      "entropy": 5.52950348854065,
      "epoch": 1.3494131521311843,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048284138419640724,
      "loss": 5.4268,
      "mean_token_accuracy": 0.1866847515106201,
      "num_tokens": 26794056.0,
      "step": 12015
    },
    {
      "entropy": 5.445260286331177,
      "epoch": 1.349974729039142,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004828260030705976,
      "loss": 5.2643,
      "mean_token_accuracy": 0.1904071569442749,
      "num_tokens": 26804501.0,
      "step": 12020
    },
    {
      "entropy": 5.52919111251831,
      "epoch": 1.3505363059470994,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048281061532752734,
      "loss": 5.3331,
      "mean_token_accuracy": 0.1926148861646652,
      "num_tokens": 26814996.0,
      "step": 12025
    },
    {
      "entropy": 5.42205777168274,
      "epoch": 1.351097882855057,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004827952209676866,
      "loss": 5.264,
      "mean_token_accuracy": 0.19196725338697435,
      "num_tokens": 26825838.0,
      "step": 12030
    },
    {
      "entropy": 5.4905433177948,
      "epoch": 1.3516594597630145,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048277981999156534,
      "loss": 5.2835,
      "mean_token_accuracy": 0.19127179831266403,
      "num_tokens": 26836320.0,
      "step": 12035
    },
    {
      "entropy": 5.467498636245727,
      "epoch": 1.3522210366709722,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00048276441239965414,
      "loss": 5.2667,
      "mean_token_accuracy": 0.19495543092489243,
      "num_tokens": 26846485.0,
      "step": 12040
    },
    {
      "entropy": 5.533906412124634,
      "epoch": 1.3527826135789296,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004827489981924434,
      "loss": 5.39,
      "mean_token_accuracy": 0.1856178715825081,
      "num_tokens": 26858554.0,
      "step": 12045
    },
    {
      "entropy": 5.4507801055908205,
      "epoch": 1.3533441904868873,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048273357737042393,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19157515317201615,
      "num_tokens": 26869538.0,
      "step": 12050
    },
    {
      "entropy": 5.546870136260987,
      "epoch": 1.3539057673948447,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048271814993408694,
      "loss": 5.3835,
      "mean_token_accuracy": 0.18661566078662872,
      "num_tokens": 26880284.0,
      "step": 12055
    },
    {
      "entropy": 5.46618504524231,
      "epoch": 1.3544673443028024,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004827027158839233,
      "loss": 5.3688,
      "mean_token_accuracy": 0.17917487770318985,
      "num_tokens": 26893180.0,
      "step": 12060
    },
    {
      "entropy": 5.427161645889282,
      "epoch": 1.3550289212107598,
      "grad_norm": 1.125,
      "learning_rate": 0.0004826872752204248,
      "loss": 5.2195,
      "mean_token_accuracy": 0.19233309775590895,
      "num_tokens": 26904213.0,
      "step": 12065
    },
    {
      "entropy": 5.47069902420044,
      "epoch": 1.3555904981187172,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048267182794408276,
      "loss": 5.3129,
      "mean_token_accuracy": 0.19588389992713928,
      "num_tokens": 26915566.0,
      "step": 12070
    },
    {
      "entropy": 5.362538528442383,
      "epoch": 1.356152075026675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004826563740553893,
      "loss": 5.2713,
      "mean_token_accuracy": 0.1930746078491211,
      "num_tokens": 26926253.0,
      "step": 12075
    },
    {
      "entropy": 5.418523454666138,
      "epoch": 1.3567136519346326,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004826409135548363,
      "loss": 5.2523,
      "mean_token_accuracy": 0.20055679082870484,
      "num_tokens": 26936723.0,
      "step": 12080
    },
    {
      "entropy": 5.404415273666382,
      "epoch": 1.35727522884259,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004826254464429161,
      "loss": 5.2517,
      "mean_token_accuracy": 0.191728575527668,
      "num_tokens": 26947551.0,
      "step": 12085
    },
    {
      "entropy": 5.466691732406616,
      "epoch": 1.3578368057505474,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00048260997272012114,
      "loss": 5.3716,
      "mean_token_accuracy": 0.19475919157266616,
      "num_tokens": 26959969.0,
      "step": 12090
    },
    {
      "entropy": 5.554465961456299,
      "epoch": 1.358398382658505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004825944923869441,
      "loss": 5.4293,
      "mean_token_accuracy": 0.1870420902967453,
      "num_tokens": 26972511.0,
      "step": 12095
    },
    {
      "entropy": 5.548736619949341,
      "epoch": 1.3589599595664628,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048257900544387795,
      "loss": 5.4203,
      "mean_token_accuracy": 0.1859400674700737,
      "num_tokens": 26983827.0,
      "step": 12100
    },
    {
      "entropy": 5.601272964477539,
      "epoch": 1.3595215364744202,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048256351189141574,
      "loss": 5.3745,
      "mean_token_accuracy": 0.18305257111787795,
      "num_tokens": 26994668.0,
      "step": 12105
    },
    {
      "entropy": 5.433708953857422,
      "epoch": 1.3600831133823776,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004825480117300508,
      "loss": 5.2366,
      "mean_token_accuracy": 0.20103898048400878,
      "num_tokens": 27006313.0,
      "step": 12110
    },
    {
      "entropy": 5.4824872493743895,
      "epoch": 1.3606446902903353,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048253250496027675,
      "loss": 5.2942,
      "mean_token_accuracy": 0.18631143718957902,
      "num_tokens": 27017628.0,
      "step": 12115
    },
    {
      "entropy": 5.372732543945313,
      "epoch": 1.3612062671982927,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00048251699158258714,
      "loss": 5.1994,
      "mean_token_accuracy": 0.19352100193500518,
      "num_tokens": 27028303.0,
      "step": 12120
    },
    {
      "entropy": 5.510053825378418,
      "epoch": 1.3617678441062504,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004825014715974761,
      "loss": 5.3425,
      "mean_token_accuracy": 0.1854233518242836,
      "num_tokens": 27039307.0,
      "step": 12125
    },
    {
      "entropy": 5.461513376235962,
      "epoch": 1.3623294210142078,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048248594500543767,
      "loss": 5.2457,
      "mean_token_accuracy": 0.18978635519742965,
      "num_tokens": 27049271.0,
      "step": 12130
    },
    {
      "entropy": 5.338725471496582,
      "epoch": 1.3628909979221655,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048247041180696633,
      "loss": 5.1516,
      "mean_token_accuracy": 0.19708142578601837,
      "num_tokens": 27059510.0,
      "step": 12135
    },
    {
      "entropy": 5.561393070220947,
      "epoch": 1.363452574830123,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004824548720025566,
      "loss": 5.4858,
      "mean_token_accuracy": 0.1819046527147293,
      "num_tokens": 27071685.0,
      "step": 12140
    },
    {
      "entropy": 5.436061429977417,
      "epoch": 1.3640141517380806,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004824393255927033,
      "loss": 5.2177,
      "mean_token_accuracy": 0.18922069072723388,
      "num_tokens": 27081957.0,
      "step": 12145
    },
    {
      "entropy": 5.364203596115113,
      "epoch": 1.364575728646038,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048242377257790145,
      "loss": 5.2283,
      "mean_token_accuracy": 0.1956916019320488,
      "num_tokens": 27091609.0,
      "step": 12150
    },
    {
      "entropy": 5.478212928771972,
      "epoch": 1.3651373055539957,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048240821295864614,
      "loss": 5.3984,
      "mean_token_accuracy": 0.1872815817594528,
      "num_tokens": 27103889.0,
      "step": 12155
    },
    {
      "entropy": 5.576805353164673,
      "epoch": 1.3656988824619531,
      "grad_norm": 1.265625,
      "learning_rate": 0.000482392646735433,
      "loss": 5.3325,
      "mean_token_accuracy": 0.1915147140622139,
      "num_tokens": 27114186.0,
      "step": 12160
    },
    {
      "entropy": 5.393768548965454,
      "epoch": 1.3662604593699106,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004823770739087574,
      "loss": 5.1987,
      "mean_token_accuracy": 0.19757619351148606,
      "num_tokens": 27124677.0,
      "step": 12165
    },
    {
      "entropy": 5.54158878326416,
      "epoch": 1.3668220362778682,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004823614944791154,
      "loss": 5.381,
      "mean_token_accuracy": 0.1940017744898796,
      "num_tokens": 27134873.0,
      "step": 12170
    },
    {
      "entropy": 5.44172854423523,
      "epoch": 1.367383613185826,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004823459084470031,
      "loss": 5.3838,
      "mean_token_accuracy": 0.19058656692504883,
      "num_tokens": 27146107.0,
      "step": 12175
    },
    {
      "entropy": 5.425854063034057,
      "epoch": 1.3679451900937833,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004823303158129165,
      "loss": 5.2325,
      "mean_token_accuracy": 0.2001013457775116,
      "num_tokens": 27158821.0,
      "step": 12180
    },
    {
      "entropy": 5.47525315284729,
      "epoch": 1.3685067670017408,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004823147165773523,
      "loss": 5.2474,
      "mean_token_accuracy": 0.19808011054992675,
      "num_tokens": 27169513.0,
      "step": 12185
    },
    {
      "entropy": 5.359856700897216,
      "epoch": 1.3690683439096984,
      "grad_norm": 1.171875,
      "learning_rate": 0.00048229911074080715,
      "loss": 5.2599,
      "mean_token_accuracy": 0.1965148538351059,
      "num_tokens": 27180059.0,
      "step": 12190
    },
    {
      "entropy": 5.41556134223938,
      "epoch": 1.369629920817656,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00048228349830377787,
      "loss": 5.3241,
      "mean_token_accuracy": 0.19290325343608855,
      "num_tokens": 27190738.0,
      "step": 12195
    },
    {
      "entropy": 5.410951280593872,
      "epoch": 1.3701914977256135,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048226787926676165,
      "loss": 5.21,
      "mean_token_accuracy": 0.19297966957092286,
      "num_tokens": 27200664.0,
      "step": 12200
    },
    {
      "entropy": 5.411569213867187,
      "epoch": 1.370753074633571,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00048225225363025576,
      "loss": 5.2791,
      "mean_token_accuracy": 0.19224051982164383,
      "num_tokens": 27211011.0,
      "step": 12205
    },
    {
      "entropy": 5.511159753799438,
      "epoch": 1.3713146515415287,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048223662139475773,
      "loss": 5.328,
      "mean_token_accuracy": 0.19868140518665314,
      "num_tokens": 27221888.0,
      "step": 12210
    },
    {
      "entropy": 5.530559873580932,
      "epoch": 1.371876228449486,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048222098256076523,
      "loss": 5.3189,
      "mean_token_accuracy": 0.18549761325120925,
      "num_tokens": 27234671.0,
      "step": 12215
    },
    {
      "entropy": 5.337631416320801,
      "epoch": 1.3724378053574438,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004822053371287763,
      "loss": 5.1291,
      "mean_token_accuracy": 0.19670447409152986,
      "num_tokens": 27245283.0,
      "step": 12220
    },
    {
      "entropy": 5.431076765060425,
      "epoch": 1.3729993822654012,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004821896850992891,
      "loss": 5.2849,
      "mean_token_accuracy": 0.1954860806465149,
      "num_tokens": 27255056.0,
      "step": 12225
    },
    {
      "entropy": 5.537234783172607,
      "epoch": 1.3735609591733589,
      "grad_norm": 1.3125,
      "learning_rate": 0.000482174026472802,
      "loss": 5.3464,
      "mean_token_accuracy": 0.1846200332045555,
      "num_tokens": 27266893.0,
      "step": 12230
    },
    {
      "entropy": 5.425251054763794,
      "epoch": 1.3741225360813163,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004821583612498135,
      "loss": 5.37,
      "mean_token_accuracy": 0.1895558714866638,
      "num_tokens": 27279504.0,
      "step": 12235
    },
    {
      "entropy": 5.557827091217041,
      "epoch": 1.374684112989274,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048214268943082246,
      "loss": 5.4115,
      "mean_token_accuracy": 0.18655518293380738,
      "num_tokens": 27290525.0,
      "step": 12240
    },
    {
      "entropy": 5.536990976333618,
      "epoch": 1.3752456898972314,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004821270110163278,
      "loss": 5.2658,
      "mean_token_accuracy": 0.1923648402094841,
      "num_tokens": 27301404.0,
      "step": 12245
    },
    {
      "entropy": 5.405010080337524,
      "epoch": 1.375807266805189,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004821113260068288,
      "loss": 5.2202,
      "mean_token_accuracy": 0.19178383201360702,
      "num_tokens": 27312635.0,
      "step": 12250
    },
    {
      "entropy": 5.390475606918335,
      "epoch": 1.3763688437131465,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004820956344028249,
      "loss": 5.2653,
      "mean_token_accuracy": 0.1974032089114189,
      "num_tokens": 27322485.0,
      "step": 12255
    },
    {
      "entropy": 5.439991807937622,
      "epoch": 1.376930420621104,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048207993620481566,
      "loss": 5.311,
      "mean_token_accuracy": 0.1886539027094841,
      "num_tokens": 27332568.0,
      "step": 12260
    },
    {
      "entropy": 5.554746627807617,
      "epoch": 1.3774919975290616,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004820642314133009,
      "loss": 5.3523,
      "mean_token_accuracy": 0.18777467012405397,
      "num_tokens": 27343058.0,
      "step": 12265
    },
    {
      "entropy": 5.442495822906494,
      "epoch": 1.3780535744370193,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048204852002878073,
      "loss": 5.2438,
      "mean_token_accuracy": 0.19678584784269332,
      "num_tokens": 27355316.0,
      "step": 12270
    },
    {
      "entropy": 5.371298599243164,
      "epoch": 1.3786151513449767,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048203280205175544,
      "loss": 5.1865,
      "mean_token_accuracy": 0.19887161254882812,
      "num_tokens": 27366098.0,
      "step": 12275
    },
    {
      "entropy": 5.399563932418824,
      "epoch": 1.3791767282529341,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004820170774827253,
      "loss": 5.2109,
      "mean_token_accuracy": 0.2009422779083252,
      "num_tokens": 27376923.0,
      "step": 12280
    },
    {
      "entropy": 5.504350757598877,
      "epoch": 1.3797383051608918,
      "grad_norm": 1.6015625,
      "learning_rate": 0.0004820013463221912,
      "loss": 5.3272,
      "mean_token_accuracy": 0.18912267982959746,
      "num_tokens": 27386789.0,
      "step": 12285
    },
    {
      "entropy": 5.415982341766357,
      "epoch": 1.3802998820688495,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000481985608570654,
      "loss": 5.2565,
      "mean_token_accuracy": 0.1912071317434311,
      "num_tokens": 27399241.0,
      "step": 12290
    },
    {
      "entropy": 5.35224404335022,
      "epoch": 1.380861458976807,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004819698642286147,
      "loss": 5.1633,
      "mean_token_accuracy": 0.20056203454732896,
      "num_tokens": 27409123.0,
      "step": 12295
    },
    {
      "entropy": 5.385395812988281,
      "epoch": 1.3814230358847643,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048195411329657464,
      "loss": 5.2321,
      "mean_token_accuracy": 0.19377838224172592,
      "num_tokens": 27419920.0,
      "step": 12300
    },
    {
      "entropy": 5.328425931930542,
      "epoch": 1.381984612792722,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004819383557750353,
      "loss": 5.1114,
      "mean_token_accuracy": 0.20217813104391097,
      "num_tokens": 27429552.0,
      "step": 12305
    },
    {
      "entropy": 5.482989549636841,
      "epoch": 1.3825461897006794,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048192259166449845,
      "loss": 5.321,
      "mean_token_accuracy": 0.19576178789138793,
      "num_tokens": 27440404.0,
      "step": 12310
    },
    {
      "entropy": 5.417792940139771,
      "epoch": 1.383107766608637,
      "grad_norm": 1.25,
      "learning_rate": 0.00048190682096546606,
      "loss": 5.3603,
      "mean_token_accuracy": 0.1929946705698967,
      "num_tokens": 27452231.0,
      "step": 12315
    },
    {
      "entropy": 5.418830013275146,
      "epoch": 1.3836693435165945,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048189104367844034,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19370912462472917,
      "num_tokens": 27463658.0,
      "step": 12320
    },
    {
      "entropy": 5.4801037311553955,
      "epoch": 1.3842309204245522,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048187525980392343,
      "loss": 5.3015,
      "mean_token_accuracy": 0.18862779885530473,
      "num_tokens": 27475403.0,
      "step": 12325
    },
    {
      "entropy": 5.5829824924469,
      "epoch": 1.3847924973325096,
      "grad_norm": 1.125,
      "learning_rate": 0.000481859469342418,
      "loss": 5.3862,
      "mean_token_accuracy": 0.1868772551417351,
      "num_tokens": 27487056.0,
      "step": 12330
    },
    {
      "entropy": 5.380106639862061,
      "epoch": 1.3853540742404673,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004818436722944269,
      "loss": 5.2528,
      "mean_token_accuracy": 0.1970052793622017,
      "num_tokens": 27498659.0,
      "step": 12335
    },
    {
      "entropy": 5.552019262313843,
      "epoch": 1.3859156511484247,
      "grad_norm": 1.4140625,
      "learning_rate": 0.0004818278686604531,
      "loss": 5.3662,
      "mean_token_accuracy": 0.19028019458055495,
      "num_tokens": 27510577.0,
      "step": 12340
    },
    {
      "entropy": 5.458247280120849,
      "epoch": 1.3864772280563824,
      "grad_norm": 1.25,
      "learning_rate": 0.0004818120584409996,
      "loss": 5.3048,
      "mean_token_accuracy": 0.1978745251893997,
      "num_tokens": 27521441.0,
      "step": 12345
    },
    {
      "entropy": 5.454830980300903,
      "epoch": 1.3870388049643398,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004817962416365701,
      "loss": 5.2361,
      "mean_token_accuracy": 0.19582115411758422,
      "num_tokens": 27533443.0,
      "step": 12350
    },
    {
      "entropy": 5.430822229385376,
      "epoch": 1.3876003818722973,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048178041824766796,
      "loss": 5.2832,
      "mean_token_accuracy": 0.19403909295797347,
      "num_tokens": 27543872.0,
      "step": 12355
    },
    {
      "entropy": 5.595981216430664,
      "epoch": 1.388161958780255,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004817645882747971,
      "loss": 5.4776,
      "mean_token_accuracy": 0.18229304552078246,
      "num_tokens": 27556556.0,
      "step": 12360
    },
    {
      "entropy": 5.489650917053223,
      "epoch": 1.3887235356882126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004817487517184616,
      "loss": 5.3393,
      "mean_token_accuracy": 0.18858043998479843,
      "num_tokens": 27570739.0,
      "step": 12365
    },
    {
      "entropy": 5.477973365783692,
      "epoch": 1.38928511259617,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004817329085791657,
      "loss": 5.3895,
      "mean_token_accuracy": 0.1846841812133789,
      "num_tokens": 27582438.0,
      "step": 12370
    },
    {
      "entropy": 5.435638380050659,
      "epoch": 1.3898466895041275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004817170588574138,
      "loss": 5.2839,
      "mean_token_accuracy": 0.1947597473859787,
      "num_tokens": 27593854.0,
      "step": 12375
    },
    {
      "entropy": 5.493896341323852,
      "epoch": 1.3904082664120851,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004817012025537105,
      "loss": 5.267,
      "mean_token_accuracy": 0.1963128924369812,
      "num_tokens": 27604459.0,
      "step": 12380
    },
    {
      "entropy": 5.539251184463501,
      "epoch": 1.3909698433200428,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004816853396685608,
      "loss": 5.4516,
      "mean_token_accuracy": 0.17707674503326415,
      "num_tokens": 27617901.0,
      "step": 12385
    },
    {
      "entropy": 5.547840309143067,
      "epoch": 1.3915314202280002,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048166947020246973,
      "loss": 5.3846,
      "mean_token_accuracy": 0.18501852750778197,
      "num_tokens": 27628574.0,
      "step": 12390
    },
    {
      "entropy": 5.482903671264649,
      "epoch": 1.3920929971359577,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048165359415594245,
      "loss": 5.2429,
      "mean_token_accuracy": 0.1928139090538025,
      "num_tokens": 27639195.0,
      "step": 12395
    },
    {
      "entropy": 5.37654938697815,
      "epoch": 1.3926545740439153,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048163771152948473,
      "loss": 5.1734,
      "mean_token_accuracy": 0.20025649070739746,
      "num_tokens": 27649898.0,
      "step": 12400
    },
    {
      "entropy": 5.448194599151611,
      "epoch": 1.3932161509518728,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000481621822323602,
      "loss": 5.3172,
      "mean_token_accuracy": 0.18566079586744308,
      "num_tokens": 27663617.0,
      "step": 12405
    },
    {
      "entropy": 5.382668590545654,
      "epoch": 1.3937777278598305,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004816059265388004,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19757655411958694,
      "num_tokens": 27674823.0,
      "step": 12410
    },
    {
      "entropy": 5.510045528411865,
      "epoch": 1.394339304767788,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048159002417558596,
      "loss": 5.3236,
      "mean_token_accuracy": 0.19013630896806716,
      "num_tokens": 27686483.0,
      "step": 12415
    },
    {
      "entropy": 5.437432193756104,
      "epoch": 1.3949008816757456,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004815741152344651,
      "loss": 5.2429,
      "mean_token_accuracy": 0.19867898672819137,
      "num_tokens": 27697081.0,
      "step": 12420
    },
    {
      "entropy": 5.447075271606446,
      "epoch": 1.395462458583703,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004815581997159442,
      "loss": 5.3184,
      "mean_token_accuracy": 0.18892811983823776,
      "num_tokens": 27706710.0,
      "step": 12425
    },
    {
      "entropy": 5.471676063537598,
      "epoch": 1.3960240354916607,
      "grad_norm": 1.09375,
      "learning_rate": 0.00048154227762053003,
      "loss": 5.3467,
      "mean_token_accuracy": 0.1842469573020935,
      "num_tokens": 27718213.0,
      "step": 12430
    },
    {
      "entropy": 5.440064191818237,
      "epoch": 1.396585612399618,
      "grad_norm": 1.46875,
      "learning_rate": 0.0004815263489487298,
      "loss": 5.2775,
      "mean_token_accuracy": 0.18891369551420212,
      "num_tokens": 27729064.0,
      "step": 12435
    },
    {
      "entropy": 5.4657378673553465,
      "epoch": 1.3971471893075758,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004815104137010504,
      "loss": 5.2618,
      "mean_token_accuracy": 0.19351084679365158,
      "num_tokens": 27740121.0,
      "step": 12440
    },
    {
      "entropy": 5.406544733047485,
      "epoch": 1.3977087662155332,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004814944718779993,
      "loss": 5.3259,
      "mean_token_accuracy": 0.19067021757364272,
      "num_tokens": 27751597.0,
      "step": 12445
    },
    {
      "entropy": 5.411537933349609,
      "epoch": 1.3982703431234906,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004814785234800843,
      "loss": 5.2134,
      "mean_token_accuracy": 0.1942956641316414,
      "num_tokens": 27761539.0,
      "step": 12450
    },
    {
      "entropy": 5.422007513046265,
      "epoch": 1.3988319200314483,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048146256850781287,
      "loss": 5.2453,
      "mean_token_accuracy": 0.1897090658545494,
      "num_tokens": 27772904.0,
      "step": 12455
    },
    {
      "entropy": 5.500157976150513,
      "epoch": 1.399393496939406,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00048144660696169327,
      "loss": 5.245,
      "mean_token_accuracy": 0.1969943732023239,
      "num_tokens": 27783706.0,
      "step": 12460
    },
    {
      "entropy": 5.399797868728638,
      "epoch": 1.3999550738473634,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048143063884223365,
      "loss": 5.2528,
      "mean_token_accuracy": 0.18924653232097627,
      "num_tokens": 27794503.0,
      "step": 12465
    },
    {
      "entropy": 5.381992435455322,
      "epoch": 1.4005166507553208,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004814146641499424,
      "loss": 5.2722,
      "mean_token_accuracy": 0.19058639854192733,
      "num_tokens": 27805952.0,
      "step": 12470
    },
    {
      "entropy": 5.465413188934326,
      "epoch": 1.4010782276632785,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00048139868288532816,
      "loss": 5.2485,
      "mean_token_accuracy": 0.1952443838119507,
      "num_tokens": 27817530.0,
      "step": 12475
    },
    {
      "entropy": 5.495134878158569,
      "epoch": 1.4016398045712362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004813826950488998,
      "loss": 5.3336,
      "mean_token_accuracy": 0.1834641456604004,
      "num_tokens": 27829932.0,
      "step": 12480
    },
    {
      "entropy": 5.522836923599243,
      "epoch": 1.4022013814791936,
      "grad_norm": 1.109375,
      "learning_rate": 0.00048136670064116645,
      "loss": 5.3161,
      "mean_token_accuracy": 0.19068046510219575,
      "num_tokens": 27840993.0,
      "step": 12485
    },
    {
      "entropy": 5.449536848068237,
      "epoch": 1.402762958387151,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004813506996626373,
      "loss": 5.2432,
      "mean_token_accuracy": 0.19361874163150788,
      "num_tokens": 27852727.0,
      "step": 12490
    },
    {
      "entropy": 5.4345457553863525,
      "epoch": 1.4033245352951087,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004813346921138218,
      "loss": 5.405,
      "mean_token_accuracy": 0.19102840572595597,
      "num_tokens": 27865183.0,
      "step": 12495
    },
    {
      "entropy": 5.415298843383789,
      "epoch": 1.4038861122030661,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004813186779952297,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20114682763814926,
      "num_tokens": 27877266.0,
      "step": 12500
    },
    {
      "entropy": 5.485949516296387,
      "epoch": 1.4044476891110238,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004813026573073708,
      "loss": 5.305,
      "mean_token_accuracy": 0.18829656541347503,
      "num_tokens": 27888863.0,
      "step": 12505
    },
    {
      "entropy": 5.411046409606934,
      "epoch": 1.4050092660189812,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004812866300507554,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1942449614405632,
      "num_tokens": 27900443.0,
      "step": 12510
    },
    {
      "entropy": 5.490506553649903,
      "epoch": 1.405570842926939,
      "grad_norm": 1.484375,
      "learning_rate": 0.0004812705962258936,
      "loss": 5.2219,
      "mean_token_accuracy": 0.1921681597828865,
      "num_tokens": 27911241.0,
      "step": 12515
    },
    {
      "entropy": 5.424184274673462,
      "epoch": 1.4061324198348963,
      "grad_norm": 1.125,
      "learning_rate": 0.0004812545558332961,
      "loss": 5.2512,
      "mean_token_accuracy": 0.1953155905008316,
      "num_tokens": 27922171.0,
      "step": 12520
    },
    {
      "entropy": 5.341803121566772,
      "epoch": 1.406693996742854,
      "grad_norm": 1.140625,
      "learning_rate": 0.00048123850887347346,
      "loss": 5.2039,
      "mean_token_accuracy": 0.19879241585731505,
      "num_tokens": 27933627.0,
      "step": 12525
    },
    {
      "entropy": 5.275621795654297,
      "epoch": 1.4072555736508114,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048122245534693676,
      "loss": 5.1699,
      "mean_token_accuracy": 0.1921117275953293,
      "num_tokens": 27943585.0,
      "step": 12530
    },
    {
      "entropy": 5.482760095596314,
      "epoch": 1.407817150558769,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004812063952541971,
      "loss": 5.29,
      "mean_token_accuracy": 0.1889578953385353,
      "num_tokens": 27955554.0,
      "step": 12535
    },
    {
      "entropy": 5.481731128692627,
      "epoch": 1.4083787274667265,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048119032859576577,
      "loss": 5.1666,
      "mean_token_accuracy": 0.2004450485110283,
      "num_tokens": 27966150.0,
      "step": 12540
    },
    {
      "entropy": 5.326528739929199,
      "epoch": 1.408940304374684,
      "grad_norm": 1.3359375,
      "learning_rate": 0.0004811742553721545,
      "loss": 5.2121,
      "mean_token_accuracy": 0.19881806075572966,
      "num_tokens": 27977436.0,
      "step": 12545
    },
    {
      "entropy": 5.486036205291748,
      "epoch": 1.4095018812826416,
      "grad_norm": 1.296875,
      "learning_rate": 0.00048115817558387494,
      "loss": 5.3782,
      "mean_token_accuracy": 0.18544383198022843,
      "num_tokens": 27988325.0,
      "step": 12550
    },
    {
      "entropy": 5.578227758407593,
      "epoch": 1.4100634581905993,
      "grad_norm": 1.25,
      "learning_rate": 0.0004811420892314391,
      "loss": 5.2036,
      "mean_token_accuracy": 0.1986285537481308,
      "num_tokens": 27999760.0,
      "step": 12555
    },
    {
      "entropy": 5.464999103546143,
      "epoch": 1.4106250350985567,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004811259963153592,
      "loss": 5.3657,
      "mean_token_accuracy": 0.18548568338155746,
      "num_tokens": 28011333.0,
      "step": 12560
    },
    {
      "entropy": 5.3490626335144045,
      "epoch": 1.4111866120065142,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004811098968361476,
      "loss": 5.323,
      "mean_token_accuracy": 0.19722088128328324,
      "num_tokens": 28023464.0,
      "step": 12565
    },
    {
      "entropy": 5.425189542770386,
      "epoch": 1.4117481889144718,
      "grad_norm": 1.28125,
      "learning_rate": 0.000481093790794317,
      "loss": 5.2643,
      "mean_token_accuracy": 0.1885809928178787,
      "num_tokens": 28034205.0,
      "step": 12570
    },
    {
      "entropy": 5.542120027542114,
      "epoch": 1.4123097658224295,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004810776781903801,
      "loss": 5.264,
      "mean_token_accuracy": 0.18553368151187896,
      "num_tokens": 28045846.0,
      "step": 12575
    },
    {
      "entropy": 5.464288377761841,
      "epoch": 1.412871342730387,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048106155902485003,
      "loss": 5.2748,
      "mean_token_accuracy": 0.19653246998786927,
      "num_tokens": 28057365.0,
      "step": 12580
    },
    {
      "entropy": 5.467927646636963,
      "epoch": 1.4134329196383444,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004810454332982401,
      "loss": 5.3498,
      "mean_token_accuracy": 0.18708886355161666,
      "num_tokens": 28069098.0,
      "step": 12585
    },
    {
      "entropy": 5.458622312545776,
      "epoch": 1.413994496546302,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004810293010110635,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19389213174581527,
      "num_tokens": 28080008.0,
      "step": 12590
    },
    {
      "entropy": 5.5056215763092045,
      "epoch": 1.4145560734542595,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004810131621638342,
      "loss": 5.314,
      "mean_token_accuracy": 0.19194261580705643,
      "num_tokens": 28091657.0,
      "step": 12595
    },
    {
      "entropy": 5.381366348266601,
      "epoch": 1.4151176503622171,
      "grad_norm": 1.234375,
      "learning_rate": 0.00048099701675706583,
      "loss": 5.2096,
      "mean_token_accuracy": 0.19305577129125595,
      "num_tokens": 28102586.0,
      "step": 12600
    },
    {
      "entropy": 5.431563520431519,
      "epoch": 1.4156792272701746,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004809808647912725,
      "loss": 5.3097,
      "mean_token_accuracy": 0.19521011114120485,
      "num_tokens": 28113066.0,
      "step": 12605
    },
    {
      "entropy": 5.434861183166504,
      "epoch": 1.4162408041781323,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004809647062669686,
      "loss": 5.2927,
      "mean_token_accuracy": 0.1846861571073532,
      "num_tokens": 28124814.0,
      "step": 12610
    },
    {
      "entropy": 5.505294561386108,
      "epoch": 1.4168023810860897,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00048094854118466856,
      "loss": 5.3405,
      "mean_token_accuracy": 0.18603829592466353,
      "num_tokens": 28138009.0,
      "step": 12615
    },
    {
      "entropy": 5.38302001953125,
      "epoch": 1.4173639579940474,
      "grad_norm": 1.28125,
      "learning_rate": 0.00048093236954488715,
      "loss": 5.2327,
      "mean_token_accuracy": 0.19298163503408433,
      "num_tokens": 28149222.0,
      "step": 12620
    },
    {
      "entropy": 5.463192176818848,
      "epoch": 1.4179255349020048,
      "grad_norm": 1.125,
      "learning_rate": 0.0004809161913481392,
      "loss": 5.2919,
      "mean_token_accuracy": 0.18981857299804689,
      "num_tokens": 28161323.0,
      "step": 12625
    },
    {
      "entropy": 5.432276821136474,
      "epoch": 1.4184871118099625,
      "grad_norm": 1.328125,
      "learning_rate": 0.00048090000659493985,
      "loss": 5.2737,
      "mean_token_accuracy": 0.19063198417425156,
      "num_tokens": 28173242.0,
      "step": 12630
    },
    {
      "entropy": 5.42558479309082,
      "epoch": 1.41904868871792,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048088381528580445,
      "loss": 5.2362,
      "mean_token_accuracy": 0.194551520049572,
      "num_tokens": 28183927.0,
      "step": 12635
    },
    {
      "entropy": 5.438029098510742,
      "epoch": 1.4196102656258773,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00048086761742124846,
      "loss": 5.2477,
      "mean_token_accuracy": 0.1936724454164505,
      "num_tokens": 28194375.0,
      "step": 12640
    },
    {
      "entropy": 5.476759767532348,
      "epoch": 1.420171842533835,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00048085141300178774,
      "loss": 5.3128,
      "mean_token_accuracy": 0.19387400150299072,
      "num_tokens": 28206492.0,
      "step": 12645
    },
    {
      "entropy": 5.457561016082764,
      "epoch": 1.4207334194417927,
      "grad_norm": 1.25,
      "learning_rate": 0.00048083520202793815,
      "loss": 5.3287,
      "mean_token_accuracy": 0.18830227106809616,
      "num_tokens": 28217795.0,
      "step": 12650
    },
    {
      "entropy": 5.4019232273101805,
      "epoch": 1.42129499634975,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004808189845002159,
      "loss": 5.2836,
      "mean_token_accuracy": 0.192978173494339,
      "num_tokens": 28228796.0,
      "step": 12655
    },
    {
      "entropy": 5.475896978378296,
      "epoch": 1.4218565732577075,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004808027604191374,
      "loss": 5.3088,
      "mean_token_accuracy": 0.19147174507379533,
      "num_tokens": 28238873.0,
      "step": 12660
    },
    {
      "entropy": 5.380402851104736,
      "epoch": 1.4224181501656652,
      "grad_norm": 1.125,
      "learning_rate": 0.00048078652978521914,
      "loss": 5.0968,
      "mean_token_accuracy": 0.20774092823266982,
      "num_tokens": 28249832.0,
      "step": 12665
    },
    {
      "entropy": 5.37987756729126,
      "epoch": 1.4229797270736229,
      "grad_norm": 1.125,
      "learning_rate": 0.00048077029259897797,
      "loss": 5.3111,
      "mean_token_accuracy": 0.18598005622625352,
      "num_tokens": 28261532.0,
      "step": 12670
    },
    {
      "entropy": 5.548765707015991,
      "epoch": 1.4235413039815803,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00048075404886093094,
      "loss": 5.3204,
      "mean_token_accuracy": 0.19131788462400437,
      "num_tokens": 28272008.0,
      "step": 12675
    },
    {
      "entropy": 5.457288789749145,
      "epoch": 1.4241028808895377,
      "grad_norm": 1.3125,
      "learning_rate": 0.00048073779857159504,
      "loss": 5.2962,
      "mean_token_accuracy": 0.1940157651901245,
      "num_tokens": 28283239.0,
      "step": 12680
    },
    {
      "entropy": 5.49482479095459,
      "epoch": 1.4246644577974954,
      "grad_norm": 1.125,
      "learning_rate": 0.000480721541731488,
      "loss": 5.343,
      "mean_token_accuracy": 0.18934712707996368,
      "num_tokens": 28293597.0,
      "step": 12685
    },
    {
      "entropy": 5.458812284469604,
      "epoch": 1.4252260347054528,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048070527834112727,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19474604725837708,
      "num_tokens": 28305124.0,
      "step": 12690
    },
    {
      "entropy": 5.368135786056518,
      "epoch": 1.4257876116134105,
      "grad_norm": 1.125,
      "learning_rate": 0.0004806890084010306,
      "loss": 5.2543,
      "mean_token_accuracy": 0.1978668138384819,
      "num_tokens": 28315875.0,
      "step": 12695
    },
    {
      "entropy": 5.413513040542602,
      "epoch": 1.426349188521368,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00048067273191171615,
      "loss": 5.2547,
      "mean_token_accuracy": 0.1947520613670349,
      "num_tokens": 28327705.0,
      "step": 12700
    },
    {
      "entropy": 5.416609907150269,
      "epoch": 1.4269107654293256,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004806564488737022,
      "loss": 5.2323,
      "mean_token_accuracy": 0.19762539863586426,
      "num_tokens": 28338300.0,
      "step": 12705
    },
    {
      "entropy": 5.429306983947754,
      "epoch": 1.427472342337283,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004806401592875071,
      "loss": 5.1799,
      "mean_token_accuracy": 0.20126354098320007,
      "num_tokens": 28348485.0,
      "step": 12710
    },
    {
      "entropy": 5.433665609359741,
      "epoch": 1.4280339192452407,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048062386315364964,
      "loss": 5.2313,
      "mean_token_accuracy": 0.19867527037858962,
      "num_tokens": 28359118.0,
      "step": 12715
    },
    {
      "entropy": 5.405245542526245,
      "epoch": 1.4285954961531981,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004806075604726485,
      "loss": 5.2499,
      "mean_token_accuracy": 0.19774582087993622,
      "num_tokens": 28369608.0,
      "step": 12720
    },
    {
      "entropy": 5.458242464065552,
      "epoch": 1.4291570730611558,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048059125124502304,
      "loss": 5.2385,
      "mean_token_accuracy": 0.1999989166855812,
      "num_tokens": 28379709.0,
      "step": 12725
    },
    {
      "entropy": 5.433728170394898,
      "epoch": 1.4297186499691132,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004805749354712924,
      "loss": 5.2744,
      "mean_token_accuracy": 0.1899983122944832,
      "num_tokens": 28390758.0,
      "step": 12730
    },
    {
      "entropy": 5.600304508209229,
      "epoch": 1.4302802268770707,
      "grad_norm": 1.203125,
      "learning_rate": 0.000480558613151976,
      "loss": 5.5315,
      "mean_token_accuracy": 0.17958250641822815,
      "num_tokens": 28403485.0,
      "step": 12735
    },
    {
      "entropy": 5.542997264862061,
      "epoch": 1.4308418037850283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004805422842875936,
      "loss": 5.3274,
      "mean_token_accuracy": 0.18867387771606445,
      "num_tokens": 28413627.0,
      "step": 12740
    },
    {
      "entropy": 5.412322616577148,
      "epoch": 1.431403380692986,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00048052594887866524,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19367787837982178,
      "num_tokens": 28424202.0,
      "step": 12745
    },
    {
      "entropy": 5.3527992248535154,
      "epoch": 1.4319649576009434,
      "grad_norm": 1.1875,
      "learning_rate": 0.00048050960692571086,
      "loss": 5.1895,
      "mean_token_accuracy": 0.19925883859395982,
      "num_tokens": 28435136.0,
      "step": 12750
    },
    {
      "entropy": 5.4476696968078615,
      "epoch": 1.4325265345089009,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000480493258429251,
      "loss": 5.21,
      "mean_token_accuracy": 0.1932688683271408,
      "num_tokens": 28444915.0,
      "step": 12755
    },
    {
      "entropy": 5.436851072311401,
      "epoch": 1.4330881114168585,
      "grad_norm": 1.234375,
      "learning_rate": 0.000480476903389806,
      "loss": 5.2779,
      "mean_token_accuracy": 0.18963428288698198,
      "num_tokens": 28458295.0,
      "step": 12760
    },
    {
      "entropy": 5.468523788452148,
      "epoch": 1.4336496883248162,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004804605418078968,
      "loss": 5.3194,
      "mean_token_accuracy": 0.1955967664718628,
      "num_tokens": 28469587.0,
      "step": 12765
    },
    {
      "entropy": 5.521590709686279,
      "epoch": 1.4342112652327736,
      "grad_norm": 1.203125,
      "learning_rate": 0.00048044417368404426,
      "loss": 5.3062,
      "mean_token_accuracy": 0.18945180326700212,
      "num_tokens": 28481498.0,
      "step": 12770
    },
    {
      "entropy": 5.4691017150878904,
      "epoch": 1.434772842140731,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048042779901876946,
      "loss": 5.4276,
      "mean_token_accuracy": 0.18312709778547287,
      "num_tokens": 28492626.0,
      "step": 12775
    },
    {
      "entropy": 5.551679277420044,
      "epoch": 1.4353344190486887,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004804114178125939,
      "loss": 5.3065,
      "mean_token_accuracy": 0.19160436391830443,
      "num_tokens": 28503882.0,
      "step": 12780
    },
    {
      "entropy": 5.469602108001709,
      "epoch": 1.4358959959566462,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004803950300660391,
      "loss": 5.2334,
      "mean_token_accuracy": 0.20204749554395676,
      "num_tokens": 28514786.0,
      "step": 12785
    },
    {
      "entropy": 5.443823385238647,
      "epoch": 1.4364575728646038,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004803786357796269,
      "loss": 5.2495,
      "mean_token_accuracy": 0.1930367171764374,
      "num_tokens": 28526045.0,
      "step": 12790
    },
    {
      "entropy": 5.516070890426636,
      "epoch": 1.4370191497725613,
      "grad_norm": 1.265625,
      "learning_rate": 0.00048036223495387934,
      "loss": 5.2752,
      "mean_token_accuracy": 0.1906992256641388,
      "num_tokens": 28536773.0,
      "step": 12795
    },
    {
      "entropy": 5.389249515533447,
      "epoch": 1.437580726680519,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004803458275893185,
      "loss": 5.2256,
      "mean_token_accuracy": 0.19098111242055893,
      "num_tokens": 28548099.0,
      "step": 12800
    },
    {
      "entropy": 5.417944860458374,
      "epoch": 1.4381423035884764,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004803294136864669,
      "loss": 5.2811,
      "mean_token_accuracy": 0.19215987920761107,
      "num_tokens": 28559372.0,
      "step": 12805
    },
    {
      "entropy": 5.469154977798462,
      "epoch": 1.438703880496434,
      "grad_norm": 1.078125,
      "learning_rate": 0.00048031299324584716,
      "loss": 5.3043,
      "mean_token_accuracy": 0.1955375552177429,
      "num_tokens": 28570106.0,
      "step": 12810
    },
    {
      "entropy": 5.458250570297241,
      "epoch": 1.4392654574043915,
      "grad_norm": 1.15625,
      "learning_rate": 0.000480296566267982,
      "loss": 5.2883,
      "mean_token_accuracy": 0.19549554586410522,
      "num_tokens": 28582417.0,
      "step": 12815
    },
    {
      "entropy": 5.495559787750244,
      "epoch": 1.4398270343123492,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004802801327533947,
      "loss": 5.4153,
      "mean_token_accuracy": 0.19176921397447586,
      "num_tokens": 28594341.0,
      "step": 12820
    },
    {
      "entropy": 5.45522723197937,
      "epoch": 1.4403886112203066,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00048026369270260814,
      "loss": 5.29,
      "mean_token_accuracy": 0.19270292669534683,
      "num_tokens": 28605580.0,
      "step": 12825
    },
    {
      "entropy": 5.576689147949219,
      "epoch": 1.440950188128264,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004802472461161462,
      "loss": 5.3284,
      "mean_token_accuracy": 0.18985722810029984,
      "num_tokens": 28617779.0,
      "step": 12830
    },
    {
      "entropy": 5.459845542907715,
      "epoch": 1.4415117650362217,
      "grad_norm": 1.25,
      "learning_rate": 0.0004802307929945322,
      "loss": 5.328,
      "mean_token_accuracy": 0.18627672493457795,
      "num_tokens": 28629714.0,
      "step": 12835
    },
    {
      "entropy": 5.495349311828614,
      "epoch": 1.4420733419441794,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048021433333829015,
      "loss": 5.3235,
      "mean_token_accuracy": 0.18933959901332856,
      "num_tokens": 28641381.0,
      "step": 12840
    },
    {
      "entropy": 5.488303232192993,
      "epoch": 1.4426349188521368,
      "grad_norm": 1.21875,
      "learning_rate": 0.00048019786714794424,
      "loss": 5.3117,
      "mean_token_accuracy": 0.19368012845516205,
      "num_tokens": 28653650.0,
      "step": 12845
    },
    {
      "entropy": 5.424508380889892,
      "epoch": 1.4431964957600942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004801813944240186,
      "loss": 5.2492,
      "mean_token_accuracy": 0.1951059490442276,
      "num_tokens": 28665252.0,
      "step": 12850
    },
    {
      "entropy": 5.402679777145385,
      "epoch": 1.443758072668052,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004801649151670378,
      "loss": 5.2795,
      "mean_token_accuracy": 0.19484080374240875,
      "num_tokens": 28677072.0,
      "step": 12855
    },
    {
      "entropy": 5.391095542907715,
      "epoch": 1.4443196495760096,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004801484293775265,
      "loss": 5.2147,
      "mean_token_accuracy": 0.19180794805288315,
      "num_tokens": 28688767.0,
      "step": 12860
    },
    {
      "entropy": 5.510450029373169,
      "epoch": 1.444881226483967,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004801319370560097,
      "loss": 5.4707,
      "mean_token_accuracy": 0.18604153394699097,
      "num_tokens": 28699773.0,
      "step": 12865
    },
    {
      "entropy": 5.505920028686523,
      "epoch": 1.4454428033919244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004801154382030124,
      "loss": 5.3372,
      "mean_token_accuracy": 0.19013401716947556,
      "num_tokens": 28710971.0,
      "step": 12870
    },
    {
      "entropy": 5.421435356140137,
      "epoch": 1.446004380299882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00048009893281906004,
      "loss": 5.1995,
      "mean_token_accuracy": 0.19837294667959213,
      "num_tokens": 28721544.0,
      "step": 12875
    },
    {
      "entropy": 5.490336751937866,
      "epoch": 1.4465659572078395,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00048008242090467803,
      "loss": 5.4159,
      "mean_token_accuracy": 0.18639136701822281,
      "num_tokens": 28733790.0,
      "step": 12880
    },
    {
      "entropy": 5.52124981880188,
      "epoch": 1.4471275341157972,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004800659024603923,
      "loss": 5.32,
      "mean_token_accuracy": 0.19438826888799668,
      "num_tokens": 28745218.0,
      "step": 12885
    },
    {
      "entropy": 5.516373491287231,
      "epoch": 1.4476891110237546,
      "grad_norm": 1.15625,
      "learning_rate": 0.00048004937748672864,
      "loss": 5.3739,
      "mean_token_accuracy": 0.18650627583265306,
      "num_tokens": 28758061.0,
      "step": 12890
    },
    {
      "entropy": 5.440051794052124,
      "epoch": 1.4482506879317123,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00048003284598421327,
      "loss": 5.187,
      "mean_token_accuracy": 0.2026249349117279,
      "num_tokens": 28770037.0,
      "step": 12895
    },
    {
      "entropy": 5.446256256103515,
      "epoch": 1.4488122648396697,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004800163079533726,
      "loss": 5.3558,
      "mean_token_accuracy": 0.18772219866514206,
      "num_tokens": 28781052.0,
      "step": 12900
    },
    {
      "entropy": 5.3748777389526365,
      "epoch": 1.4493738417476274,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004799997633947332,
      "loss": 5.2267,
      "mean_token_accuracy": 0.18345671743154526,
      "num_tokens": 28791225.0,
      "step": 12905
    },
    {
      "entropy": 5.498763561248779,
      "epoch": 1.4499354186555848,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047998321230882174,
      "loss": 5.2493,
      "mean_token_accuracy": 0.1936606228351593,
      "num_tokens": 28803022.0,
      "step": 12910
    },
    {
      "entropy": 5.385029983520508,
      "epoch": 1.4504969955635425,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047996665469616536,
      "loss": 5.24,
      "mean_token_accuracy": 0.1913703441619873,
      "num_tokens": 28813789.0,
      "step": 12915
    },
    {
      "entropy": 5.505116653442383,
      "epoch": 1.4510585724715,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047995009055729114,
      "loss": 5.3018,
      "mean_token_accuracy": 0.19192540943622588,
      "num_tokens": 28825015.0,
      "step": 12920
    },
    {
      "entropy": 5.5920023918151855,
      "epoch": 1.4516201493794574,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004799335198927266,
      "loss": 5.3917,
      "mean_token_accuracy": 0.1863776206970215,
      "num_tokens": 28835852.0,
      "step": 12925
    },
    {
      "entropy": 5.371799468994141,
      "epoch": 1.452181726287415,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004799169427029993,
      "loss": 5.3142,
      "mean_token_accuracy": 0.19285976141691208,
      "num_tokens": 28846684.0,
      "step": 12930
    },
    {
      "entropy": 5.436971616744995,
      "epoch": 1.4527433031953727,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047990035898863706,
      "loss": 5.3183,
      "mean_token_accuracy": 0.18753400295972825,
      "num_tokens": 28857402.0,
      "step": 12935
    },
    {
      "entropy": 5.396050834655762,
      "epoch": 1.4533048801033301,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004798837687501679,
      "loss": 5.1537,
      "mean_token_accuracy": 0.20105982273817063,
      "num_tokens": 28867866.0,
      "step": 12940
    },
    {
      "entropy": 5.407264661788941,
      "epoch": 1.4538664570112876,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004798671719881201,
      "loss": 5.2607,
      "mean_token_accuracy": 0.19380312114953996,
      "num_tokens": 28878102.0,
      "step": 12945
    },
    {
      "entropy": 5.519525527954102,
      "epoch": 1.4544280339192452,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004798505687030221,
      "loss": 5.373,
      "mean_token_accuracy": 0.18923720419406892,
      "num_tokens": 28888981.0,
      "step": 12950
    },
    {
      "entropy": 5.370624828338623,
      "epoch": 1.454989610827203,
      "grad_norm": 1.3828125,
      "learning_rate": 0.0004798339588954025,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19115737229585647,
      "num_tokens": 28899733.0,
      "step": 12955
    },
    {
      "entropy": 5.393942880630493,
      "epoch": 1.4555511877351603,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004798173425657903,
      "loss": 5.2102,
      "mean_token_accuracy": 0.19257485568523408,
      "num_tokens": 28910701.0,
      "step": 12960
    },
    {
      "entropy": 5.511909294128418,
      "epoch": 1.4561127646431178,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004798007197147144,
      "loss": 5.2921,
      "mean_token_accuracy": 0.18703228533267974,
      "num_tokens": 28922627.0,
      "step": 12965
    },
    {
      "entropy": 5.515297889709473,
      "epoch": 1.4566743415510754,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797840903427042,
      "loss": 5.3349,
      "mean_token_accuracy": 0.18688380867242813,
      "num_tokens": 28933292.0,
      "step": 12970
    },
    {
      "entropy": 5.380740165710449,
      "epoch": 1.4572359184590329,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047976745445028906,
      "loss": 5.1994,
      "mean_token_accuracy": 0.1961256742477417,
      "num_tokens": 28943898.0,
      "step": 12975
    },
    {
      "entropy": 5.387813949584961,
      "epoch": 1.4577974953669905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00047975081203799883,
      "loss": 5.3048,
      "mean_token_accuracy": 0.18965622335672377,
      "num_tokens": 28955248.0,
      "step": 12980
    },
    {
      "entropy": 5.4105384826660154,
      "epoch": 1.458359072274948,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047973416310636327,
      "loss": 5.1696,
      "mean_token_accuracy": 0.19757775515317916,
      "num_tokens": 28966905.0,
      "step": 12985
    },
    {
      "entropy": 5.480452346801758,
      "epoch": 1.4589206491829056,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004797175076559126,
      "loss": 5.3195,
      "mean_token_accuracy": 0.19408757388591766,
      "num_tokens": 28978515.0,
      "step": 12990
    },
    {
      "entropy": 5.422642230987549,
      "epoch": 1.459482226090863,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047970084568717703,
      "loss": 5.2711,
      "mean_token_accuracy": 0.18703113049268721,
      "num_tokens": 28990185.0,
      "step": 12995
    },
    {
      "entropy": 5.475670051574707,
      "epoch": 1.4600438029988208,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004796841772006872,
      "loss": 5.3003,
      "mean_token_accuracy": 0.18595077693462372,
      "num_tokens": 29000779.0,
      "step": 13000
    },
    {
      "entropy": 5.531659746170044,
      "epoch": 1.4606053799067782,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047966750219697376,
      "loss": 5.3994,
      "mean_token_accuracy": 0.18907374739646912,
      "num_tokens": 29012066.0,
      "step": 13005
    },
    {
      "entropy": 5.571883249282837,
      "epoch": 1.4611669568147359,
      "grad_norm": 1.125,
      "learning_rate": 0.0004796508206765677,
      "loss": 5.4076,
      "mean_token_accuracy": 0.1872843846678734,
      "num_tokens": 29023416.0,
      "step": 13010
    },
    {
      "entropy": 5.459448385238647,
      "epoch": 1.4617285337226933,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047963413264000005,
      "loss": 5.2423,
      "mean_token_accuracy": 0.1921609327197075,
      "num_tokens": 29034469.0,
      "step": 13015
    },
    {
      "entropy": 5.428943872451782,
      "epoch": 1.4622901106306507,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004796174380878023,
      "loss": 5.2262,
      "mean_token_accuracy": 0.18830537050962448,
      "num_tokens": 29045306.0,
      "step": 13020
    },
    {
      "entropy": 5.358241128921509,
      "epoch": 1.4628516875386084,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000479600737020506,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19370398223400115,
      "num_tokens": 29057114.0,
      "step": 13025
    },
    {
      "entropy": 5.41494779586792,
      "epoch": 1.463413264446566,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004795840294386429,
      "loss": 5.2044,
      "mean_token_accuracy": 0.18830911070108414,
      "num_tokens": 29067250.0,
      "step": 13030
    },
    {
      "entropy": 5.345307350158691,
      "epoch": 1.4639748413545235,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047956731534274483,
      "loss": 5.2335,
      "mean_token_accuracy": 0.1982669487595558,
      "num_tokens": 29078582.0,
      "step": 13035
    },
    {
      "entropy": 5.417934465408325,
      "epoch": 1.464536418262481,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047955059473334425,
      "loss": 5.2763,
      "mean_token_accuracy": 0.1979276642203331,
      "num_tokens": 29090113.0,
      "step": 13040
    },
    {
      "entropy": 5.424048709869385,
      "epoch": 1.4650979951704386,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004795338676109733,
      "loss": 5.2662,
      "mean_token_accuracy": 0.19380177706480026,
      "num_tokens": 29101117.0,
      "step": 13045
    },
    {
      "entropy": 5.513153839111328,
      "epoch": 1.4656595720783963,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004795171339761647,
      "loss": 5.2693,
      "mean_token_accuracy": 0.19244939982891082,
      "num_tokens": 29112569.0,
      "step": 13050
    },
    {
      "entropy": 5.4974036693573,
      "epoch": 1.4662211489863537,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047950039382945126,
      "loss": 5.3662,
      "mean_token_accuracy": 0.1952400267124176,
      "num_tokens": 29124462.0,
      "step": 13055
    },
    {
      "entropy": 5.45172438621521,
      "epoch": 1.4667827258943111,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047948364717136587,
      "loss": 5.221,
      "mean_token_accuracy": 0.1987057611346245,
      "num_tokens": 29135401.0,
      "step": 13060
    },
    {
      "entropy": 5.453978872299194,
      "epoch": 1.4673443028022688,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047946689400244196,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18796801567077637,
      "num_tokens": 29146601.0,
      "step": 13065
    },
    {
      "entropy": 5.430497837066651,
      "epoch": 1.4679058797102262,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047945013432321284,
      "loss": 5.2524,
      "mean_token_accuracy": 0.19588834941387176,
      "num_tokens": 29157299.0,
      "step": 13070
    },
    {
      "entropy": 5.535222053527832,
      "epoch": 1.468467456618184,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047943336813421205,
      "loss": 5.346,
      "mean_token_accuracy": 0.18766935169696808,
      "num_tokens": 29169101.0,
      "step": 13075
    },
    {
      "entropy": 5.3304276943206785,
      "epoch": 1.4690290335261413,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004794165954359736,
      "loss": 5.1704,
      "mean_token_accuracy": 0.1956697404384613,
      "num_tokens": 29181286.0,
      "step": 13080
    },
    {
      "entropy": 5.472303009033203,
      "epoch": 1.469590610434099,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004793998162290315,
      "loss": 5.3482,
      "mean_token_accuracy": 0.18582204580307007,
      "num_tokens": 29191181.0,
      "step": 13085
    },
    {
      "entropy": 5.379495334625244,
      "epoch": 1.4701521873420564,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004793830305139199,
      "loss": 5.203,
      "mean_token_accuracy": 0.19309724271297454,
      "num_tokens": 29202426.0,
      "step": 13090
    },
    {
      "entropy": 5.354828977584839,
      "epoch": 1.470713764250014,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004793662382911734,
      "loss": 5.2156,
      "mean_token_accuracy": 0.19222575426101685,
      "num_tokens": 29212817.0,
      "step": 13095
    },
    {
      "entropy": 5.387184906005859,
      "epoch": 1.4712753411579715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004793494395613265,
      "loss": 5.2205,
      "mean_token_accuracy": 0.1931103527545929,
      "num_tokens": 29224644.0,
      "step": 13100
    },
    {
      "entropy": 5.4386228084564205,
      "epoch": 1.4718369180659292,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004793326343249144,
      "loss": 5.2362,
      "mean_token_accuracy": 0.19114460200071334,
      "num_tokens": 29236443.0,
      "step": 13105
    },
    {
      "entropy": 5.489503860473633,
      "epoch": 1.4723984949738866,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004793158225824718,
      "loss": 5.3902,
      "mean_token_accuracy": 0.18319057524204255,
      "num_tokens": 29247687.0,
      "step": 13110
    },
    {
      "entropy": 5.491920757293701,
      "epoch": 1.4729600718818443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047929900433453423,
      "loss": 5.2834,
      "mean_token_accuracy": 0.19282291382551192,
      "num_tokens": 29259739.0,
      "step": 13115
    },
    {
      "entropy": 5.470120477676391,
      "epoch": 1.4735216487898017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047928217958163703,
      "loss": 5.3047,
      "mean_token_accuracy": 0.19429208636283873,
      "num_tokens": 29270834.0,
      "step": 13120
    },
    {
      "entropy": 5.454315948486328,
      "epoch": 1.4740832256977594,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047926534832431615,
      "loss": 5.2738,
      "mean_token_accuracy": 0.19493975788354873,
      "num_tokens": 29282291.0,
      "step": 13125
    },
    {
      "entropy": 5.4565431594848635,
      "epoch": 1.4746448026057168,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004792485105631073,
      "loss": 5.3154,
      "mean_token_accuracy": 0.1895140990614891,
      "num_tokens": 29293836.0,
      "step": 13130
    },
    {
      "entropy": 5.473170185089112,
      "epoch": 1.4752063795136743,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004792316662985467,
      "loss": 5.3428,
      "mean_token_accuracy": 0.18391882330179216,
      "num_tokens": 29305550.0,
      "step": 13135
    },
    {
      "entropy": 5.509187078475952,
      "epoch": 1.475767956421632,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047921481553117053,
      "loss": 5.3219,
      "mean_token_accuracy": 0.18953231275081633,
      "num_tokens": 29316973.0,
      "step": 13140
    },
    {
      "entropy": 5.420145750045776,
      "epoch": 1.4763295333295896,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047919795826151544,
      "loss": 5.1845,
      "mean_token_accuracy": 0.19443155229091644,
      "num_tokens": 29327743.0,
      "step": 13145
    },
    {
      "entropy": 5.366609811782837,
      "epoch": 1.476891110237547,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791810944901183,
      "loss": 5.1709,
      "mean_token_accuracy": 0.18876693695783614,
      "num_tokens": 29338286.0,
      "step": 13150
    },
    {
      "entropy": 5.408519220352173,
      "epoch": 1.4774526871455045,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004791642242175158,
      "loss": 5.3075,
      "mean_token_accuracy": 0.1913115054368973,
      "num_tokens": 29348877.0,
      "step": 13155
    },
    {
      "entropy": 5.356501150131225,
      "epoch": 1.4780142640534621,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047914734744424517,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20615222305059433,
      "num_tokens": 29360072.0,
      "step": 13160
    },
    {
      "entropy": 5.474372339248657,
      "epoch": 1.4785758409614196,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047913046417084393,
      "loss": 5.3284,
      "mean_token_accuracy": 0.19344312101602554,
      "num_tokens": 29370744.0,
      "step": 13165
    },
    {
      "entropy": 5.476388168334961,
      "epoch": 1.4791374178693772,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047911357439784947,
      "loss": 5.2807,
      "mean_token_accuracy": 0.18801308125257493,
      "num_tokens": 29381463.0,
      "step": 13170
    },
    {
      "entropy": 5.400535202026367,
      "epoch": 1.4796989947773347,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047909667812579964,
      "loss": 5.3041,
      "mean_token_accuracy": 0.19465888738632203,
      "num_tokens": 29394227.0,
      "step": 13175
    },
    {
      "entropy": 5.362101650238037,
      "epoch": 1.4802605716852923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047907977535523253,
      "loss": 5.2184,
      "mean_token_accuracy": 0.196198733150959,
      "num_tokens": 29404754.0,
      "step": 13180
    },
    {
      "entropy": 5.387766695022583,
      "epoch": 1.4808221485932498,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004790628660866861,
      "loss": 5.2462,
      "mean_token_accuracy": 0.20132745504379274,
      "num_tokens": 29416161.0,
      "step": 13185
    },
    {
      "entropy": 5.410443449020386,
      "epoch": 1.4813837255012074,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004790459503206989,
      "loss": 5.1448,
      "mean_token_accuracy": 0.20597066581249238,
      "num_tokens": 29426587.0,
      "step": 13190
    },
    {
      "entropy": 5.444832277297974,
      "epoch": 1.4819453024091649,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004790290280578095,
      "loss": 5.319,
      "mean_token_accuracy": 0.18633709698915482,
      "num_tokens": 29437488.0,
      "step": 13195
    },
    {
      "entropy": 5.507973957061767,
      "epoch": 1.4825068793171226,
      "grad_norm": 1.25,
      "learning_rate": 0.00047901209929855675,
      "loss": 5.2864,
      "mean_token_accuracy": 0.1919541209936142,
      "num_tokens": 29448244.0,
      "step": 13200
    },
    {
      "entropy": 5.527728509902954,
      "epoch": 1.48306845622508,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004789951640434796,
      "loss": 5.3611,
      "mean_token_accuracy": 0.1892804652452469,
      "num_tokens": 29458260.0,
      "step": 13205
    },
    {
      "entropy": 5.483075904846191,
      "epoch": 1.4836300331330377,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004789782222931173,
      "loss": 5.3082,
      "mean_token_accuracy": 0.18972558081150054,
      "num_tokens": 29470874.0,
      "step": 13210
    },
    {
      "entropy": 5.41730318069458,
      "epoch": 1.484191610040995,
      "grad_norm": 1.40625,
      "learning_rate": 0.00047896127404800935,
      "loss": 5.2242,
      "mean_token_accuracy": 0.1890203535556793,
      "num_tokens": 29481442.0,
      "step": 13215
    },
    {
      "entropy": 5.546670627593994,
      "epoch": 1.4847531869489528,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004789443193086953,
      "loss": 5.3825,
      "mean_token_accuracy": 0.1896733298897743,
      "num_tokens": 29492727.0,
      "step": 13220
    },
    {
      "entropy": 5.528610801696777,
      "epoch": 1.4853147638569102,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047892735807571505,
      "loss": 5.3051,
      "mean_token_accuracy": 0.1881225436925888,
      "num_tokens": 29502729.0,
      "step": 13225
    },
    {
      "entropy": 5.482454776763916,
      "epoch": 1.4858763407648676,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004789103903496086,
      "loss": 5.2506,
      "mean_token_accuracy": 0.1929288163781166,
      "num_tokens": 29514315.0,
      "step": 13230
    },
    {
      "entropy": 5.41616997718811,
      "epoch": 1.4864379176728253,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004788934161309161,
      "loss": 5.2126,
      "mean_token_accuracy": 0.19344741851091385,
      "num_tokens": 29526030.0,
      "step": 13235
    },
    {
      "entropy": 5.412803888320923,
      "epoch": 1.486999494580783,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047887643542017836,
      "loss": 5.2255,
      "mean_token_accuracy": 0.19977983087301254,
      "num_tokens": 29536972.0,
      "step": 13240
    },
    {
      "entropy": 5.356698608398437,
      "epoch": 1.4875610714887404,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047885944821793574,
      "loss": 5.2683,
      "mean_token_accuracy": 0.19482929557561873,
      "num_tokens": 29547325.0,
      "step": 13245
    },
    {
      "entropy": 5.461649608612061,
      "epoch": 1.4881226483966978,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047884245452472926,
      "loss": 5.2766,
      "mean_token_accuracy": 0.1961446925997734,
      "num_tokens": 29557944.0,
      "step": 13250
    },
    {
      "entropy": 5.363716077804566,
      "epoch": 1.4886842253046555,
      "grad_norm": 1.125,
      "learning_rate": 0.0004788254543410999,
      "loss": 5.2069,
      "mean_token_accuracy": 0.2030858501791954,
      "num_tokens": 29568618.0,
      "step": 13255
    },
    {
      "entropy": 5.444748592376709,
      "epoch": 1.489245802212613,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047880844766758914,
      "loss": 5.2767,
      "mean_token_accuracy": 0.18729570060968398,
      "num_tokens": 29580682.0,
      "step": 13260
    },
    {
      "entropy": 5.486226987838745,
      "epoch": 1.4898073791205706,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047879143450473824,
      "loss": 5.3242,
      "mean_token_accuracy": 0.1887418136000633,
      "num_tokens": 29592302.0,
      "step": 13265
    },
    {
      "entropy": 5.5025428295135494,
      "epoch": 1.490368956028528,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047877441485308917,
      "loss": 5.3596,
      "mean_token_accuracy": 0.18503705710172652,
      "num_tokens": 29603591.0,
      "step": 13270
    },
    {
      "entropy": 5.42810091972351,
      "epoch": 1.4909305329364857,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047875738871318356,
      "loss": 5.27,
      "mean_token_accuracy": 0.19263422787189483,
      "num_tokens": 29615542.0,
      "step": 13275
    },
    {
      "entropy": 5.5065319538116455,
      "epoch": 1.4914921098444431,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004787403560855637,
      "loss": 5.3209,
      "mean_token_accuracy": 0.19698410481214523,
      "num_tokens": 29626650.0,
      "step": 13280
    },
    {
      "entropy": 5.3547046184539795,
      "epoch": 1.4920536867524008,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004787233169707719,
      "loss": 5.2612,
      "mean_token_accuracy": 0.1899896815419197,
      "num_tokens": 29637135.0,
      "step": 13285
    },
    {
      "entropy": 5.425365734100342,
      "epoch": 1.4926152636603582,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004787062713693507,
      "loss": 5.301,
      "mean_token_accuracy": 0.1851486384868622,
      "num_tokens": 29648862.0,
      "step": 13290
    },
    {
      "entropy": 5.4676882266998295,
      "epoch": 1.493176840568316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047868921928184276,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19029929488897324,
      "num_tokens": 29661015.0,
      "step": 13295
    },
    {
      "entropy": 5.566496658325195,
      "epoch": 1.4937384174762733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004786721607087911,
      "loss": 5.2563,
      "mean_token_accuracy": 0.18760274797677995,
      "num_tokens": 29673521.0,
      "step": 13300
    },
    {
      "entropy": 5.502147912979126,
      "epoch": 1.494299994384231,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004786550956507389,
      "loss": 5.3318,
      "mean_token_accuracy": 0.19536332339048385,
      "num_tokens": 29685379.0,
      "step": 13305
    },
    {
      "entropy": 5.389399003982544,
      "epoch": 1.4948615712921884,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004786380241082294,
      "loss": 5.2431,
      "mean_token_accuracy": 0.1876935377717018,
      "num_tokens": 29696372.0,
      "step": 13310
    },
    {
      "entropy": 5.408631277084351,
      "epoch": 1.495423148200146,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047862094608180623,
      "loss": 5.2625,
      "mean_token_accuracy": 0.1975863054394722,
      "num_tokens": 29707401.0,
      "step": 13315
    },
    {
      "entropy": 5.377112245559692,
      "epoch": 1.4959847251081035,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004786038615720132,
      "loss": 5.2497,
      "mean_token_accuracy": 0.1919127434492111,
      "num_tokens": 29717524.0,
      "step": 13320
    },
    {
      "entropy": 5.464303970336914,
      "epoch": 1.496546302016061,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00047858677057939427,
      "loss": 5.3601,
      "mean_token_accuracy": 0.1918877437710762,
      "num_tokens": 29728235.0,
      "step": 13325
    },
    {
      "entropy": 5.555882692337036,
      "epoch": 1.4971078789240186,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00047856967310449356,
      "loss": 5.3577,
      "mean_token_accuracy": 0.18339023888111114,
      "num_tokens": 29738800.0,
      "step": 13330
    },
    {
      "entropy": 5.457555913925171,
      "epoch": 1.4976694558319763,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047855256914785556,
      "loss": 5.2837,
      "mean_token_accuracy": 0.18927301466464996,
      "num_tokens": 29750346.0,
      "step": 13335
    },
    {
      "entropy": 5.425393199920654,
      "epoch": 1.4982310327399337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047853545871002476,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19537228792905809,
      "num_tokens": 29761931.0,
      "step": 13340
    },
    {
      "entropy": 5.3307514667510985,
      "epoch": 1.4987926096478912,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047851834179154595,
      "loss": 5.1948,
      "mean_token_accuracy": 0.1965647429227829,
      "num_tokens": 29773333.0,
      "step": 13345
    },
    {
      "entropy": 5.382732295989991,
      "epoch": 1.4993541865558488,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004785012183929643,
      "loss": 5.1858,
      "mean_token_accuracy": 0.1968771979212761,
      "num_tokens": 29785125.0,
      "step": 13350
    },
    {
      "entropy": 5.384773302078247,
      "epoch": 1.4999157634638065,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004784840885148249,
      "loss": 5.298,
      "mean_token_accuracy": 0.19949504733085632,
      "num_tokens": 29796288.0,
      "step": 13355
    },
    {
      "entropy": 5.501663780212402,
      "epoch": 1.500477340371764,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004784669521576732,
      "loss": 5.3257,
      "mean_token_accuracy": 0.1917411595582962,
      "num_tokens": 29806348.0,
      "step": 13360
    },
    {
      "entropy": 5.484355688095093,
      "epoch": 1.5010389172797214,
      "grad_norm": 1.25,
      "learning_rate": 0.0004784498093220549,
      "loss": 5.2166,
      "mean_token_accuracy": 0.19674459248781204,
      "num_tokens": 29816342.0,
      "step": 13365
    },
    {
      "entropy": 5.354651546478271,
      "epoch": 1.501600494187679,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004784326600085157,
      "loss": 5.2064,
      "mean_token_accuracy": 0.18896224945783616,
      "num_tokens": 29826789.0,
      "step": 13370
    },
    {
      "entropy": 5.403900957107544,
      "epoch": 1.5021620710956367,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047841550421760164,
      "loss": 5.3176,
      "mean_token_accuracy": 0.1963667392730713,
      "num_tokens": 29837633.0,
      "step": 13375
    },
    {
      "entropy": 5.522979688644409,
      "epoch": 1.502723648003594,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047839834194985916,
      "loss": 5.3552,
      "mean_token_accuracy": 0.18796049952507018,
      "num_tokens": 29848498.0,
      "step": 13380
    },
    {
      "entropy": 5.566979885101318,
      "epoch": 1.5032852249115516,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004783811732058345,
      "loss": 5.3269,
      "mean_token_accuracy": 0.1891693204641342,
      "num_tokens": 29859777.0,
      "step": 13385
    },
    {
      "entropy": 5.706383609771729,
      "epoch": 1.5038468018195092,
      "grad_norm": 1.125,
      "learning_rate": 0.00047836399798607446,
      "loss": 5.5115,
      "mean_token_accuracy": 0.18352531641721725,
      "num_tokens": 29873373.0,
      "step": 13390
    },
    {
      "entropy": 5.438326644897461,
      "epoch": 1.5044083787274667,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047834681629112583,
      "loss": 5.2813,
      "mean_token_accuracy": 0.1947934851050377,
      "num_tokens": 29884749.0,
      "step": 13395
    },
    {
      "entropy": 5.414584493637085,
      "epoch": 1.5049699556354241,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004783296281215357,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19408933818340302,
      "num_tokens": 29895059.0,
      "step": 13400
    },
    {
      "entropy": 5.3966797351837155,
      "epoch": 1.5055315325433818,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004783124334778513,
      "loss": 5.2502,
      "mean_token_accuracy": 0.1891699329018593,
      "num_tokens": 29906611.0,
      "step": 13405
    },
    {
      "entropy": 5.32564058303833,
      "epoch": 1.5060931094513395,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004782952323606202,
      "loss": 5.1246,
      "mean_token_accuracy": 0.20259118676185608,
      "num_tokens": 29918179.0,
      "step": 13410
    },
    {
      "entropy": 5.484653520584106,
      "epoch": 1.506654686359297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004782780247703901,
      "loss": 5.3123,
      "mean_token_accuracy": 0.18911022692918777,
      "num_tokens": 29929612.0,
      "step": 13415
    },
    {
      "entropy": 5.448102188110352,
      "epoch": 1.5072162632672543,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004782608107077088,
      "loss": 5.1656,
      "mean_token_accuracy": 0.1987714111804962,
      "num_tokens": 29939569.0,
      "step": 13420
    },
    {
      "entropy": 5.4011975765228275,
      "epoch": 1.507777840175212,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047824359017312445,
      "loss": 5.2154,
      "mean_token_accuracy": 0.19681653380393982,
      "num_tokens": 29950344.0,
      "step": 13425
    },
    {
      "entropy": 5.4108006954193115,
      "epoch": 1.5083394170831697,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004782263631671854,
      "loss": 5.219,
      "mean_token_accuracy": 0.19945429414510726,
      "num_tokens": 29960748.0,
      "step": 13430
    },
    {
      "entropy": 5.545145893096924,
      "epoch": 1.508900993991127,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047820912969044014,
      "loss": 5.4047,
      "mean_token_accuracy": 0.18214541524648667,
      "num_tokens": 29971678.0,
      "step": 13435
    },
    {
      "entropy": 5.390454912185669,
      "epoch": 1.5094625708990845,
      "grad_norm": 1.125,
      "learning_rate": 0.00047819188974343734,
      "loss": 5.2644,
      "mean_token_accuracy": 0.19253990203142166,
      "num_tokens": 29982477.0,
      "step": 13440
    },
    {
      "entropy": 5.409023475646973,
      "epoch": 1.5100241478070422,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004781746433267259,
      "loss": 5.2947,
      "mean_token_accuracy": 0.1977451726794243,
      "num_tokens": 29994539.0,
      "step": 13445
    },
    {
      "entropy": 5.511298370361328,
      "epoch": 1.5105857247149999,
      "grad_norm": 1.328125,
      "learning_rate": 0.00047815739044085515,
      "loss": 5.2247,
      "mean_token_accuracy": 0.19928698241710663,
      "num_tokens": 30004821.0,
      "step": 13450
    },
    {
      "entropy": 5.5251940250396725,
      "epoch": 1.5111473016229573,
      "grad_norm": 1.125,
      "learning_rate": 0.00047814013108637423,
      "loss": 5.3028,
      "mean_token_accuracy": 0.18489723652601242,
      "num_tokens": 30016343.0,
      "step": 13455
    },
    {
      "entropy": 5.448993301391601,
      "epoch": 1.5117088785309147,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047812286526383274,
      "loss": 5.2008,
      "mean_token_accuracy": 0.2012137457728386,
      "num_tokens": 30027719.0,
      "step": 13460
    },
    {
      "entropy": 5.391881132125855,
      "epoch": 1.5122704554388724,
      "grad_norm": 1.125,
      "learning_rate": 0.0004781055929737804,
      "loss": 5.2196,
      "mean_token_accuracy": 0.20195548981428146,
      "num_tokens": 30037988.0,
      "step": 13465
    },
    {
      "entropy": 5.405081367492675,
      "epoch": 1.51283203234683,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047808831421676724,
      "loss": 5.2091,
      "mean_token_accuracy": 0.19408999681472777,
      "num_tokens": 30049736.0,
      "step": 13470
    },
    {
      "entropy": 5.427035188674926,
      "epoch": 1.5133936092547875,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004780710289933435,
      "loss": 5.3184,
      "mean_token_accuracy": 0.1925765722990036,
      "num_tokens": 30061963.0,
      "step": 13475
    },
    {
      "entropy": 5.476969528198242,
      "epoch": 1.513955186162745,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004780537373040593,
      "loss": 5.2613,
      "mean_token_accuracy": 0.1989409565925598,
      "num_tokens": 30072256.0,
      "step": 13480
    },
    {
      "entropy": 5.4121985912323,
      "epoch": 1.5145167630707026,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047803643914946537,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19489428251981736,
      "num_tokens": 30084804.0,
      "step": 13485
    },
    {
      "entropy": 5.460195016860962,
      "epoch": 1.51507833997866,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004780191345301125,
      "loss": 5.2763,
      "mean_token_accuracy": 0.19354714155197145,
      "num_tokens": 30096938.0,
      "step": 13490
    },
    {
      "entropy": 5.501681995391846,
      "epoch": 1.5156399168866175,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047800182344655165,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1889519840478897,
      "num_tokens": 30108233.0,
      "step": 13495
    },
    {
      "entropy": 5.444441652297973,
      "epoch": 1.5162014937945751,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000477984505899334,
      "loss": 5.2688,
      "mean_token_accuracy": 0.19504794627428054,
      "num_tokens": 30120246.0,
      "step": 13500
    },
    {
      "entropy": 5.409551906585693,
      "epoch": 1.5167630707025328,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047796718188901095,
      "loss": 5.0955,
      "mean_token_accuracy": 0.20016416609287263,
      "num_tokens": 30131213.0,
      "step": 13505
    },
    {
      "entropy": 5.302604103088379,
      "epoch": 1.5173246476104902,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047794985141613413,
      "loss": 5.1484,
      "mean_token_accuracy": 0.19683748036623,
      "num_tokens": 30143300.0,
      "step": 13510
    },
    {
      "entropy": 5.3935182094573975,
      "epoch": 1.5178862245184477,
      "grad_norm": 1.125,
      "learning_rate": 0.00047793251448125526,
      "loss": 5.2198,
      "mean_token_accuracy": 0.18978463262319564,
      "num_tokens": 30153794.0,
      "step": 13515
    },
    {
      "entropy": 5.426440858840943,
      "epoch": 1.5184478014264053,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004779151710849265,
      "loss": 5.3363,
      "mean_token_accuracy": 0.18739715069532395,
      "num_tokens": 30165152.0,
      "step": 13520
    },
    {
      "entropy": 5.4154908657073975,
      "epoch": 1.519009378334363,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004778978212276999,
      "loss": 5.2636,
      "mean_token_accuracy": 0.19496385157108306,
      "num_tokens": 30176338.0,
      "step": 13525
    },
    {
      "entropy": 5.418806791305542,
      "epoch": 1.5195709552423204,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004778804649101281,
      "loss": 5.2988,
      "mean_token_accuracy": 0.19032107889652253,
      "num_tokens": 30187666.0,
      "step": 13530
    },
    {
      "entropy": 5.455024909973145,
      "epoch": 1.5201325321502779,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004778631021327636,
      "loss": 5.321,
      "mean_token_accuracy": 0.18851346969604493,
      "num_tokens": 30198892.0,
      "step": 13535
    },
    {
      "entropy": 5.412677145004272,
      "epoch": 1.5206941090582355,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004778457328961592,
      "loss": 5.2762,
      "mean_token_accuracy": 0.18981684893369674,
      "num_tokens": 30210677.0,
      "step": 13540
    },
    {
      "entropy": 5.451454401016235,
      "epoch": 1.5212556859661932,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047782835720086797,
      "loss": 5.233,
      "mean_token_accuracy": 0.19500928074121476,
      "num_tokens": 30221217.0,
      "step": 13545
    },
    {
      "entropy": 5.440000820159912,
      "epoch": 1.5218172628741506,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004778109750474432,
      "loss": 5.3382,
      "mean_token_accuracy": 0.1924414172768593,
      "num_tokens": 30233576.0,
      "step": 13550
    },
    {
      "entropy": 5.376503372192383,
      "epoch": 1.522378839782108,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004777935864364383,
      "loss": 5.1913,
      "mean_token_accuracy": 0.19454078376293182,
      "num_tokens": 30244787.0,
      "step": 13555
    },
    {
      "entropy": 5.401934289932251,
      "epoch": 1.5229404166900657,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477776191368407,
      "loss": 5.2669,
      "mean_token_accuracy": 0.19777343273162842,
      "num_tokens": 30256656.0,
      "step": 13560
    },
    {
      "entropy": 5.46606969833374,
      "epoch": 1.5235019935980234,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004777587898439031,
      "loss": 5.3488,
      "mean_token_accuracy": 0.19227325767278672,
      "num_tokens": 30267715.0,
      "step": 13565
    },
    {
      "entropy": 5.5290178775787355,
      "epoch": 1.5240635705059808,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047774138186348073,
      "loss": 5.3216,
      "mean_token_accuracy": 0.18840609043836593,
      "num_tokens": 30278119.0,
      "step": 13570
    },
    {
      "entropy": 5.485423517227173,
      "epoch": 1.5246251474139383,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004777239674276941,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19443580955266954,
      "num_tokens": 30289899.0,
      "step": 13575
    },
    {
      "entropy": 5.383790445327759,
      "epoch": 1.525186724321896,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004777065465370977,
      "loss": 5.2369,
      "mean_token_accuracy": 0.19598551094532013,
      "num_tokens": 30300799.0,
      "step": 13580
    },
    {
      "entropy": 5.596652698516846,
      "epoch": 1.5257483012298534,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004776891191922463,
      "loss": 5.4515,
      "mean_token_accuracy": 0.1856719434261322,
      "num_tokens": 30312955.0,
      "step": 13585
    },
    {
      "entropy": 5.469523000717163,
      "epoch": 1.5263098781378108,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004776716853936947,
      "loss": 5.2266,
      "mean_token_accuracy": 0.1933372974395752,
      "num_tokens": 30323089.0,
      "step": 13590
    },
    {
      "entropy": 5.370826244354248,
      "epoch": 1.5268714550457685,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000477654245141998,
      "loss": 5.1842,
      "mean_token_accuracy": 0.19963466972112656,
      "num_tokens": 30333926.0,
      "step": 13595
    },
    {
      "entropy": 5.391605234146118,
      "epoch": 1.5274330319537262,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004776367984377116,
      "loss": 5.285,
      "mean_token_accuracy": 0.1901363894343376,
      "num_tokens": 30346347.0,
      "step": 13600
    },
    {
      "entropy": 5.45722336769104,
      "epoch": 1.5279946088616836,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047761934528139086,
      "loss": 5.2574,
      "mean_token_accuracy": 0.18455291539430618,
      "num_tokens": 30357757.0,
      "step": 13605
    },
    {
      "entropy": 5.545757484436035,
      "epoch": 1.528556185769641,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004776018856735916,
      "loss": 5.342,
      "mean_token_accuracy": 0.19106468111276625,
      "num_tokens": 30369323.0,
      "step": 13610
    },
    {
      "entropy": 5.354086685180664,
      "epoch": 1.5291177626775987,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004775844196148698,
      "loss": 5.1825,
      "mean_token_accuracy": 0.19854395389556884,
      "num_tokens": 30380335.0,
      "step": 13615
    },
    {
      "entropy": 5.361436033248902,
      "epoch": 1.5296793395855564,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047756694710578146,
      "loss": 5.2,
      "mean_token_accuracy": 0.19790864884853362,
      "num_tokens": 30391338.0,
      "step": 13620
    },
    {
      "entropy": 5.459527444839478,
      "epoch": 1.5302409164935138,
      "grad_norm": 1.015625,
      "learning_rate": 0.000477549468146883,
      "loss": 5.3037,
      "mean_token_accuracy": 0.18293204456567763,
      "num_tokens": 30402917.0,
      "step": 13625
    },
    {
      "entropy": 5.372813510894775,
      "epoch": 1.5308024934014712,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004775319827387309,
      "loss": 5.1155,
      "mean_token_accuracy": 0.2021074801683426,
      "num_tokens": 30412690.0,
      "step": 13630
    },
    {
      "entropy": 5.397270727157593,
      "epoch": 1.531364070309429,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004775144908818818,
      "loss": 5.2826,
      "mean_token_accuracy": 0.19037747383117676,
      "num_tokens": 30423020.0,
      "step": 13635
    },
    {
      "entropy": 5.559815883636475,
      "epoch": 1.5319256472173866,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047749699257689296,
      "loss": 5.5072,
      "mean_token_accuracy": 0.17775093913078308,
      "num_tokens": 30434039.0,
      "step": 13640
    },
    {
      "entropy": 5.543820905685425,
      "epoch": 1.532487224125344,
      "grad_norm": 1.125,
      "learning_rate": 0.0004774794878243212,
      "loss": 5.2814,
      "mean_token_accuracy": 0.18784320801496507,
      "num_tokens": 30446743.0,
      "step": 13645
    },
    {
      "entropy": 5.471198225021363,
      "epoch": 1.5330488010333014,
      "grad_norm": 1.1875,
      "learning_rate": 0.000477461976624724,
      "loss": 5.2483,
      "mean_token_accuracy": 0.19485463947057724,
      "num_tokens": 30456969.0,
      "step": 13650
    },
    {
      "entropy": 5.366778898239136,
      "epoch": 1.533610377941259,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047744445897865905,
      "loss": 5.306,
      "mean_token_accuracy": 0.18710312843322754,
      "num_tokens": 30467497.0,
      "step": 13655
    },
    {
      "entropy": 5.454708242416382,
      "epoch": 1.5341719548492168,
      "grad_norm": 1.109375,
      "learning_rate": 0.000477426934886684,
      "loss": 5.2719,
      "mean_token_accuracy": 0.19015054404735565,
      "num_tokens": 30479428.0,
      "step": 13660
    },
    {
      "entropy": 5.434100008010864,
      "epoch": 1.5347335317571742,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047740940434935677,
      "loss": 5.2259,
      "mean_token_accuracy": 0.19724835455417633,
      "num_tokens": 30490193.0,
      "step": 13665
    },
    {
      "entropy": 5.457854223251343,
      "epoch": 1.5352951086651316,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004773918673672356,
      "loss": 5.2854,
      "mean_token_accuracy": 0.1905955880880356,
      "num_tokens": 30500311.0,
      "step": 13670
    },
    {
      "entropy": 5.392323303222656,
      "epoch": 1.5358566855730893,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047737432394087887,
      "loss": 5.2014,
      "mean_token_accuracy": 0.19985711872577666,
      "num_tokens": 30512485.0,
      "step": 13675
    },
    {
      "entropy": 5.426626682281494,
      "epoch": 1.5364182624810467,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047735677407084525,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19498414695262908,
      "num_tokens": 30523495.0,
      "step": 13680
    },
    {
      "entropy": 5.3962806224822994,
      "epoch": 1.5369798393890042,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004773392177576934,
      "loss": 5.2452,
      "mean_token_accuracy": 0.19643473774194717,
      "num_tokens": 30533964.0,
      "step": 13685
    },
    {
      "entropy": 5.313876676559448,
      "epoch": 1.5375414162969618,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004773216550019824,
      "loss": 5.1384,
      "mean_token_accuracy": 0.19768504202365875,
      "num_tokens": 30544798.0,
      "step": 13690
    },
    {
      "entropy": 5.4216637134552,
      "epoch": 1.5381029932049195,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047730408580427136,
      "loss": 5.334,
      "mean_token_accuracy": 0.19050325453281403,
      "num_tokens": 30555991.0,
      "step": 13695
    },
    {
      "entropy": 5.469227504730225,
      "epoch": 1.538664570112877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047728651016511994,
      "loss": 5.1638,
      "mean_token_accuracy": 0.19549207389354706,
      "num_tokens": 30566848.0,
      "step": 13700
    },
    {
      "entropy": 5.40775465965271,
      "epoch": 1.5392261470208344,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047726892808508744,
      "loss": 5.1768,
      "mean_token_accuracy": 0.20165636241436005,
      "num_tokens": 30577361.0,
      "step": 13705
    },
    {
      "entropy": 5.336817836761474,
      "epoch": 1.539787723928792,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004772513395647339,
      "loss": 5.2432,
      "mean_token_accuracy": 0.19521275758743287,
      "num_tokens": 30587685.0,
      "step": 13710
    },
    {
      "entropy": 5.408463764190674,
      "epoch": 1.5403493008367497,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004772337446046192,
      "loss": 5.2946,
      "mean_token_accuracy": 0.19399577975273133,
      "num_tokens": 30599864.0,
      "step": 13715
    },
    {
      "entropy": 5.448783445358276,
      "epoch": 1.5409108777447071,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047721614320530366,
      "loss": 5.2845,
      "mean_token_accuracy": 0.19121176302433013,
      "num_tokens": 30610885.0,
      "step": 13720
    },
    {
      "entropy": 5.426961183547974,
      "epoch": 1.5414724546526646,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047719853536734764,
      "loss": 5.2338,
      "mean_token_accuracy": 0.1926610976457596,
      "num_tokens": 30622537.0,
      "step": 13725
    },
    {
      "entropy": 5.499411392211914,
      "epoch": 1.5420340315606222,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004771809210913119,
      "loss": 5.371,
      "mean_token_accuracy": 0.18665376156568528,
      "num_tokens": 30635045.0,
      "step": 13730
    },
    {
      "entropy": 5.394086933135986,
      "epoch": 1.54259560846858,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047716330037775724,
      "loss": 5.1885,
      "mean_token_accuracy": 0.19156184047460556,
      "num_tokens": 30645998.0,
      "step": 13735
    },
    {
      "entropy": 5.399102640151978,
      "epoch": 1.5431571853765373,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004771456732272446,
      "loss": 5.2562,
      "mean_token_accuracy": 0.18922141194343567,
      "num_tokens": 30658744.0,
      "step": 13740
    },
    {
      "entropy": 5.360708093643188,
      "epoch": 1.5437187622844948,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047712803964033545,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20015766322612763,
      "num_tokens": 30669420.0,
      "step": 13745
    },
    {
      "entropy": 5.278477048873901,
      "epoch": 1.5442803391924524,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047711039961759107,
      "loss": 5.2144,
      "mean_token_accuracy": 0.19358174353837967,
      "num_tokens": 30680037.0,
      "step": 13750
    },
    {
      "entropy": 5.491505527496338,
      "epoch": 1.54484191610041,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004770927531595731,
      "loss": 5.2482,
      "mean_token_accuracy": 0.1999806970357895,
      "num_tokens": 30691650.0,
      "step": 13755
    },
    {
      "entropy": 5.35825924873352,
      "epoch": 1.5454034930083675,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047707510026684353,
      "loss": 5.2918,
      "mean_token_accuracy": 0.1918620139360428,
      "num_tokens": 30701876.0,
      "step": 13760
    },
    {
      "entropy": 5.611922311782837,
      "epoch": 1.545965069916325,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047705744093996447,
      "loss": 5.4793,
      "mean_token_accuracy": 0.18509581536054612,
      "num_tokens": 30714484.0,
      "step": 13765
    },
    {
      "entropy": 5.400297355651856,
      "epoch": 1.5465266468242826,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000477039775179498,
      "loss": 5.266,
      "mean_token_accuracy": 0.18623226583004,
      "num_tokens": 30725185.0,
      "step": 13770
    },
    {
      "entropy": 5.326832056045532,
      "epoch": 1.54708822373224,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047702210298600684,
      "loss": 5.2579,
      "mean_token_accuracy": 0.19108876287937165,
      "num_tokens": 30735522.0,
      "step": 13775
    },
    {
      "entropy": 5.382134294509887,
      "epoch": 1.5476498006401975,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047700442436005356,
      "loss": 5.2583,
      "mean_token_accuracy": 0.196957728266716,
      "num_tokens": 30746353.0,
      "step": 13780
    },
    {
      "entropy": 5.4740814685821535,
      "epoch": 1.5482113775481552,
      "grad_norm": 1.125,
      "learning_rate": 0.000476986739302201,
      "loss": 5.2146,
      "mean_token_accuracy": 0.19409114867448807,
      "num_tokens": 30756245.0,
      "step": 13785
    },
    {
      "entropy": 5.464963674545288,
      "epoch": 1.5487729544561128,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047696904781301233,
      "loss": 5.331,
      "mean_token_accuracy": 0.1953594923019409,
      "num_tokens": 30767387.0,
      "step": 13790
    },
    {
      "entropy": 5.393512344360351,
      "epoch": 1.5493345313640703,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004769513498930508,
      "loss": 5.2901,
      "mean_token_accuracy": 0.19191524386405945,
      "num_tokens": 30780140.0,
      "step": 13795
    },
    {
      "entropy": 5.414828538894653,
      "epoch": 1.5498961082720277,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047693364554288,
      "loss": 5.1942,
      "mean_token_accuracy": 0.1981704980134964,
      "num_tokens": 30791137.0,
      "step": 13800
    },
    {
      "entropy": 5.404635858535767,
      "epoch": 1.5504576851799854,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047691593476306373,
      "loss": 5.2174,
      "mean_token_accuracy": 0.19825750589370728,
      "num_tokens": 30801708.0,
      "step": 13805
    },
    {
      "entropy": 5.4322082042694095,
      "epoch": 1.551019262087943,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004768982175541656,
      "loss": 5.281,
      "mean_token_accuracy": 0.19138255715370178,
      "num_tokens": 30813519.0,
      "step": 13810
    },
    {
      "entropy": 5.5216950416564945,
      "epoch": 1.5515808389959005,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047688049391675006,
      "loss": 5.3503,
      "mean_token_accuracy": 0.19128201007843018,
      "num_tokens": 30823786.0,
      "step": 13815
    },
    {
      "entropy": 5.397929763793945,
      "epoch": 1.552142415903858,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004768627638513813,
      "loss": 5.2339,
      "mean_token_accuracy": 0.19601008892059327,
      "num_tokens": 30834381.0,
      "step": 13820
    },
    {
      "entropy": 5.430441236495971,
      "epoch": 1.5527039928118156,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004768450273586237,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1950397312641144,
      "num_tokens": 30844865.0,
      "step": 13825
    },
    {
      "entropy": 5.456672143936157,
      "epoch": 1.5532655697197733,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047682728443904223,
      "loss": 5.3111,
      "mean_token_accuracy": 0.19478338062763215,
      "num_tokens": 30857088.0,
      "step": 13830
    },
    {
      "entropy": 5.396194505691528,
      "epoch": 1.5538271466277307,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047680953509320176,
      "loss": 5.3454,
      "mean_token_accuracy": 0.18512041866779327,
      "num_tokens": 30868320.0,
      "step": 13835
    },
    {
      "entropy": 5.4606321334838865,
      "epoch": 1.5543887235356881,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004767917793216674,
      "loss": 5.2482,
      "mean_token_accuracy": 0.19642228931188582,
      "num_tokens": 30878598.0,
      "step": 13840
    },
    {
      "entropy": 5.425065660476685,
      "epoch": 1.5549503004436458,
      "grad_norm": 1.296875,
      "learning_rate": 0.00047677401712500455,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19661712795495986,
      "num_tokens": 30889254.0,
      "step": 13845
    },
    {
      "entropy": 5.459064245223999,
      "epoch": 1.5555118773516035,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004767562485037787,
      "loss": 5.3015,
      "mean_token_accuracy": 0.19374869912862777,
      "num_tokens": 30901164.0,
      "step": 13850
    },
    {
      "entropy": 5.392041349411011,
      "epoch": 1.556073454259561,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047673847345855566,
      "loss": 5.2208,
      "mean_token_accuracy": 0.19375364780426024,
      "num_tokens": 30911828.0,
      "step": 13855
    },
    {
      "entropy": 5.392468214035034,
      "epoch": 1.5566350311675183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004767206919899014,
      "loss": 5.0962,
      "mean_token_accuracy": 0.2029048979282379,
      "num_tokens": 30922050.0,
      "step": 13860
    },
    {
      "entropy": 5.338767862319946,
      "epoch": 1.557196608075476,
      "grad_norm": 1.15625,
      "learning_rate": 0.000476702904098382,
      "loss": 5.1729,
      "mean_token_accuracy": 0.19713774770498277,
      "num_tokens": 30932723.0,
      "step": 13865
    },
    {
      "entropy": 5.361827182769775,
      "epoch": 1.5577581849834334,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000476685109784564,
      "loss": 5.2018,
      "mean_token_accuracy": 0.19916203022003173,
      "num_tokens": 30943699.0,
      "step": 13870
    },
    {
      "entropy": 5.432551336288452,
      "epoch": 1.5583197618913909,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047666730904901377,
      "loss": 5.2657,
      "mean_token_accuracy": 0.19288444817066192,
      "num_tokens": 30954315.0,
      "step": 13875
    },
    {
      "entropy": 5.530124235153198,
      "epoch": 1.5588813387993485,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047664950189229833,
      "loss": 5.4335,
      "mean_token_accuracy": 0.18677089214324952,
      "num_tokens": 30965840.0,
      "step": 13880
    },
    {
      "entropy": 5.521332693099976,
      "epoch": 1.5594429157073062,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004766316883149845,
      "loss": 5.3262,
      "mean_token_accuracy": 0.1931717574596405,
      "num_tokens": 30977051.0,
      "step": 13885
    },
    {
      "entropy": 5.309330415725708,
      "epoch": 1.5600044926152636,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047661386831763955,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19732474237680436,
      "num_tokens": 30988730.0,
      "step": 13890
    },
    {
      "entropy": 5.35407280921936,
      "epoch": 1.560566069523221,
      "grad_norm": 1.03125,
      "learning_rate": 0.00047659604190083076,
      "loss": 5.1921,
      "mean_token_accuracy": 0.19518263638019562,
      "num_tokens": 31000571.0,
      "step": 13895
    },
    {
      "entropy": 5.432471370697021,
      "epoch": 1.5611276464311787,
      "grad_norm": 1.125,
      "learning_rate": 0.00047657820906512585,
      "loss": 5.2139,
      "mean_token_accuracy": 0.20153809636831282,
      "num_tokens": 31011003.0,
      "step": 13900
    },
    {
      "entropy": 5.423380184173584,
      "epoch": 1.5616892233391364,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047656036981109256,
      "loss": 5.2198,
      "mean_token_accuracy": 0.1994598925113678,
      "num_tokens": 31022510.0,
      "step": 13905
    },
    {
      "entropy": 5.436857652664185,
      "epoch": 1.5622508002470938,
      "grad_norm": 1.125,
      "learning_rate": 0.00047654252413929896,
      "loss": 5.2844,
      "mean_token_accuracy": 0.19004056751728057,
      "num_tokens": 31033149.0,
      "step": 13910
    },
    {
      "entropy": 5.510715198516846,
      "epoch": 1.5628123771550513,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047652467205031317,
      "loss": 5.314,
      "mean_token_accuracy": 0.19826253354549409,
      "num_tokens": 31043626.0,
      "step": 13915
    },
    {
      "entropy": 5.493800020217895,
      "epoch": 1.563373954063009,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047650681354470363,
      "loss": 5.3532,
      "mean_token_accuracy": 0.19249965697526933,
      "num_tokens": 31055083.0,
      "step": 13920
    },
    {
      "entropy": 5.401673984527588,
      "epoch": 1.5639355309709666,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047648894862303907,
      "loss": 5.206,
      "mean_token_accuracy": 0.18956436812877656,
      "num_tokens": 31066763.0,
      "step": 13925
    },
    {
      "entropy": 5.417176008224487,
      "epoch": 1.564497107878924,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047647107728588826,
      "loss": 5.2673,
      "mean_token_accuracy": 0.19028756767511368,
      "num_tokens": 31078722.0,
      "step": 13930
    },
    {
      "entropy": 5.458354520797729,
      "epoch": 1.5650586847868815,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047645319953382024,
      "loss": 5.3032,
      "mean_token_accuracy": 0.19648596346378328,
      "num_tokens": 31090207.0,
      "step": 13935
    },
    {
      "entropy": 5.447060394287109,
      "epoch": 1.5656202616948391,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047643531536740405,
      "loss": 5.3296,
      "mean_token_accuracy": 0.1925012946128845,
      "num_tokens": 31101943.0,
      "step": 13940
    },
    {
      "entropy": 5.448119497299194,
      "epoch": 1.5661818386027968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004764174247872094,
      "loss": 5.3391,
      "mean_token_accuracy": 0.18826257288455964,
      "num_tokens": 31115032.0,
      "step": 13945
    },
    {
      "entropy": 5.514790868759155,
      "epoch": 1.5667434155107542,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047639952779380586,
      "loss": 5.2384,
      "mean_token_accuracy": 0.19772580415010452,
      "num_tokens": 31124965.0,
      "step": 13950
    },
    {
      "entropy": 5.481111717224121,
      "epoch": 1.5673049924187117,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004763816243877632,
      "loss": 5.2646,
      "mean_token_accuracy": 0.19057824611663818,
      "num_tokens": 31136692.0,
      "step": 13955
    },
    {
      "entropy": 5.420503568649292,
      "epoch": 1.5678665693266693,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004763637145696515,
      "loss": 5.2473,
      "mean_token_accuracy": 0.19159969538450242,
      "num_tokens": 31148352.0,
      "step": 13960
    },
    {
      "entropy": 5.37258677482605,
      "epoch": 1.5684281462346268,
      "grad_norm": 1.109375,
      "learning_rate": 0.000476345798340041,
      "loss": 5.2049,
      "mean_token_accuracy": 0.19525333046913146,
      "num_tokens": 31159892.0,
      "step": 13965
    },
    {
      "entropy": 5.527163314819336,
      "epoch": 1.5689897231425842,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004763278756995022,
      "loss": 5.3525,
      "mean_token_accuracy": 0.18845344930887223,
      "num_tokens": 31170087.0,
      "step": 13970
    },
    {
      "entropy": 5.3343404769897464,
      "epoch": 1.5695513000505419,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047630994664860576,
      "loss": 5.156,
      "mean_token_accuracy": 0.19562680870294571,
      "num_tokens": 31180687.0,
      "step": 13975
    },
    {
      "entropy": 5.4815186023712155,
      "epoch": 1.5701128769584995,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004762920111879226,
      "loss": 5.3235,
      "mean_token_accuracy": 0.19009089469909668,
      "num_tokens": 31192513.0,
      "step": 13980
    },
    {
      "entropy": 5.423823070526123,
      "epoch": 1.570674453866457,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004762740693180237,
      "loss": 5.2241,
      "mean_token_accuracy": 0.19401683211326598,
      "num_tokens": 31203432.0,
      "step": 13985
    },
    {
      "entropy": 5.461122465133667,
      "epoch": 1.5712360307744144,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047625612103948035,
      "loss": 5.2224,
      "mean_token_accuracy": 0.20152114778757096,
      "num_tokens": 31214474.0,
      "step": 13990
    },
    {
      "entropy": 5.4504319190979,
      "epoch": 1.571797607682372,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004762381663528641,
      "loss": 5.2833,
      "mean_token_accuracy": 0.18873070180416107,
      "num_tokens": 31226489.0,
      "step": 13995
    },
    {
      "entropy": 5.338401222229004,
      "epoch": 1.5723591845903298,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004762202052587466,
      "loss": 5.1606,
      "mean_token_accuracy": 0.19449353814125062,
      "num_tokens": 31237568.0,
      "step": 14000
    },
    {
      "entropy": 5.3929856300354,
      "epoch": 1.5729207614982872,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004762022377576996,
      "loss": 5.1603,
      "mean_token_accuracy": 0.2085471734404564,
      "num_tokens": 31248587.0,
      "step": 14005
    },
    {
      "entropy": 5.434262800216675,
      "epoch": 1.5734823384062446,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004761842638502954,
      "loss": 5.2939,
      "mean_token_accuracy": 0.19370619356632232,
      "num_tokens": 31259513.0,
      "step": 14010
    },
    {
      "entropy": 5.467734432220459,
      "epoch": 1.5740439153142023,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004761662835371062,
      "loss": 5.2881,
      "mean_token_accuracy": 0.1904546782374382,
      "num_tokens": 31270984.0,
      "step": 14015
    },
    {
      "entropy": 5.429001522064209,
      "epoch": 1.57460549222216,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004761482968187045,
      "loss": 5.2299,
      "mean_token_accuracy": 0.19307435303926468,
      "num_tokens": 31281907.0,
      "step": 14020
    },
    {
      "entropy": 5.336591100692749,
      "epoch": 1.5751670691301174,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004761303036956631,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19489356726408005,
      "num_tokens": 31292903.0,
      "step": 14025
    },
    {
      "entropy": 5.49773325920105,
      "epoch": 1.5757286460380748,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004761123041685547,
      "loss": 5.276,
      "mean_token_accuracy": 0.19539545625448226,
      "num_tokens": 31304185.0,
      "step": 14030
    },
    {
      "entropy": 5.446257448196411,
      "epoch": 1.5762902229460325,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004760942982379526,
      "loss": 5.2761,
      "mean_token_accuracy": 0.19103594720363617,
      "num_tokens": 31314806.0,
      "step": 14035
    },
    {
      "entropy": 5.442435693740845,
      "epoch": 1.5768517998539902,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047607628590443004,
      "loss": 5.2806,
      "mean_token_accuracy": 0.19540289044380188,
      "num_tokens": 31326361.0,
      "step": 14040
    },
    {
      "entropy": 5.5094770908355715,
      "epoch": 1.5774133767619476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047605826716856056,
      "loss": 5.2634,
      "mean_token_accuracy": 0.19840868562459946,
      "num_tokens": 31336549.0,
      "step": 14045
    },
    {
      "entropy": 5.515659046173096,
      "epoch": 1.577974953669905,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004760402420309178,
      "loss": 5.3669,
      "mean_token_accuracy": 0.18913905769586564,
      "num_tokens": 31346994.0,
      "step": 14050
    },
    {
      "entropy": 5.417604112625122,
      "epoch": 1.5785365305778627,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004760222104920759,
      "loss": 5.265,
      "mean_token_accuracy": 0.1914373442530632,
      "num_tokens": 31358449.0,
      "step": 14055
    },
    {
      "entropy": 5.479715585708618,
      "epoch": 1.5790981074858201,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004760041725526088,
      "loss": 5.3139,
      "mean_token_accuracy": 0.190072163939476,
      "num_tokens": 31368964.0,
      "step": 14060
    },
    {
      "entropy": 5.4369758605957035,
      "epoch": 1.5796596843937776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047598612821309077,
      "loss": 5.2478,
      "mean_token_accuracy": 0.19750716984272004,
      "num_tokens": 31379388.0,
      "step": 14065
    },
    {
      "entropy": 5.389810609817505,
      "epoch": 1.5802212613017352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004759680774740966,
      "loss": 5.2802,
      "mean_token_accuracy": 0.19735356718301772,
      "num_tokens": 31390757.0,
      "step": 14070
    },
    {
      "entropy": 5.39409065246582,
      "epoch": 1.580782838209693,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004759500203362008,
      "loss": 5.2388,
      "mean_token_accuracy": 0.19244916439056398,
      "num_tokens": 31402309.0,
      "step": 14075
    },
    {
      "entropy": 5.49221978187561,
      "epoch": 1.5813444151176503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004759319567999785,
      "loss": 5.2472,
      "mean_token_accuracy": 0.19421395361423494,
      "num_tokens": 31412559.0,
      "step": 14080
    },
    {
      "entropy": 5.460115957260132,
      "epoch": 1.5819059920256078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004759138868660048,
      "loss": 5.3995,
      "mean_token_accuracy": 0.18740309476852418,
      "num_tokens": 31424982.0,
      "step": 14085
    },
    {
      "entropy": 5.393653297424317,
      "epoch": 1.5824675689335654,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047589581053485494,
      "loss": 5.0997,
      "mean_token_accuracy": 0.19733640551567078,
      "num_tokens": 31434472.0,
      "step": 14090
    },
    {
      "entropy": 5.393371963500977,
      "epoch": 1.583029145841523,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047587772780710455,
      "loss": 5.3348,
      "mean_token_accuracy": 0.18950925022363663,
      "num_tokens": 31445458.0,
      "step": 14095
    },
    {
      "entropy": 5.449988937377929,
      "epoch": 1.5835907227494805,
      "grad_norm": 1.125,
      "learning_rate": 0.00047585963868332945,
      "loss": 5.228,
      "mean_token_accuracy": 0.19983907639980317,
      "num_tokens": 31455851.0,
      "step": 14100
    },
    {
      "entropy": 5.364946174621582,
      "epoch": 1.584152299657438,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004758415431641055,
      "loss": 5.1808,
      "mean_token_accuracy": 0.1970400482416153,
      "num_tokens": 31466279.0,
      "step": 14105
    },
    {
      "entropy": 5.478248119354248,
      "epoch": 1.5847138765653956,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004758234412500089,
      "loss": 5.2672,
      "mean_token_accuracy": 0.19903184324502946,
      "num_tokens": 31477933.0,
      "step": 14110
    },
    {
      "entropy": 5.460322284698487,
      "epoch": 1.5852754534733533,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047580533294161616,
      "loss": 5.299,
      "mean_token_accuracy": 0.18935396373271943,
      "num_tokens": 31487640.0,
      "step": 14115
    },
    {
      "entropy": 5.4627386093139645,
      "epoch": 1.5858370303813107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004757872182395036,
      "loss": 5.2292,
      "mean_token_accuracy": 0.2002650171518326,
      "num_tokens": 31499059.0,
      "step": 14120
    },
    {
      "entropy": 5.407925128936768,
      "epoch": 1.5863986072892682,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047576909714424823,
      "loss": 5.246,
      "mean_token_accuracy": 0.1941057860851288,
      "num_tokens": 31510264.0,
      "step": 14125
    },
    {
      "entropy": 5.40975775718689,
      "epoch": 1.5869601841972258,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047575096965642687,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19318452328443528,
      "num_tokens": 31520873.0,
      "step": 14130
    },
    {
      "entropy": 5.420286655426025,
      "epoch": 1.5875217611051835,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047573283577661684,
      "loss": 5.2699,
      "mean_token_accuracy": 0.19530697762966157,
      "num_tokens": 31533836.0,
      "step": 14135
    },
    {
      "entropy": 5.37579345703125,
      "epoch": 1.588083338013141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004757146955053955,
      "loss": 5.2688,
      "mean_token_accuracy": 0.1891054779291153,
      "num_tokens": 31545091.0,
      "step": 14140
    },
    {
      "entropy": 5.414040040969849,
      "epoch": 1.5886449149210984,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047569654884334036,
      "loss": 5.177,
      "mean_token_accuracy": 0.1941921278834343,
      "num_tokens": 31557139.0,
      "step": 14145
    },
    {
      "entropy": 5.491841602325439,
      "epoch": 1.589206491829056,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004756783957910292,
      "loss": 5.274,
      "mean_token_accuracy": 0.1915229231119156,
      "num_tokens": 31567928.0,
      "step": 14150
    },
    {
      "entropy": 5.368353700637817,
      "epoch": 1.5897680687370135,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047566023634904015,
      "loss": 5.2847,
      "mean_token_accuracy": 0.19151749312877656,
      "num_tokens": 31578081.0,
      "step": 14155
    },
    {
      "entropy": 5.379028081893921,
      "epoch": 1.590329645644971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004756420705179513,
      "loss": 5.1782,
      "mean_token_accuracy": 0.19942757487297058,
      "num_tokens": 31589648.0,
      "step": 14160
    },
    {
      "entropy": 5.4311606884002686,
      "epoch": 1.5908912225529286,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004756238982983412,
      "loss": 5.2642,
      "mean_token_accuracy": 0.19276570677757263,
      "num_tokens": 31601085.0,
      "step": 14165
    },
    {
      "entropy": 5.437676286697387,
      "epoch": 1.5914527994608862,
      "grad_norm": 1.125,
      "learning_rate": 0.0004756057196907883,
      "loss": 5.3209,
      "mean_token_accuracy": 0.1951897397637367,
      "num_tokens": 31612005.0,
      "step": 14170
    },
    {
      "entropy": 5.471052598953247,
      "epoch": 1.5920143763688437,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047558753469587154,
      "loss": 5.272,
      "mean_token_accuracy": 0.1923879563808441,
      "num_tokens": 31623427.0,
      "step": 14175
    },
    {
      "entropy": 5.564110136032104,
      "epoch": 1.5925759532768011,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047556934331416986,
      "loss": 5.4101,
      "mean_token_accuracy": 0.18384131640195847,
      "num_tokens": 31635086.0,
      "step": 14180
    },
    {
      "entropy": 5.445308208465576,
      "epoch": 1.5931375301847588,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004755511455462624,
      "loss": 5.2593,
      "mean_token_accuracy": 0.19438714832067489,
      "num_tokens": 31646707.0,
      "step": 14185
    },
    {
      "entropy": 5.372581958770752,
      "epoch": 1.5936991070927164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004755329413927287,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1959525838494301,
      "num_tokens": 31657327.0,
      "step": 14190
    },
    {
      "entropy": 5.436516952514649,
      "epoch": 1.5942606840006739,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004755147308541485,
      "loss": 5.1731,
      "mean_token_accuracy": 0.19982902109622955,
      "num_tokens": 31668075.0,
      "step": 14195
    },
    {
      "entropy": 5.419575786590576,
      "epoch": 1.5948222609086313,
      "grad_norm": 1.125,
      "learning_rate": 0.00047549651393110145,
      "loss": 5.2995,
      "mean_token_accuracy": 0.18631188273429872,
      "num_tokens": 31680270.0,
      "step": 14200
    },
    {
      "entropy": 5.369943714141845,
      "epoch": 1.595383837816589,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047547829062416757,
      "loss": 5.148,
      "mean_token_accuracy": 0.20311090350151062,
      "num_tokens": 31691271.0,
      "step": 14205
    },
    {
      "entropy": 5.378877115249634,
      "epoch": 1.5959454147245467,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047546006093392726,
      "loss": 5.172,
      "mean_token_accuracy": 0.200679974257946,
      "num_tokens": 31702982.0,
      "step": 14210
    },
    {
      "entropy": 5.439656972885132,
      "epoch": 1.596506991632504,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004754418248609608,
      "loss": 5.2923,
      "mean_token_accuracy": 0.19118187874555587,
      "num_tokens": 31715129.0,
      "step": 14215
    },
    {
      "entropy": 5.495428419113159,
      "epoch": 1.5970685685404615,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004754235824058489,
      "loss": 5.3622,
      "mean_token_accuracy": 0.1909250497817993,
      "num_tokens": 31726392.0,
      "step": 14220
    },
    {
      "entropy": 5.497851705551147,
      "epoch": 1.5976301454484192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004754053335691724,
      "loss": 5.3008,
      "mean_token_accuracy": 0.19638860672712327,
      "num_tokens": 31738143.0,
      "step": 14225
    },
    {
      "entropy": 5.463070869445801,
      "epoch": 1.5981917223563769,
      "grad_norm": 1.125,
      "learning_rate": 0.00047538707835151237,
      "loss": 5.2717,
      "mean_token_accuracy": 0.19194953441619872,
      "num_tokens": 31748510.0,
      "step": 14230
    },
    {
      "entropy": 5.3450291633605955,
      "epoch": 1.5987532992643343,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047536881675345,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19870778322219848,
      "num_tokens": 31759378.0,
      "step": 14235
    },
    {
      "entropy": 5.5384656429290775,
      "epoch": 1.5993148761722917,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004753505487755669,
      "loss": 5.392,
      "mean_token_accuracy": 0.18203849196434022,
      "num_tokens": 31771627.0,
      "step": 14240
    },
    {
      "entropy": 5.393084239959717,
      "epoch": 1.5998764530802494,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047533227441844453,
      "loss": 5.1827,
      "mean_token_accuracy": 0.201456318795681,
      "num_tokens": 31781990.0,
      "step": 14245
    },
    {
      "entropy": 5.404367303848266,
      "epoch": 1.6004380299882068,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047531399368266485,
      "loss": 5.2439,
      "mean_token_accuracy": 0.19421983808279036,
      "num_tokens": 31792144.0,
      "step": 14250
    },
    {
      "entropy": 5.493143844604492,
      "epoch": 1.6009996068961643,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004752957065688099,
      "loss": 5.2699,
      "mean_token_accuracy": 0.1930365815758705,
      "num_tokens": 31804022.0,
      "step": 14255
    },
    {
      "entropy": 5.49880747795105,
      "epoch": 1.601561183804122,
      "grad_norm": 1.109375,
      "learning_rate": 0.000475277413077462,
      "loss": 5.3135,
      "mean_token_accuracy": 0.1915546253323555,
      "num_tokens": 31815108.0,
      "step": 14260
    },
    {
      "entropy": 5.356485748291016,
      "epoch": 1.6021227607120796,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047525911320920363,
      "loss": 5.2877,
      "mean_token_accuracy": 0.19406869858503342,
      "num_tokens": 31825727.0,
      "step": 14265
    },
    {
      "entropy": 5.4112602233886715,
      "epoch": 1.602684337620037,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004752408069646174,
      "loss": 5.2272,
      "mean_token_accuracy": 0.2054294154047966,
      "num_tokens": 31836168.0,
      "step": 14270
    },
    {
      "entropy": 5.545200300216675,
      "epoch": 1.6032459145279945,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047522249434428623,
      "loss": 5.3445,
      "mean_token_accuracy": 0.1859061151742935,
      "num_tokens": 31847383.0,
      "step": 14275
    },
    {
      "entropy": 5.361667442321777,
      "epoch": 1.6038074914359521,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004752041753487931,
      "loss": 5.1673,
      "mean_token_accuracy": 0.20096057206392287,
      "num_tokens": 31858679.0,
      "step": 14280
    },
    {
      "entropy": 5.38459734916687,
      "epoch": 1.6043690683439098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004751858499787215,
      "loss": 5.2551,
      "mean_token_accuracy": 0.198748616874218,
      "num_tokens": 31870195.0,
      "step": 14285
    },
    {
      "entropy": 5.552084827423096,
      "epoch": 1.6049306452518672,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004751675182346547,
      "loss": 5.2656,
      "mean_token_accuracy": 0.19633141309022903,
      "num_tokens": 31881175.0,
      "step": 14290
    },
    {
      "entropy": 5.407534217834472,
      "epoch": 1.6054922221598247,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047514918011717656,
      "loss": 5.1882,
      "mean_token_accuracy": 0.20395768582820892,
      "num_tokens": 31891199.0,
      "step": 14295
    },
    {
      "entropy": 5.560690259933471,
      "epoch": 1.6060537990677823,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047513083562687085,
      "loss": 5.3851,
      "mean_token_accuracy": 0.18717293292284012,
      "num_tokens": 31903379.0,
      "step": 14300
    },
    {
      "entropy": 5.464174461364746,
      "epoch": 1.60661537597574,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004751124847643217,
      "loss": 5.1851,
      "mean_token_accuracy": 0.19634411185979844,
      "num_tokens": 31913958.0,
      "step": 14305
    },
    {
      "entropy": 5.307919645309449,
      "epoch": 1.6071769528836974,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004750941275301135,
      "loss": 5.1568,
      "mean_token_accuracy": 0.1933925211429596,
      "num_tokens": 31924547.0,
      "step": 14310
    },
    {
      "entropy": 5.453706884384156,
      "epoch": 1.6077385297916549,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004750757639248306,
      "loss": 5.3411,
      "mean_token_accuracy": 0.19382108002901077,
      "num_tokens": 31936987.0,
      "step": 14315
    },
    {
      "entropy": 5.4506134510040285,
      "epoch": 1.6083001066996125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047505739394905776,
      "loss": 5.2071,
      "mean_token_accuracy": 0.1954615518450737,
      "num_tokens": 31948165.0,
      "step": 14320
    },
    {
      "entropy": 5.483512210845947,
      "epoch": 1.6088616836075702,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047503901760338003,
      "loss": 5.2704,
      "mean_token_accuracy": 0.19478003084659576,
      "num_tokens": 31959206.0,
      "step": 14325
    },
    {
      "entropy": 5.406462669372559,
      "epoch": 1.6094232605155276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047502063488838225,
      "loss": 5.2548,
      "mean_token_accuracy": 0.19421267956495286,
      "num_tokens": 31970804.0,
      "step": 14330
    },
    {
      "entropy": 5.344238185882569,
      "epoch": 1.609984837423485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047500224580464994,
      "loss": 5.1695,
      "mean_token_accuracy": 0.19820151180028917,
      "num_tokens": 31982961.0,
      "step": 14335
    },
    {
      "entropy": 5.47658953666687,
      "epoch": 1.6105464143314427,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00047498385035276856,
      "loss": 5.223,
      "mean_token_accuracy": 0.19264578968286514,
      "num_tokens": 31993812.0,
      "step": 14340
    },
    {
      "entropy": 5.397758817672729,
      "epoch": 1.6111079912394002,
      "grad_norm": 1.125,
      "learning_rate": 0.0004749654485333238,
      "loss": 5.2958,
      "mean_token_accuracy": 0.19647812694311143,
      "num_tokens": 32004688.0,
      "step": 14345
    },
    {
      "entropy": 5.387997341156006,
      "epoch": 1.6116695681473576,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047494704034690163,
      "loss": 5.1612,
      "mean_token_accuracy": 0.19436236917972566,
      "num_tokens": 32014653.0,
      "step": 14350
    },
    {
      "entropy": 5.375695848464966,
      "epoch": 1.6122311450553153,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047492862579408816,
      "loss": 5.2258,
      "mean_token_accuracy": 0.19059600830078124,
      "num_tokens": 32025638.0,
      "step": 14355
    },
    {
      "entropy": 5.323161935806274,
      "epoch": 1.612792721963273,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047491020487546966,
      "loss": 5.1364,
      "mean_token_accuracy": 0.20041654258966446,
      "num_tokens": 32036316.0,
      "step": 14360
    },
    {
      "entropy": 5.402593994140625,
      "epoch": 1.6133542988712304,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004748917775916327,
      "loss": 5.1686,
      "mean_token_accuracy": 0.19886563569307328,
      "num_tokens": 32047355.0,
      "step": 14365
    },
    {
      "entropy": 5.414781951904297,
      "epoch": 1.6139158757791878,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047487334394316405,
      "loss": 5.2875,
      "mean_token_accuracy": 0.18405257165431976,
      "num_tokens": 32058897.0,
      "step": 14370
    },
    {
      "entropy": 5.489324474334717,
      "epoch": 1.6144774526871455,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004748549039306506,
      "loss": 5.2602,
      "mean_token_accuracy": 0.19208630174398422,
      "num_tokens": 32069932.0,
      "step": 14375
    },
    {
      "entropy": 5.3789068222045895,
      "epoch": 1.6150390295951031,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004748364575546795,
      "loss": 5.1719,
      "mean_token_accuracy": 0.1965685799717903,
      "num_tokens": 32081359.0,
      "step": 14380
    },
    {
      "entropy": 5.341517734527588,
      "epoch": 1.6156006065030606,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047481800481583806,
      "loss": 5.1326,
      "mean_token_accuracy": 0.1974394753575325,
      "num_tokens": 32093050.0,
      "step": 14385
    },
    {
      "entropy": 5.494117784500122,
      "epoch": 1.616162183411018,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004747995457147139,
      "loss": 5.2779,
      "mean_token_accuracy": 0.19301560819149016,
      "num_tokens": 32104623.0,
      "step": 14390
    },
    {
      "entropy": 5.3884031772613525,
      "epoch": 1.6167237603189757,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047478108025189467,
      "loss": 5.2427,
      "mean_token_accuracy": 0.18624853789806367,
      "num_tokens": 32115845.0,
      "step": 14395
    },
    {
      "entropy": 5.395812511444092,
      "epoch": 1.6172853372269334,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047476260842796837,
      "loss": 5.2104,
      "mean_token_accuracy": 0.1907674178481102,
      "num_tokens": 32127066.0,
      "step": 14400
    },
    {
      "entropy": 5.4036780834198,
      "epoch": 1.6178469141348908,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004747441302435231,
      "loss": 5.3225,
      "mean_token_accuracy": 0.1909378781914711,
      "num_tokens": 32138344.0,
      "step": 14405
    },
    {
      "entropy": 5.47201771736145,
      "epoch": 1.6184084910428482,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047472564569914726,
      "loss": 5.2809,
      "mean_token_accuracy": 0.1864091068506241,
      "num_tokens": 32149054.0,
      "step": 14410
    },
    {
      "entropy": 5.438974523544312,
      "epoch": 1.618970067950806,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00047470715479542945,
      "loss": 5.1827,
      "mean_token_accuracy": 0.18727043271064758,
      "num_tokens": 32160385.0,
      "step": 14415
    },
    {
      "entropy": 5.43072738647461,
      "epoch": 1.6195316448587636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004746886575329584,
      "loss": 5.3308,
      "mean_token_accuracy": 0.18574255257844924,
      "num_tokens": 32171856.0,
      "step": 14420
    },
    {
      "entropy": 5.402473831176758,
      "epoch": 1.620093221766721,
      "grad_norm": 1.046875,
      "learning_rate": 0.00047467015391232305,
      "loss": 5.2583,
      "mean_token_accuracy": 0.19282355457544326,
      "num_tokens": 32182690.0,
      "step": 14425
    },
    {
      "entropy": 5.418821048736572,
      "epoch": 1.6206547986746784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047465164393411245,
      "loss": 5.2318,
      "mean_token_accuracy": 0.19286864697933198,
      "num_tokens": 32194045.0,
      "step": 14430
    },
    {
      "entropy": 5.453915452957153,
      "epoch": 1.621216375582636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047463312759891615,
      "loss": 5.2889,
      "mean_token_accuracy": 0.19005917012691498,
      "num_tokens": 32205699.0,
      "step": 14435
    },
    {
      "entropy": 5.492156648635865,
      "epoch": 1.6217779524905935,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004746146049073236,
      "loss": 5.2908,
      "mean_token_accuracy": 0.1873659834265709,
      "num_tokens": 32217974.0,
      "step": 14440
    },
    {
      "entropy": 5.290875244140625,
      "epoch": 1.622339529398551,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047459607585992466,
      "loss": 5.1834,
      "mean_token_accuracy": 0.19084254950284957,
      "num_tokens": 32228723.0,
      "step": 14445
    },
    {
      "entropy": 5.402580642700196,
      "epoch": 1.6229011063065086,
      "grad_norm": 1.0,
      "learning_rate": 0.00047457754045730923,
      "loss": 5.3338,
      "mean_token_accuracy": 0.18994588851928712,
      "num_tokens": 32241048.0,
      "step": 14450
    },
    {
      "entropy": 5.406540584564209,
      "epoch": 1.6234626832144663,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047455899870006745,
      "loss": 5.1998,
      "mean_token_accuracy": 0.19607583433389664,
      "num_tokens": 32251255.0,
      "step": 14455
    },
    {
      "entropy": 5.384950923919678,
      "epoch": 1.6240242601224237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004745404505887899,
      "loss": 5.2477,
      "mean_token_accuracy": 0.19308674186468125,
      "num_tokens": 32262954.0,
      "step": 14460
    },
    {
      "entropy": 5.399402093887329,
      "epoch": 1.6245858370303812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004745218961240669,
      "loss": 5.2402,
      "mean_token_accuracy": 0.19668905884027482,
      "num_tokens": 32275445.0,
      "step": 14465
    },
    {
      "entropy": 5.436430358886719,
      "epoch": 1.6251474139383388,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004745033353064893,
      "loss": 5.2018,
      "mean_token_accuracy": 0.20040460228919982,
      "num_tokens": 32285976.0,
      "step": 14470
    },
    {
      "entropy": 5.3414772033691404,
      "epoch": 1.6257089908462965,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004744847681366482,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19918595105409623,
      "num_tokens": 32296304.0,
      "step": 14475
    },
    {
      "entropy": 5.347096681594849,
      "epoch": 1.626270567754254,
      "grad_norm": 1.125,
      "learning_rate": 0.00047446619461513466,
      "loss": 5.2393,
      "mean_token_accuracy": 0.20142198652029036,
      "num_tokens": 32307085.0,
      "step": 14480
    },
    {
      "entropy": 5.355552816390992,
      "epoch": 1.6268321446622114,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004744476147425402,
      "loss": 5.147,
      "mean_token_accuracy": 0.2082793414592743,
      "num_tokens": 32316814.0,
      "step": 14485
    },
    {
      "entropy": 5.371435976028442,
      "epoch": 1.627393721570169,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047442902851945634,
      "loss": 5.2367,
      "mean_token_accuracy": 0.1946148991584778,
      "num_tokens": 32327820.0,
      "step": 14490
    },
    {
      "entropy": 5.529321336746216,
      "epoch": 1.6279552984781267,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047441043594647486,
      "loss": 5.2352,
      "mean_token_accuracy": 0.19784968942403794,
      "num_tokens": 32338185.0,
      "step": 14495
    },
    {
      "entropy": 5.383028173446656,
      "epoch": 1.6285168753860841,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047439183702418765,
      "loss": 5.1397,
      "mean_token_accuracy": 0.19972691982984542,
      "num_tokens": 32348537.0,
      "step": 14500
    },
    {
      "entropy": 5.375629091262818,
      "epoch": 1.6290784522940416,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047437323175318716,
      "loss": 5.2788,
      "mean_token_accuracy": 0.19280828088521956,
      "num_tokens": 32359094.0,
      "step": 14505
    },
    {
      "entropy": 5.293658447265625,
      "epoch": 1.6296400292019992,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047435462013406554,
      "loss": 5.1099,
      "mean_token_accuracy": 0.2046833321452141,
      "num_tokens": 32370811.0,
      "step": 14510
    },
    {
      "entropy": 5.438618755340576,
      "epoch": 1.630201606109957,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004743360021674155,
      "loss": 5.3102,
      "mean_token_accuracy": 0.19130280762910842,
      "num_tokens": 32381522.0,
      "step": 14515
    },
    {
      "entropy": 5.381268882751465,
      "epoch": 1.6307631830179143,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047431737785382996,
      "loss": 5.2594,
      "mean_token_accuracy": 0.19215675592422485,
      "num_tokens": 32393221.0,
      "step": 14520
    },
    {
      "entropy": 5.4749448776245115,
      "epoch": 1.6313247599258718,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047429874719390165,
      "loss": 5.2093,
      "mean_token_accuracy": 0.1909129500389099,
      "num_tokens": 32404007.0,
      "step": 14525
    },
    {
      "entropy": 5.438856172561645,
      "epoch": 1.6318863368338294,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047428011018822404,
      "loss": 5.3313,
      "mean_token_accuracy": 0.18930992633104324,
      "num_tokens": 32414318.0,
      "step": 14530
    },
    {
      "entropy": 5.420485925674439,
      "epoch": 1.6324479137417869,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047426146683739043,
      "loss": 5.2675,
      "mean_token_accuracy": 0.19623674899339677,
      "num_tokens": 32425493.0,
      "step": 14535
    },
    {
      "entropy": 5.45293345451355,
      "epoch": 1.6330094906497443,
      "grad_norm": 1.28125,
      "learning_rate": 0.00047424281714199436,
      "loss": 5.2135,
      "mean_token_accuracy": 0.19775812327861786,
      "num_tokens": 32435765.0,
      "step": 14540
    },
    {
      "entropy": 5.494131755828858,
      "epoch": 1.633571067557702,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047422416110262975,
      "loss": 5.2882,
      "mean_token_accuracy": 0.19806757420301438,
      "num_tokens": 32446478.0,
      "step": 14545
    },
    {
      "entropy": 5.34685320854187,
      "epoch": 1.6341326444656596,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047420549871989053,
      "loss": 5.2689,
      "mean_token_accuracy": 0.20041225105524063,
      "num_tokens": 32456866.0,
      "step": 14550
    },
    {
      "entropy": 5.42845573425293,
      "epoch": 1.634694221373617,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000474186829994371,
      "loss": 5.1848,
      "mean_token_accuracy": 0.2027999296784401,
      "num_tokens": 32466973.0,
      "step": 14555
    },
    {
      "entropy": 5.506003427505493,
      "epoch": 1.6352557982815745,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047416815492666564,
      "loss": 5.3389,
      "mean_token_accuracy": 0.18898268938064575,
      "num_tokens": 32478118.0,
      "step": 14560
    },
    {
      "entropy": 5.326248979568481,
      "epoch": 1.6358173751895322,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004741494735173689,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20445750653743744,
      "num_tokens": 32488255.0,
      "step": 14565
    },
    {
      "entropy": 5.413556337356567,
      "epoch": 1.6363789520974898,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004741307857670757,
      "loss": 5.2541,
      "mean_token_accuracy": 0.19524148404598235,
      "num_tokens": 32498685.0,
      "step": 14570
    },
    {
      "entropy": 5.417917490005493,
      "epoch": 1.6369405290054473,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00047411209167638106,
      "loss": 5.2591,
      "mean_token_accuracy": 0.195913402736187,
      "num_tokens": 32510457.0,
      "step": 14575
    },
    {
      "entropy": 5.404718732833862,
      "epoch": 1.6375021059134047,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004740933912458801,
      "loss": 5.2292,
      "mean_token_accuracy": 0.19901012480258942,
      "num_tokens": 32522495.0,
      "step": 14580
    },
    {
      "entropy": 5.5153501510620115,
      "epoch": 1.6380636828213624,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047407468447616845,
      "loss": 5.3279,
      "mean_token_accuracy": 0.19305233657360077,
      "num_tokens": 32534891.0,
      "step": 14585
    },
    {
      "entropy": 5.400802993774414,
      "epoch": 1.63862525972932,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047405597136784166,
      "loss": 5.191,
      "mean_token_accuracy": 0.1986464574933052,
      "num_tokens": 32545555.0,
      "step": 14590
    },
    {
      "entropy": 5.348240613937378,
      "epoch": 1.6391868366372775,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004740372519214955,
      "loss": 5.2389,
      "mean_token_accuracy": 0.1956884890794754,
      "num_tokens": 32556851.0,
      "step": 14595
    },
    {
      "entropy": 5.529170227050781,
      "epoch": 1.639748413545235,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047401852613772604,
      "loss": 5.3885,
      "mean_token_accuracy": 0.1852370023727417,
      "num_tokens": 32569028.0,
      "step": 14600
    },
    {
      "entropy": 5.570987367630005,
      "epoch": 1.6403099904531926,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004739997940171295,
      "loss": 5.3221,
      "mean_token_accuracy": 0.19265488535165787,
      "num_tokens": 32581510.0,
      "step": 14605
    },
    {
      "entropy": 5.347806787490844,
      "epoch": 1.6408715673611503,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047398105556030234,
      "loss": 5.1891,
      "mean_token_accuracy": 0.19193412661552428,
      "num_tokens": 32592914.0,
      "step": 14610
    },
    {
      "entropy": 5.303299188613892,
      "epoch": 1.6414331442691077,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047396231076784123,
      "loss": 5.1293,
      "mean_token_accuracy": 0.20402445793151855,
      "num_tokens": 32603931.0,
      "step": 14615
    },
    {
      "entropy": 5.464503335952759,
      "epoch": 1.6419947211770651,
      "grad_norm": 1.0,
      "learning_rate": 0.0004739435596403429,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1941636845469475,
      "num_tokens": 32614508.0,
      "step": 14620
    },
    {
      "entropy": 5.308857107162476,
      "epoch": 1.6425562980850228,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004739248021784046,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20576320588588715,
      "num_tokens": 32625393.0,
      "step": 14625
    },
    {
      "entropy": 5.274083423614502,
      "epoch": 1.6431178749929805,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004739060383826234,
      "loss": 5.1423,
      "mean_token_accuracy": 0.20271852165460585,
      "num_tokens": 32636645.0,
      "step": 14630
    },
    {
      "entropy": 5.525800704956055,
      "epoch": 1.6436794519009377,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047388726825359673,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18888948261737823,
      "num_tokens": 32648792.0,
      "step": 14635
    },
    {
      "entropy": 5.439349031448364,
      "epoch": 1.6442410288088953,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047386849179192235,
      "loss": 5.2292,
      "mean_token_accuracy": 0.2039506956934929,
      "num_tokens": 32658825.0,
      "step": 14640
    },
    {
      "entropy": 5.48520245552063,
      "epoch": 1.644802605716853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047384970899819813,
      "loss": 5.2539,
      "mean_token_accuracy": 0.19663765281438828,
      "num_tokens": 32670141.0,
      "step": 14645
    },
    {
      "entropy": 5.294895792007447,
      "epoch": 1.6453641826248104,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004738309198730219,
      "loss": 5.0899,
      "mean_token_accuracy": 0.2075846239924431,
      "num_tokens": 32681804.0,
      "step": 14650
    },
    {
      "entropy": 5.534151315689087,
      "epoch": 1.6459257595327679,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047381212441699217,
      "loss": 5.3486,
      "mean_token_accuracy": 0.19073261469602584,
      "num_tokens": 32692550.0,
      "step": 14655
    },
    {
      "entropy": 5.392971849441528,
      "epoch": 1.6464873364407255,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004737933226307072,
      "loss": 5.2199,
      "mean_token_accuracy": 0.1999617710709572,
      "num_tokens": 32703833.0,
      "step": 14660
    },
    {
      "entropy": 5.396618604660034,
      "epoch": 1.6470489133486832,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047377451451476586,
      "loss": 5.2524,
      "mean_token_accuracy": 0.1935041531920433,
      "num_tokens": 32713928.0,
      "step": 14665
    },
    {
      "entropy": 5.502590131759644,
      "epoch": 1.6476104902566406,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004737557000697668,
      "loss": 5.2803,
      "mean_token_accuracy": 0.19424055516719818,
      "num_tokens": 32725393.0,
      "step": 14670
    },
    {
      "entropy": 5.462159299850464,
      "epoch": 1.648172067164598,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004737368792963092,
      "loss": 5.2732,
      "mean_token_accuracy": 0.19286412596702576,
      "num_tokens": 32736047.0,
      "step": 14675
    },
    {
      "entropy": 5.41311240196228,
      "epoch": 1.6487336440725557,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047371805219499223,
      "loss": 5.2895,
      "mean_token_accuracy": 0.18969098329544068,
      "num_tokens": 32747960.0,
      "step": 14680
    },
    {
      "entropy": 5.3996007442474365,
      "epoch": 1.6492952209805134,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047369921876641547,
      "loss": 5.217,
      "mean_token_accuracy": 0.1975078195333481,
      "num_tokens": 32757851.0,
      "step": 14685
    },
    {
      "entropy": 5.380012035369873,
      "epoch": 1.6498567978884708,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047368037901117854,
      "loss": 5.2336,
      "mean_token_accuracy": 0.19729177802801132,
      "num_tokens": 32767317.0,
      "step": 14690
    },
    {
      "entropy": 5.474935579299927,
      "epoch": 1.6504183747964283,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004736615329298812,
      "loss": 5.2582,
      "mean_token_accuracy": 0.19319434016942977,
      "num_tokens": 32777729.0,
      "step": 14695
    },
    {
      "entropy": 5.444044065475464,
      "epoch": 1.650979951704386,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047364268052312367,
      "loss": 5.3153,
      "mean_token_accuracy": 0.18542221337556838,
      "num_tokens": 32789547.0,
      "step": 14700
    },
    {
      "entropy": 5.44497389793396,
      "epoch": 1.6515415286123436,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004736238217915061,
      "loss": 5.2019,
      "mean_token_accuracy": 0.20148316770792007,
      "num_tokens": 32801487.0,
      "step": 14705
    },
    {
      "entropy": 5.505596351623535,
      "epoch": 1.652103105520301,
      "grad_norm": 1.015625,
      "learning_rate": 0.00047360495673562906,
      "loss": 5.2696,
      "mean_token_accuracy": 0.20080070942640305,
      "num_tokens": 32813042.0,
      "step": 14710
    },
    {
      "entropy": 5.481604480743409,
      "epoch": 1.6526646824282585,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004735860853560931,
      "loss": 5.2714,
      "mean_token_accuracy": 0.19620713144540786,
      "num_tokens": 32825467.0,
      "step": 14715
    },
    {
      "entropy": 5.412279796600342,
      "epoch": 1.6532262593362161,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047356720765349926,
      "loss": 5.2298,
      "mean_token_accuracy": 0.1911209911108017,
      "num_tokens": 32837113.0,
      "step": 14720
    },
    {
      "entropy": 5.502401494979859,
      "epoch": 1.6537878362441738,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004735483236284485,
      "loss": 5.3302,
      "mean_token_accuracy": 0.19263520389795302,
      "num_tokens": 32849779.0,
      "step": 14725
    },
    {
      "entropy": 5.397288465499878,
      "epoch": 1.654349413152131,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047352943328154203,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19277156442403792,
      "num_tokens": 32861626.0,
      "step": 14730
    },
    {
      "entropy": 5.480932998657226,
      "epoch": 1.6549109900600887,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047351053661338145,
      "loss": 5.2888,
      "mean_token_accuracy": 0.19959497600793838,
      "num_tokens": 32872875.0,
      "step": 14735
    },
    {
      "entropy": 5.411078834533692,
      "epoch": 1.6554725669680463,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004734916336245684,
      "loss": 5.1149,
      "mean_token_accuracy": 0.19652225077152252,
      "num_tokens": 32883899.0,
      "step": 14740
    },
    {
      "entropy": 5.323318910598755,
      "epoch": 1.6560341438760038,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004734727243157047,
      "loss": 5.1303,
      "mean_token_accuracy": 0.20477212220430374,
      "num_tokens": 32894255.0,
      "step": 14745
    },
    {
      "entropy": 5.417320346832275,
      "epoch": 1.6565957207839612,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047345380868739253,
      "loss": 5.2293,
      "mean_token_accuracy": 0.19585029631853104,
      "num_tokens": 32903822.0,
      "step": 14750
    },
    {
      "entropy": 5.376106119155883,
      "epoch": 1.6571572976919189,
      "grad_norm": 1.125,
      "learning_rate": 0.0004734348867402341,
      "loss": 5.2054,
      "mean_token_accuracy": 0.2016550585627556,
      "num_tokens": 32915879.0,
      "step": 14755
    },
    {
      "entropy": 5.3328827857971195,
      "epoch": 1.6577188745998765,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004734159584748319,
      "loss": 5.2303,
      "mean_token_accuracy": 0.19117604345083236,
      "num_tokens": 32928161.0,
      "step": 14760
    },
    {
      "entropy": 5.427958154678345,
      "epoch": 1.658280451507834,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047339702389178874,
      "loss": 5.195,
      "mean_token_accuracy": 0.198985655605793,
      "num_tokens": 32938399.0,
      "step": 14765
    },
    {
      "entropy": 5.387226486206055,
      "epoch": 1.6588420284157914,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047337808299170725,
      "loss": 5.1503,
      "mean_token_accuracy": 0.19986325949430467,
      "num_tokens": 32948221.0,
      "step": 14770
    },
    {
      "entropy": 5.3802423000335695,
      "epoch": 1.659403605323749,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047335913577519075,
      "loss": 5.2274,
      "mean_token_accuracy": 0.2009157195687294,
      "num_tokens": 32959802.0,
      "step": 14775
    },
    {
      "entropy": 5.445639991760254,
      "epoch": 1.6599651822317067,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047334018224284244,
      "loss": 5.2518,
      "mean_token_accuracy": 0.19378770738840104,
      "num_tokens": 32970222.0,
      "step": 14780
    },
    {
      "entropy": 5.440185165405273,
      "epoch": 1.6605267591396642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004733212223952657,
      "loss": 5.2059,
      "mean_token_accuracy": 0.19996944814920425,
      "num_tokens": 32982241.0,
      "step": 14785
    },
    {
      "entropy": 5.387702178955078,
      "epoch": 1.6610883360476216,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047330225623306443,
      "loss": 5.1497,
      "mean_token_accuracy": 0.20301618427038193,
      "num_tokens": 32992584.0,
      "step": 14790
    },
    {
      "entropy": 5.405397272109985,
      "epoch": 1.6616499129555793,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047328328375684237,
      "loss": 5.3697,
      "mean_token_accuracy": 0.18841946125030518,
      "num_tokens": 33004080.0,
      "step": 14795
    },
    {
      "entropy": 5.431455039978028,
      "epoch": 1.662211489863537,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004732643049672036,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19864599257707596,
      "num_tokens": 33015290.0,
      "step": 14800
    },
    {
      "entropy": 5.398479986190796,
      "epoch": 1.6627730667714944,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004732453198647525,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19738767743110658,
      "num_tokens": 33026470.0,
      "step": 14805
    },
    {
      "entropy": 5.331287479400634,
      "epoch": 1.6633346436794518,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004732263284500936,
      "loss": 5.1606,
      "mean_token_accuracy": 0.20289145410060883,
      "num_tokens": 33036766.0,
      "step": 14810
    },
    {
      "entropy": 5.347796535491943,
      "epoch": 1.6638962205874095,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004732073307238314,
      "loss": 5.1628,
      "mean_token_accuracy": 0.1937631919980049,
      "num_tokens": 33047675.0,
      "step": 14815
    },
    {
      "entropy": 5.4432840824127195,
      "epoch": 1.6644577974953672,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000473188326686571,
      "loss": 5.2324,
      "mean_token_accuracy": 0.20022669136524202,
      "num_tokens": 33057613.0,
      "step": 14820
    },
    {
      "entropy": 5.555335330963135,
      "epoch": 1.6650193744033244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047316931633891736,
      "loss": 5.3306,
      "mean_token_accuracy": 0.18785106092691423,
      "num_tokens": 33070498.0,
      "step": 14825
    },
    {
      "entropy": 5.409649753570557,
      "epoch": 1.665580951311282,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004731502996814759,
      "loss": 5.1834,
      "mean_token_accuracy": 0.20277881473302842,
      "num_tokens": 33082026.0,
      "step": 14830
    },
    {
      "entropy": 5.377809000015259,
      "epoch": 1.6661425282192397,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00047313127671485194,
      "loss": 5.2047,
      "mean_token_accuracy": 0.19617745280265808,
      "num_tokens": 33093841.0,
      "step": 14835
    },
    {
      "entropy": 5.459502840042115,
      "epoch": 1.6667041051271971,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004731122474396514,
      "loss": 5.2399,
      "mean_token_accuracy": 0.1920927196741104,
      "num_tokens": 33103817.0,
      "step": 14840
    },
    {
      "entropy": 5.473745441436767,
      "epoch": 1.6672656820351546,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047309321185648,
      "loss": 5.3086,
      "mean_token_accuracy": 0.19243767112493515,
      "num_tokens": 33116006.0,
      "step": 14845
    },
    {
      "entropy": 5.445277452468872,
      "epoch": 1.6678272589431122,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047307416996594386,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19417233318090438,
      "num_tokens": 33127649.0,
      "step": 14850
    },
    {
      "entropy": 5.374706697463989,
      "epoch": 1.66838883585107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004730551217686495,
      "loss": 5.1248,
      "mean_token_accuracy": 0.20579309910535812,
      "num_tokens": 33138748.0,
      "step": 14855
    },
    {
      "entropy": 5.434906530380249,
      "epoch": 1.6689504127590273,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004730360672652031,
      "loss": 5.2859,
      "mean_token_accuracy": 0.19392026960849762,
      "num_tokens": 33150504.0,
      "step": 14860
    },
    {
      "entropy": 5.413250064849853,
      "epoch": 1.6695119896669848,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047301700645621154,
      "loss": 5.3042,
      "mean_token_accuracy": 0.19078892916440965,
      "num_tokens": 33160625.0,
      "step": 14865
    },
    {
      "entropy": 5.475200271606445,
      "epoch": 1.6700735665749424,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004729979393422817,
      "loss": 5.1975,
      "mean_token_accuracy": 0.19851555079221725,
      "num_tokens": 33172229.0,
      "step": 14870
    },
    {
      "entropy": 5.4406898021698,
      "epoch": 1.6706351434829,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004729788659240206,
      "loss": 5.2393,
      "mean_token_accuracy": 0.1951851576566696,
      "num_tokens": 33183234.0,
      "step": 14875
    },
    {
      "entropy": 5.451408100128174,
      "epoch": 1.6711967203908575,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004729597862020358,
      "loss": 5.2781,
      "mean_token_accuracy": 0.19518179148435594,
      "num_tokens": 33193976.0,
      "step": 14880
    },
    {
      "entropy": 5.434476375579834,
      "epoch": 1.671758297298815,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047294070017693453,
      "loss": 5.2353,
      "mean_token_accuracy": 0.2009290486574173,
      "num_tokens": 33205959.0,
      "step": 14885
    },
    {
      "entropy": 5.40474500656128,
      "epoch": 1.6723198742067726,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047292160784932454,
      "loss": 5.1615,
      "mean_token_accuracy": 0.19979517310857772,
      "num_tokens": 33218405.0,
      "step": 14890
    },
    {
      "entropy": 5.395599269866944,
      "epoch": 1.6728814511147303,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047290250921981386,
      "loss": 5.2316,
      "mean_token_accuracy": 0.1934172570705414,
      "num_tokens": 33230778.0,
      "step": 14895
    },
    {
      "entropy": 5.386380577087403,
      "epoch": 1.6734430280226877,
      "grad_norm": 1.3984375,
      "learning_rate": 0.00047288340428901053,
      "loss": 5.1812,
      "mean_token_accuracy": 0.1974200651049614,
      "num_tokens": 33242069.0,
      "step": 14900
    },
    {
      "entropy": 5.406944036483765,
      "epoch": 1.6740046049306452,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047286429305752283,
      "loss": 5.2514,
      "mean_token_accuracy": 0.19577035158872605,
      "num_tokens": 33252942.0,
      "step": 14905
    },
    {
      "entropy": 5.536192321777344,
      "epoch": 1.6745661818386028,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047284517552595933,
      "loss": 5.2942,
      "mean_token_accuracy": 0.19309330135583877,
      "num_tokens": 33263892.0,
      "step": 14910
    },
    {
      "entropy": 5.3496537685394285,
      "epoch": 1.6751277587465605,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004728260516949287,
      "loss": 5.1875,
      "mean_token_accuracy": 0.19644092172384262,
      "num_tokens": 33274383.0,
      "step": 14915
    },
    {
      "entropy": 5.4718180179595945,
      "epoch": 1.6756893356545177,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004728069215650399,
      "loss": 5.1911,
      "mean_token_accuracy": 0.19964352250099182,
      "num_tokens": 33285897.0,
      "step": 14920
    },
    {
      "entropy": 5.459086656570435,
      "epoch": 1.6762509125624754,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004727877851369019,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1925385370850563,
      "num_tokens": 33297192.0,
      "step": 14925
    },
    {
      "entropy": 5.431684923171997,
      "epoch": 1.676812489470433,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047276864241112417,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20503935664892198,
      "num_tokens": 33308250.0,
      "step": 14930
    },
    {
      "entropy": 5.346336936950683,
      "epoch": 1.6773740663783905,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004727494933883161,
      "loss": 5.1997,
      "mean_token_accuracy": 0.19514390528202058,
      "num_tokens": 33318515.0,
      "step": 14935
    },
    {
      "entropy": 5.477863216400147,
      "epoch": 1.677935643286348,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004727303380690876,
      "loss": 5.3403,
      "mean_token_accuracy": 0.19418278187513352,
      "num_tokens": 33329762.0,
      "step": 14940
    },
    {
      "entropy": 5.475515985488892,
      "epoch": 1.6784972201943056,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004727111764540484,
      "loss": 5.3259,
      "mean_token_accuracy": 0.19105661660432816,
      "num_tokens": 33341054.0,
      "step": 14945
    },
    {
      "entropy": 5.412602043151855,
      "epoch": 1.6790587971022632,
      "grad_norm": 1.21875,
      "learning_rate": 0.00047269200854380854,
      "loss": 5.1757,
      "mean_token_accuracy": 0.19944616854190828,
      "num_tokens": 33351524.0,
      "step": 14950
    },
    {
      "entropy": 5.399216508865356,
      "epoch": 1.6796203740102207,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047267283433897854,
      "loss": 5.2622,
      "mean_token_accuracy": 0.1858847290277481,
      "num_tokens": 33362953.0,
      "step": 14955
    },
    {
      "entropy": 5.330742311477661,
      "epoch": 1.6801819509181781,
      "grad_norm": 1.25,
      "learning_rate": 0.00047265365384016887,
      "loss": 5.1386,
      "mean_token_accuracy": 0.20772411525249482,
      "num_tokens": 33373209.0,
      "step": 14960
    },
    {
      "entropy": 5.340386152267456,
      "epoch": 1.6807435278261358,
      "grad_norm": 1.359375,
      "learning_rate": 0.00047263446704799017,
      "loss": 5.1434,
      "mean_token_accuracy": 0.19760730415582656,
      "num_tokens": 33384775.0,
      "step": 14965
    },
    {
      "entropy": 5.348722839355469,
      "epoch": 1.6813051047340934,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047261527396305343,
      "loss": 5.197,
      "mean_token_accuracy": 0.19891173392534256,
      "num_tokens": 33395457.0,
      "step": 14970
    },
    {
      "entropy": 5.51993842124939,
      "epoch": 1.6818666816420509,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00047259607458596966,
      "loss": 5.3297,
      "mean_token_accuracy": 0.19377135187387468,
      "num_tokens": 33407506.0,
      "step": 14975
    },
    {
      "entropy": 5.486283636093139,
      "epoch": 1.6824282585500083,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047257686891735023,
      "loss": 5.2085,
      "mean_token_accuracy": 0.20010705888271332,
      "num_tokens": 33418154.0,
      "step": 14980
    },
    {
      "entropy": 5.342317390441894,
      "epoch": 1.682989835457966,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047255765695780665,
      "loss": 5.2463,
      "mean_token_accuracy": 0.20051402896642684,
      "num_tokens": 33428319.0,
      "step": 14985
    },
    {
      "entropy": 5.366711139678955,
      "epoch": 1.6835514123659236,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004725384387079507,
      "loss": 5.1404,
      "mean_token_accuracy": 0.20145986676216127,
      "num_tokens": 33440429.0,
      "step": 14990
    },
    {
      "entropy": 5.617105007171631,
      "epoch": 1.684112989273881,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047251921416839425,
      "loss": 5.3626,
      "mean_token_accuracy": 0.18662802278995513,
      "num_tokens": 33451736.0,
      "step": 14995
    },
    {
      "entropy": 5.413031911849975,
      "epoch": 1.6846745661818385,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004724999833397494,
      "loss": 5.1445,
      "mean_token_accuracy": 0.19501809924840927,
      "num_tokens": 33461991.0,
      "step": 15000
    },
    {
      "epoch": 1.6846745661818385,
      "eval_entropy": 5.212921412174992,
      "eval_loss": 5.337553977966309,
      "eval_mean_token_accuracy": 0.1982726291214064,
      "eval_num_tokens": 33461991.0,
      "eval_runtime": 23.9676,
      "eval_samples_per_second": 1293.957,
      "eval_steps_per_second": 161.76,
      "step": 15000
    },
    {
      "entropy": 5.383431911468506,
      "epoch": 1.6852361430897962,
      "grad_norm": 1.25,
      "learning_rate": 0.0004724807462226285,
      "loss": 5.2965,
      "mean_token_accuracy": 0.18771005421876907,
      "num_tokens": 33473646.0,
      "step": 15005
    },
    {
      "entropy": 5.38461971282959,
      "epoch": 1.6857977199977539,
      "grad_norm": 1.171875,
      "learning_rate": 0.000472461502817644,
      "loss": 5.2285,
      "mean_token_accuracy": 0.19694608300924302,
      "num_tokens": 33485444.0,
      "step": 15010
    },
    {
      "entropy": 5.569370222091675,
      "epoch": 1.6863592969057113,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004724422531254087,
      "loss": 5.394,
      "mean_token_accuracy": 0.1867316946387291,
      "num_tokens": 33497199.0,
      "step": 15015
    },
    {
      "entropy": 5.517810392379761,
      "epoch": 1.6869208738136687,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004724229971465355,
      "loss": 5.332,
      "mean_token_accuracy": 0.19116003066301346,
      "num_tokens": 33509045.0,
      "step": 15020
    },
    {
      "entropy": 5.433034801483155,
      "epoch": 1.6874824507216264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004724037348816375,
      "loss": 5.1117,
      "mean_token_accuracy": 0.20560626089572906,
      "num_tokens": 33519730.0,
      "step": 15025
    },
    {
      "entropy": 5.402930784225464,
      "epoch": 1.6880440276295838,
      "grad_norm": 1.234375,
      "learning_rate": 0.000472384466331328,
      "loss": 5.3168,
      "mean_token_accuracy": 0.1895298182964325,
      "num_tokens": 33530393.0,
      "step": 15030
    },
    {
      "entropy": 5.416666841506958,
      "epoch": 1.6886056045375413,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004723651914962205,
      "loss": 5.2653,
      "mean_token_accuracy": 0.19534376561641692,
      "num_tokens": 33543592.0,
      "step": 15035
    },
    {
      "entropy": 5.471719741821289,
      "epoch": 1.689167181445499,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047234591037692884,
      "loss": 5.215,
      "mean_token_accuracy": 0.19633906185626984,
      "num_tokens": 33554736.0,
      "step": 15040
    },
    {
      "entropy": 5.433762741088867,
      "epoch": 1.6897287583534566,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047232662297406683,
      "loss": 5.1352,
      "mean_token_accuracy": 0.20254287719726563,
      "num_tokens": 33565428.0,
      "step": 15045
    },
    {
      "entropy": 5.479749774932861,
      "epoch": 1.690290335261414,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004723073292882485,
      "loss": 5.445,
      "mean_token_accuracy": 0.18747391700744628,
      "num_tokens": 33577102.0,
      "step": 15050
    },
    {
      "entropy": 5.514101982116699,
      "epoch": 1.6908519121693715,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047228802932008836,
      "loss": 5.2735,
      "mean_token_accuracy": 0.19936016499996184,
      "num_tokens": 33587325.0,
      "step": 15055
    },
    {
      "entropy": 5.525373077392578,
      "epoch": 1.6914134890773291,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004722687230702009,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19051536172628403,
      "num_tokens": 33599294.0,
      "step": 15060
    },
    {
      "entropy": 5.500714445114136,
      "epoch": 1.6919750659852868,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004722494105392007,
      "loss": 5.358,
      "mean_token_accuracy": 0.19168773740530015,
      "num_tokens": 33609025.0,
      "step": 15065
    },
    {
      "entropy": 5.498642110824585,
      "epoch": 1.6925366428932442,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004722300917277029,
      "loss": 5.2887,
      "mean_token_accuracy": 0.1920127436518669,
      "num_tokens": 33620881.0,
      "step": 15070
    },
    {
      "entropy": 5.437159395217895,
      "epoch": 1.6930982198012017,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004722107666363224,
      "loss": 5.2363,
      "mean_token_accuracy": 0.19724693447351455,
      "num_tokens": 33631505.0,
      "step": 15075
    },
    {
      "entropy": 5.450789976119995,
      "epoch": 1.6936597967091593,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004721914352656746,
      "loss": 5.3922,
      "mean_token_accuracy": 0.19202485531568528,
      "num_tokens": 33642560.0,
      "step": 15080
    },
    {
      "entropy": 5.514165830612183,
      "epoch": 1.694221373617117,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000472172097616375,
      "loss": 5.2622,
      "mean_token_accuracy": 0.19481472671031952,
      "num_tokens": 33654706.0,
      "step": 15085
    },
    {
      "entropy": 5.440787887573242,
      "epoch": 1.6947829505250744,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004721527536890394,
      "loss": 5.2892,
      "mean_token_accuracy": 0.19063674360513688,
      "num_tokens": 33665392.0,
      "step": 15090
    },
    {
      "entropy": 5.349322700500489,
      "epoch": 1.6953445274330319,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004721334034842836,
      "loss": 5.1341,
      "mean_token_accuracy": 0.206089523434639,
      "num_tokens": 33675504.0,
      "step": 15095
    },
    {
      "entropy": 5.491777658462524,
      "epoch": 1.6959061043409895,
      "grad_norm": 1.125,
      "learning_rate": 0.0004721140470027238,
      "loss": 5.2805,
      "mean_token_accuracy": 0.19446851313114166,
      "num_tokens": 33686505.0,
      "step": 15100
    },
    {
      "entropy": 5.50444803237915,
      "epoch": 1.6964676812489472,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047209468424497627,
      "loss": 5.2741,
      "mean_token_accuracy": 0.19369671046733855,
      "num_tokens": 33698752.0,
      "step": 15105
    },
    {
      "entropy": 5.502308368682861,
      "epoch": 1.6970292581569046,
      "grad_norm": 1.3984375,
      "learning_rate": 0.0004720753152116576,
      "loss": 5.3052,
      "mean_token_accuracy": 0.18724352717399598,
      "num_tokens": 33710323.0,
      "step": 15110
    },
    {
      "entropy": 5.481440734863281,
      "epoch": 1.697590835064862,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047205593990338443,
      "loss": 5.3867,
      "mean_token_accuracy": 0.18112934231758118,
      "num_tokens": 33721875.0,
      "step": 15115
    },
    {
      "entropy": 5.415646934509278,
      "epoch": 1.6981524119728197,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004720365583207737,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20474787503480912,
      "num_tokens": 33731520.0,
      "step": 15120
    },
    {
      "entropy": 5.4886468887329105,
      "epoch": 1.6987139888807772,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004720171704644425,
      "loss": 5.3384,
      "mean_token_accuracy": 0.1902194231748581,
      "num_tokens": 33743351.0,
      "step": 15125
    },
    {
      "entropy": 5.42932710647583,
      "epoch": 1.6992755657887346,
      "grad_norm": 1.125,
      "learning_rate": 0.0004719977763350083,
      "loss": 5.2721,
      "mean_token_accuracy": 0.19842470437288284,
      "num_tokens": 33754879.0,
      "step": 15130
    },
    {
      "entropy": 5.396392679214477,
      "epoch": 1.6998371426966923,
      "grad_norm": 1.234375,
      "learning_rate": 0.00047197837593308833,
      "loss": 5.1983,
      "mean_token_accuracy": 0.19502635002136232,
      "num_tokens": 33766671.0,
      "step": 15135
    },
    {
      "entropy": 5.500061511993408,
      "epoch": 1.70039871960465,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004719589692593006,
      "loss": 5.2458,
      "mean_token_accuracy": 0.19489597529172897,
      "num_tokens": 33778236.0,
      "step": 15140
    },
    {
      "entropy": 5.416738033294678,
      "epoch": 1.7009602965126074,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004719395563142629,
      "loss": 5.1457,
      "mean_token_accuracy": 0.19538179486989976,
      "num_tokens": 33788898.0,
      "step": 15145
    },
    {
      "entropy": 5.46825647354126,
      "epoch": 1.7015218734205648,
      "grad_norm": 1.34375,
      "learning_rate": 0.0004719201370985933,
      "loss": 5.3223,
      "mean_token_accuracy": 0.1866288736462593,
      "num_tokens": 33800641.0,
      "step": 15150
    },
    {
      "entropy": 5.46290340423584,
      "epoch": 1.7020834503285225,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047190071161291007,
      "loss": 5.2098,
      "mean_token_accuracy": 0.19810257405042647,
      "num_tokens": 33812581.0,
      "step": 15155
    },
    {
      "entropy": 5.367129755020142,
      "epoch": 1.7026450272364801,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047188127985783195,
      "loss": 5.1685,
      "mean_token_accuracy": 0.20178072303533554,
      "num_tokens": 33823378.0,
      "step": 15160
    },
    {
      "entropy": 5.450681495666504,
      "epoch": 1.7032066041444376,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047186184183397754,
      "loss": 5.2912,
      "mean_token_accuracy": 0.1933170422911644,
      "num_tokens": 33834804.0,
      "step": 15165
    },
    {
      "entropy": 5.376850128173828,
      "epoch": 1.703768181052395,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047184239754196564,
      "loss": 5.2404,
      "mean_token_accuracy": 0.19305022209882736,
      "num_tokens": 33847044.0,
      "step": 15170
    },
    {
      "entropy": 5.440149402618408,
      "epoch": 1.7043297579603527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004718229469824155,
      "loss": 5.1744,
      "mean_token_accuracy": 0.19201573729515076,
      "num_tokens": 33858077.0,
      "step": 15175
    },
    {
      "entropy": 5.426899051666259,
      "epoch": 1.7048913348683103,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047180349015594634,
      "loss": 5.2237,
      "mean_token_accuracy": 0.19823005944490432,
      "num_tokens": 33868765.0,
      "step": 15180
    },
    {
      "entropy": 5.372467422485352,
      "epoch": 1.7054529117762678,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004717840270631778,
      "loss": 5.1936,
      "mean_token_accuracy": 0.1976988673210144,
      "num_tokens": 33880415.0,
      "step": 15185
    },
    {
      "entropy": 5.397086715698242,
      "epoch": 1.7060144886842252,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047176455770472946,
      "loss": 5.2425,
      "mean_token_accuracy": 0.19397920072078706,
      "num_tokens": 33892081.0,
      "step": 15190
    },
    {
      "entropy": 5.509957408905029,
      "epoch": 1.706576065592183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047174508208122135,
      "loss": 5.3638,
      "mean_token_accuracy": 0.18950413316488265,
      "num_tokens": 33903433.0,
      "step": 15195
    },
    {
      "entropy": 5.443167448043823,
      "epoch": 1.7071376425001406,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004717256001932735,
      "loss": 5.2289,
      "mean_token_accuracy": 0.19547265619039536,
      "num_tokens": 33914706.0,
      "step": 15200
    },
    {
      "entropy": 5.400796842575073,
      "epoch": 1.707699219408098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00047170611204150624,
      "loss": 5.3659,
      "mean_token_accuracy": 0.19442434012889862,
      "num_tokens": 33927195.0,
      "step": 15205
    },
    {
      "entropy": 5.487416458129883,
      "epoch": 1.7082607963160554,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004716866176265401,
      "loss": 5.2326,
      "mean_token_accuracy": 0.1965354010462761,
      "num_tokens": 33938422.0,
      "step": 15210
    },
    {
      "entropy": 5.495540523529053,
      "epoch": 1.708822373224013,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047166711694899573,
      "loss": 5.2613,
      "mean_token_accuracy": 0.19832492619752884,
      "num_tokens": 33949775.0,
      "step": 15215
    },
    {
      "entropy": 5.380752372741699,
      "epoch": 1.7093839501319705,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047164761000949404,
      "loss": 5.2192,
      "mean_token_accuracy": 0.19707373529672623,
      "num_tokens": 33960371.0,
      "step": 15220
    },
    {
      "entropy": 5.376278257369995,
      "epoch": 1.709945527039928,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004716280968086563,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19339862763881682,
      "num_tokens": 33971326.0,
      "step": 15225
    },
    {
      "entropy": 5.507417392730713,
      "epoch": 1.7105071039478856,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004716085773471036,
      "loss": 5.2908,
      "mean_token_accuracy": 0.18933467119932174,
      "num_tokens": 33982744.0,
      "step": 15230
    },
    {
      "entropy": 5.500482034683228,
      "epoch": 1.7110686808558433,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047158905162545756,
      "loss": 5.3342,
      "mean_token_accuracy": 0.1914701297879219,
      "num_tokens": 33993647.0,
      "step": 15235
    },
    {
      "entropy": 5.387461996078491,
      "epoch": 1.7116302577638007,
      "grad_norm": 1.25,
      "learning_rate": 0.00047156951964433984,
      "loss": 5.1943,
      "mean_token_accuracy": 0.19284010231494902,
      "num_tokens": 34004026.0,
      "step": 15240
    },
    {
      "entropy": 5.423922538757324,
      "epoch": 1.7121918346717582,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047154998140437246,
      "loss": 5.2593,
      "mean_token_accuracy": 0.20143596678972245,
      "num_tokens": 34014882.0,
      "step": 15245
    },
    {
      "entropy": 5.394704532623291,
      "epoch": 1.7127534115797158,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047153043690617737,
      "loss": 5.2837,
      "mean_token_accuracy": 0.19749995023012162,
      "num_tokens": 34026548.0,
      "step": 15250
    },
    {
      "entropy": 5.4196600914001465,
      "epoch": 1.7133149884876735,
      "grad_norm": 1.328125,
      "learning_rate": 0.000471510886150377,
      "loss": 5.1451,
      "mean_token_accuracy": 0.20060894638299942,
      "num_tokens": 34038393.0,
      "step": 15255
    },
    {
      "entropy": 5.455565118789673,
      "epoch": 1.713876565395631,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004714913291375938,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19010669738054276,
      "num_tokens": 34050257.0,
      "step": 15260
    },
    {
      "entropy": 5.432827091217041,
      "epoch": 1.7144381423035884,
      "grad_norm": 1.125,
      "learning_rate": 0.00047147176586845036,
      "loss": 5.2403,
      "mean_token_accuracy": 0.19605116695165634,
      "num_tokens": 34061860.0,
      "step": 15265
    },
    {
      "entropy": 5.453761911392212,
      "epoch": 1.714999719211546,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004714521963435698,
      "loss": 5.2955,
      "mean_token_accuracy": 0.1907349243760109,
      "num_tokens": 34075247.0,
      "step": 15270
    },
    {
      "entropy": 5.478574180603028,
      "epoch": 1.7155612961195037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00047143262056357506,
      "loss": 5.1603,
      "mean_token_accuracy": 0.19145914912223816,
      "num_tokens": 34085801.0,
      "step": 15275
    },
    {
      "entropy": 5.4751685619354244,
      "epoch": 1.7161228730274611,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047141303852908965,
      "loss": 5.2615,
      "mean_token_accuracy": 0.18904366493225097,
      "num_tokens": 34095686.0,
      "step": 15280
    },
    {
      "entropy": 5.388732862472534,
      "epoch": 1.7166844499354186,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004713934502407367,
      "loss": 5.1907,
      "mean_token_accuracy": 0.20004222095012664,
      "num_tokens": 34106518.0,
      "step": 15285
    },
    {
      "entropy": 5.447650527954101,
      "epoch": 1.7172460268433762,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004713738556991402,
      "loss": 5.2396,
      "mean_token_accuracy": 0.2010048061609268,
      "num_tokens": 34117793.0,
      "step": 15290
    },
    {
      "entropy": 5.4680756568908695,
      "epoch": 1.717807603751334,
      "grad_norm": 1.2890625,
      "learning_rate": 0.000471354254904924,
      "loss": 5.3362,
      "mean_token_accuracy": 0.19700620770454408,
      "num_tokens": 34128748.0,
      "step": 15295
    },
    {
      "entropy": 5.398885488510132,
      "epoch": 1.7183691806592913,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047133464785871216,
      "loss": 5.2511,
      "mean_token_accuracy": 0.19895714223384858,
      "num_tokens": 34140597.0,
      "step": 15300
    },
    {
      "entropy": 5.4876796245574955,
      "epoch": 1.7189307575672488,
      "grad_norm": 1.234375,
      "learning_rate": 0.000471315034561129,
      "loss": 5.226,
      "mean_token_accuracy": 0.1842323884367943,
      "num_tokens": 34151823.0,
      "step": 15305
    },
    {
      "entropy": 5.392983865737915,
      "epoch": 1.7194923344752064,
      "grad_norm": 1.203125,
      "learning_rate": 0.00047129541501279897,
      "loss": 5.1714,
      "mean_token_accuracy": 0.2045551657676697,
      "num_tokens": 34163697.0,
      "step": 15310
    },
    {
      "entropy": 5.437891817092895,
      "epoch": 1.7200539113831639,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004712757892143468,
      "loss": 5.1868,
      "mean_token_accuracy": 0.20002228915691375,
      "num_tokens": 34175058.0,
      "step": 15315
    },
    {
      "entropy": 5.439745807647705,
      "epoch": 1.7206154882911213,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047125615716639744,
      "loss": 5.3461,
      "mean_token_accuracy": 0.19974631816148758,
      "num_tokens": 34186883.0,
      "step": 15320
    },
    {
      "entropy": 5.429936742782592,
      "epoch": 1.721177065199079,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004712365188695758,
      "loss": 5.2068,
      "mean_token_accuracy": 0.20141966342926027,
      "num_tokens": 34198103.0,
      "step": 15325
    },
    {
      "entropy": 5.411020374298095,
      "epoch": 1.7217386421070366,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004712168743245074,
      "loss": 5.2016,
      "mean_token_accuracy": 0.19631707668304443,
      "num_tokens": 34208747.0,
      "step": 15330
    },
    {
      "entropy": 5.417437124252319,
      "epoch": 1.722300219014994,
      "grad_norm": 1.125,
      "learning_rate": 0.00047119722353181754,
      "loss": 5.3154,
      "mean_token_accuracy": 0.19800781160593034,
      "num_tokens": 34220775.0,
      "step": 15335
    },
    {
      "entropy": 5.4675750732421875,
      "epoch": 1.7228617959229515,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00047117756649213204,
      "loss": 5.1722,
      "mean_token_accuracy": 0.20411159098148346,
      "num_tokens": 34231031.0,
      "step": 15340
    },
    {
      "entropy": 5.471322154998779,
      "epoch": 1.7234233728309092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004711579032060767,
      "loss": 5.2756,
      "mean_token_accuracy": 0.19664516150951386,
      "num_tokens": 34241850.0,
      "step": 15345
    },
    {
      "entropy": 5.453921985626221,
      "epoch": 1.7239849497388668,
      "grad_norm": 1.3671875,
      "learning_rate": 0.0004711382336742777,
      "loss": 5.1672,
      "mean_token_accuracy": 0.19878093600273133,
      "num_tokens": 34252732.0,
      "step": 15350
    },
    {
      "entropy": 5.373421144485474,
      "epoch": 1.7245465266468243,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004711185578973612,
      "loss": 5.1904,
      "mean_token_accuracy": 0.1908824324607849,
      "num_tokens": 34264066.0,
      "step": 15355
    },
    {
      "entropy": 5.523460483551025,
      "epoch": 1.7251081035547817,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004710988758759538,
      "loss": 5.3077,
      "mean_token_accuracy": 0.1866416096687317,
      "num_tokens": 34275239.0,
      "step": 15360
    },
    {
      "entropy": 5.350469350814819,
      "epoch": 1.7256696804627394,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004710791876106821,
      "loss": 5.1662,
      "mean_token_accuracy": 0.19310637414455414,
      "num_tokens": 34285723.0,
      "step": 15365
    },
    {
      "entropy": 5.405348443984986,
      "epoch": 1.726231257370697,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047105949310217304,
      "loss": 5.2028,
      "mean_token_accuracy": 0.20403896570205687,
      "num_tokens": 34296292.0,
      "step": 15370
    },
    {
      "entropy": 5.435498952865601,
      "epoch": 1.7267928342786545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004710397923510537,
      "loss": 5.2523,
      "mean_token_accuracy": 0.19956592172384263,
      "num_tokens": 34307232.0,
      "step": 15375
    },
    {
      "entropy": 5.431132555007935,
      "epoch": 1.727354411186612,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047102008535795134,
      "loss": 5.2552,
      "mean_token_accuracy": 0.20467858463525773,
      "num_tokens": 34318933.0,
      "step": 15380
    },
    {
      "entropy": 5.439201354980469,
      "epoch": 1.7279159880945696,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004710003721234934,
      "loss": 5.184,
      "mean_token_accuracy": 0.20439097434282302,
      "num_tokens": 34330431.0,
      "step": 15385
    },
    {
      "entropy": 5.479485416412354,
      "epoch": 1.7284775650025272,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047098065264830754,
      "loss": 5.2913,
      "mean_token_accuracy": 0.19338566660881043,
      "num_tokens": 34342006.0,
      "step": 15390
    },
    {
      "entropy": 5.432756614685059,
      "epoch": 1.7290391419104847,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047096092693302174,
      "loss": 5.1974,
      "mean_token_accuracy": 0.19618502706289292,
      "num_tokens": 34353390.0,
      "step": 15395
    },
    {
      "entropy": 5.34984712600708,
      "epoch": 1.7296007188184421,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047094119497826406,
      "loss": 5.202,
      "mean_token_accuracy": 0.1976607397198677,
      "num_tokens": 34364211.0,
      "step": 15400
    },
    {
      "entropy": 5.45908489227295,
      "epoch": 1.7301622957263998,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00047092145678466273,
      "loss": 5.2154,
      "mean_token_accuracy": 0.188534776866436,
      "num_tokens": 34375362.0,
      "step": 15405
    },
    {
      "entropy": 5.367363214492798,
      "epoch": 1.7307238726343572,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004709017123528462,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20604494512081145,
      "num_tokens": 34385511.0,
      "step": 15410
    },
    {
      "entropy": 5.3922755241394045,
      "epoch": 1.7312854495423147,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00047088196168344315,
      "loss": 5.3258,
      "mean_token_accuracy": 0.19064933359622954,
      "num_tokens": 34397060.0,
      "step": 15415
    },
    {
      "entropy": 5.379887866973877,
      "epoch": 1.7318470264502723,
      "grad_norm": 1.3203125,
      "learning_rate": 0.00047086220477708256,
      "loss": 5.223,
      "mean_token_accuracy": 0.19640448093414306,
      "num_tokens": 34407703.0,
      "step": 15420
    },
    {
      "entropy": 5.462883520126343,
      "epoch": 1.73240860335823,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00047084244163439335,
      "loss": 5.2523,
      "mean_token_accuracy": 0.19942819476127624,
      "num_tokens": 34418929.0,
      "step": 15425
    },
    {
      "entropy": 5.392499828338623,
      "epoch": 1.7329701802661874,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004708226722560049,
      "loss": 5.1344,
      "mean_token_accuracy": 0.20433100014925004,
      "num_tokens": 34430693.0,
      "step": 15430
    },
    {
      "entropy": 5.355180263519287,
      "epoch": 1.7335317571741449,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004708028966425465,
      "loss": 5.201,
      "mean_token_accuracy": 0.19676101058721543,
      "num_tokens": 34442299.0,
      "step": 15435
    },
    {
      "entropy": 5.441939973831177,
      "epoch": 1.7340933340821025,
      "grad_norm": 1.171875,
      "learning_rate": 0.00047078311479464797,
      "loss": 5.1938,
      "mean_token_accuracy": 0.195698519051075,
      "num_tokens": 34453426.0,
      "step": 15440
    },
    {
      "entropy": 5.409383916854859,
      "epoch": 1.7346549109900602,
      "grad_norm": 1.1875,
      "learning_rate": 0.00047076332671293917,
      "loss": 5.2072,
      "mean_token_accuracy": 0.20008658319711686,
      "num_tokens": 34463601.0,
      "step": 15445
    },
    {
      "entropy": 5.4003551483154295,
      "epoch": 1.7352164878980176,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00047074353239805005,
      "loss": 5.189,
      "mean_token_accuracy": 0.20200234204530715,
      "num_tokens": 34474938.0,
      "step": 15450
    },
    {
      "entropy": 5.372933435440063,
      "epoch": 1.735778064805975,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004707237318506111,
      "loss": 5.1338,
      "mean_token_accuracy": 0.20336511731147766,
      "num_tokens": 34485244.0,
      "step": 15455
    },
    {
      "entropy": 5.3205399990081785,
      "epoch": 1.7363396417139327,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00047070392507125245,
      "loss": 5.1342,
      "mean_token_accuracy": 0.20206877142190932,
      "num_tokens": 34496912.0,
      "step": 15460
    },
    {
      "entropy": 5.3523680686950685,
      "epoch": 1.7369012186218904,
      "grad_norm": 1.046875,
      "learning_rate": 0.000470684112060605,
      "loss": 5.2086,
      "mean_token_accuracy": 0.1984914094209671,
      "num_tokens": 34508122.0,
      "step": 15465
    },
    {
      "entropy": 5.425387811660767,
      "epoch": 1.7374627955298478,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00047066429281929944,
      "loss": 5.2145,
      "mean_token_accuracy": 0.19969786703586578,
      "num_tokens": 34519522.0,
      "step": 15470
    },
    {
      "entropy": 5.386717462539673,
      "epoch": 1.7380243724378053,
      "grad_norm": 1.015625,
      "learning_rate": 0.000470644467347967,
      "loss": 5.207,
      "mean_token_accuracy": 0.19789092540740966,
      "num_tokens": 34532832.0,
      "step": 15475
    },
    {
      "entropy": 5.442133283615112,
      "epoch": 1.738585949345763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00047062463564723884,
      "loss": 5.3224,
      "mean_token_accuracy": 0.19839340150356294,
      "num_tokens": 34544817.0,
      "step": 15480
    },
    {
      "entropy": 5.401953792572021,
      "epoch": 1.7391475262537206,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004706047977177464,
      "loss": 5.1485,
      "mean_token_accuracy": 0.1971690073609352,
      "num_tokens": 34555391.0,
      "step": 15485
    },
    {
      "entropy": 5.348716735839844,
      "epoch": 1.739709103161678,
      "grad_norm": 1.0625,
      "learning_rate": 0.00047058495356012135,
      "loss": 5.2006,
      "mean_token_accuracy": 0.19901503324508668,
      "num_tokens": 34567531.0,
      "step": 15490
    },
    {
      "entropy": 5.425822353363037,
      "epoch": 1.7402706800696355,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047056510317499545,
      "loss": 5.2269,
      "mean_token_accuracy": 0.19578372687101364,
      "num_tokens": 34578212.0,
      "step": 15495
    },
    {
      "entropy": 5.335443162918091,
      "epoch": 1.7408322569775931,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00047054524656300087,
      "loss": 5.1226,
      "mean_token_accuracy": 0.20769066959619523,
      "num_tokens": 34589331.0,
      "step": 15500
    },
    {
      "entropy": 5.365932416915894,
      "epoch": 1.7413938338855506,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00047052538372476985,
      "loss": 5.2123,
      "mean_token_accuracy": 0.1986692562699318,
      "num_tokens": 34601467.0,
      "step": 15505
    },
    {
      "entropy": 5.393848991394043,
      "epoch": 1.741955410793508,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00047050551466093466,
      "loss": 5.1852,
      "mean_token_accuracy": 0.20289252549409867,
      "num_tokens": 34611608.0,
      "step": 15510
    },
    {
      "entropy": 5.435182285308838,
      "epoch": 1.7425169877014657,
      "grad_norm": 1.125,
      "learning_rate": 0.0004704856393721281,
      "loss": 5.1458,
      "mean_token_accuracy": 0.20582075715065,
      "num_tokens": 34622495.0,
      "step": 15515
    },
    {
      "entropy": 5.461090612411499,
      "epoch": 1.7430785646094233,
      "grad_norm": 1.265625,
      "learning_rate": 0.00047046575785898297,
      "loss": 5.3516,
      "mean_token_accuracy": 0.19112324118614196,
      "num_tokens": 34633279.0,
      "step": 15520
    },
    {
      "entropy": 5.322464466094971,
      "epoch": 1.7436401415173808,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004704458701221323,
      "loss": 5.1968,
      "mean_token_accuracy": 0.19832037687301635,
      "num_tokens": 34644030.0,
      "step": 15525
    },
    {
      "entropy": 5.519890213012696,
      "epoch": 1.7442017184253382,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004704259761622094,
      "loss": 5.3144,
      "mean_token_accuracy": 0.18552902191877366,
      "num_tokens": 34655540.0,
      "step": 15530
    },
    {
      "entropy": 5.444535779953003,
      "epoch": 1.7447632953332959,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047040607597984745,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19759119898080826,
      "num_tokens": 34666752.0,
      "step": 15535
    },
    {
      "entropy": 5.4212061882019045,
      "epoch": 1.7453248722412535,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004703861695756804,
      "loss": 5.2412,
      "mean_token_accuracy": 0.18881354480981827,
      "num_tokens": 34678670.0,
      "step": 15540
    },
    {
      "entropy": 5.426372480392456,
      "epoch": 1.745886449149211,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004703662569503419,
      "loss": 5.2324,
      "mean_token_accuracy": 0.19504690915346146,
      "num_tokens": 34690848.0,
      "step": 15545
    },
    {
      "entropy": 5.426721000671387,
      "epoch": 1.7464480260571684,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00047034633810446604,
      "loss": 5.2684,
      "mean_token_accuracy": 0.18876962959766388,
      "num_tokens": 34703135.0,
      "step": 15550
    },
    {
      "entropy": 5.379529333114624,
      "epoch": 1.747009602965126,
      "grad_norm": 1.421875,
      "learning_rate": 0.0004703264130386869,
      "loss": 5.2077,
      "mean_token_accuracy": 0.19939943104982377,
      "num_tokens": 34714329.0,
      "step": 15555
    },
    {
      "entropy": 5.432118034362793,
      "epoch": 1.7475711798730837,
      "grad_norm": 1.09375,
      "learning_rate": 0.00047030648175363913,
      "loss": 5.2476,
      "mean_token_accuracy": 0.19203222244977952,
      "num_tokens": 34727032.0,
      "step": 15560
    },
    {
      "entropy": 5.477253580093384,
      "epoch": 1.7481327567810412,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004702865442499572,
      "loss": 5.2411,
      "mean_token_accuracy": 0.19594147354364394,
      "num_tokens": 34738600.0,
      "step": 15565
    },
    {
      "entropy": 5.483676338195801,
      "epoch": 1.7486943336889986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004702666005282759,
      "loss": 5.2912,
      "mean_token_accuracy": 0.19124583750963212,
      "num_tokens": 34749810.0,
      "step": 15570
    },
    {
      "entropy": 5.560154581069947,
      "epoch": 1.7492559105969563,
      "grad_norm": 1.2734375,
      "learning_rate": 0.0004702466505892304,
      "loss": 5.4411,
      "mean_token_accuracy": 0.18477430939674377,
      "num_tokens": 34762029.0,
      "step": 15575
    },
    {
      "entropy": 5.401522397994995,
      "epoch": 1.749817487504914,
      "grad_norm": 1.140625,
      "learning_rate": 0.00047022669443345575,
      "loss": 5.2968,
      "mean_token_accuracy": 0.1960921511054039,
      "num_tokens": 34773160.0,
      "step": 15580
    },
    {
      "entropy": 5.379447984695434,
      "epoch": 1.7503790644128714,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00047020673206158743,
      "loss": 5.2584,
      "mean_token_accuracy": 0.19484034329652786,
      "num_tokens": 34784659.0,
      "step": 15585
    },
    {
      "entropy": 5.554268836975098,
      "epoch": 1.7509406413208288,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000470186763474261,
      "loss": 5.2213,
      "mean_token_accuracy": 0.20007944107055664,
      "num_tokens": 34796976.0,
      "step": 15590
    },
    {
      "entropy": 5.3810502052307125,
      "epoch": 1.7515022182287865,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004701667886721125,
      "loss": 5.1789,
      "mean_token_accuracy": 0.19425161480903624,
      "num_tokens": 34809271.0,
      "step": 15595
    },
    {
      "entropy": 5.427535724639893,
      "epoch": 1.752063795136744,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004701468076557776,
      "loss": 5.2627,
      "mean_token_accuracy": 0.19548626095056534,
      "num_tokens": 34820291.0,
      "step": 15600
    },
    {
      "entropy": 5.389868831634521,
      "epoch": 1.7526253720447014,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047012682042589277,
      "loss": 5.1737,
      "mean_token_accuracy": 0.20092639923095704,
      "num_tokens": 34831224.0,
      "step": 15605
    },
    {
      "entropy": 5.451478624343872,
      "epoch": 1.753186948952659,
      "grad_norm": 1.109375,
      "learning_rate": 0.00047010682698309423,
      "loss": 5.248,
      "mean_token_accuracy": 0.19867469221353531,
      "num_tokens": 34842947.0,
      "step": 15610
    },
    {
      "entropy": 5.36272668838501,
      "epoch": 1.7537485258606167,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004700868273280186,
      "loss": 5.1901,
      "mean_token_accuracy": 0.20350303053855895,
      "num_tokens": 34853422.0,
      "step": 15615
    },
    {
      "entropy": 5.377013540267944,
      "epoch": 1.7543101027685741,
      "grad_norm": 1.078125,
      "learning_rate": 0.00047006682146130273,
      "loss": 5.2079,
      "mean_token_accuracy": 0.19281482249498366,
      "num_tokens": 34864082.0,
      "step": 15620
    },
    {
      "entropy": 5.450811672210693,
      "epoch": 1.7548716796765316,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004700468093835837,
      "loss": 5.1923,
      "mean_token_accuracy": 0.20265578478574753,
      "num_tokens": 34874094.0,
      "step": 15625
    },
    {
      "entropy": 5.352462530136108,
      "epoch": 1.7554332565844892,
      "grad_norm": 1.15625,
      "learning_rate": 0.00047002679109549864,
      "loss": 5.2287,
      "mean_token_accuracy": 0.1989588662981987,
      "num_tokens": 34884597.0,
      "step": 15630
    },
    {
      "entropy": 5.320963335037232,
      "epoch": 1.755994833492447,
      "grad_norm": 1.125,
      "learning_rate": 0.00047000676659768477,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20798184126615524,
      "num_tokens": 34894746.0,
      "step": 15635
    },
    {
      "entropy": 5.414913940429687,
      "epoch": 1.7565564104004043,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004699867358907799,
      "loss": 5.1102,
      "mean_token_accuracy": 0.20844535529613495,
      "num_tokens": 34906250.0,
      "step": 15640
    },
    {
      "entropy": 5.429441118240357,
      "epoch": 1.7571179873083618,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004699666989754217,
      "loss": 5.3798,
      "mean_token_accuracy": 0.19056003391742707,
      "num_tokens": 34917772.0,
      "step": 15645
    },
    {
      "entropy": 5.404391956329346,
      "epoch": 1.7576795642163194,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004699466558522482,
      "loss": 5.2203,
      "mean_token_accuracy": 0.19978155195713043,
      "num_tokens": 34930243.0,
      "step": 15650
    },
    {
      "entropy": 5.389616298675537,
      "epoch": 1.758241141124277,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046992660652189755,
      "loss": 5.2166,
      "mean_token_accuracy": 0.19589090943336487,
      "num_tokens": 34941357.0,
      "step": 15655
    },
    {
      "entropy": 5.523631238937378,
      "epoch": 1.7588027180322345,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046990655098500824,
      "loss": 5.3625,
      "mean_token_accuracy": 0.18431807011365892,
      "num_tokens": 34952116.0,
      "step": 15660
    },
    {
      "entropy": 5.4699948787689205,
      "epoch": 1.759364294940192,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046988648924221866,
      "loss": 5.2521,
      "mean_token_accuracy": 0.19595665484666824,
      "num_tokens": 34963707.0,
      "step": 15665
    },
    {
      "entropy": 5.4870342254638675,
      "epoch": 1.7599258718481496,
      "grad_norm": 1.2890625,
      "learning_rate": 0.0004698664212941677,
      "loss": 5.3513,
      "mean_token_accuracy": 0.19349755495786666,
      "num_tokens": 34974376.0,
      "step": 15670
    },
    {
      "entropy": 5.432519626617432,
      "epoch": 1.7604874487561073,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046984634714149437,
      "loss": 5.2319,
      "mean_token_accuracy": 0.19587890952825546,
      "num_tokens": 34986183.0,
      "step": 15675
    },
    {
      "entropy": 5.51599268913269,
      "epoch": 1.7610490256640647,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004698262667848377,
      "loss": 5.258,
      "mean_token_accuracy": 0.19934463798999785,
      "num_tokens": 34997137.0,
      "step": 15680
    },
    {
      "entropy": 5.527636432647705,
      "epoch": 1.7616106025720222,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046980618022483714,
      "loss": 5.4044,
      "mean_token_accuracy": 0.18971830308437349,
      "num_tokens": 35008944.0,
      "step": 15685
    },
    {
      "entropy": 5.414469957351685,
      "epoch": 1.7621721794799798,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046978608746213224,
      "loss": 5.3067,
      "mean_token_accuracy": 0.18819376826286316,
      "num_tokens": 35019972.0,
      "step": 15690
    },
    {
      "entropy": 5.3997743129730225,
      "epoch": 1.7627337563879373,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004697659884973628,
      "loss": 5.175,
      "mean_token_accuracy": 0.19715268164873123,
      "num_tokens": 35031285.0,
      "step": 15695
    },
    {
      "entropy": 5.383348608016968,
      "epoch": 1.7632953332958947,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004697458833311687,
      "loss": 5.1467,
      "mean_token_accuracy": 0.19552053213119508,
      "num_tokens": 35042553.0,
      "step": 15700
    },
    {
      "entropy": 5.482905864715576,
      "epoch": 1.7638569102038524,
      "grad_norm": 1.125,
      "learning_rate": 0.0004697257719641902,
      "loss": 5.369,
      "mean_token_accuracy": 0.18421116918325425,
      "num_tokens": 35053637.0,
      "step": 15705
    },
    {
      "entropy": 5.4792650699615475,
      "epoch": 1.76441848711181,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004697056543970676,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19651597887277603,
      "num_tokens": 35064754.0,
      "step": 15710
    },
    {
      "entropy": 5.4876007556915285,
      "epoch": 1.7649800640197675,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046968553063044146,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19566460847854614,
      "num_tokens": 35075816.0,
      "step": 15715
    },
    {
      "entropy": 5.441746807098388,
      "epoch": 1.765541640927725,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046966540066495247,
      "loss": 5.3119,
      "mean_token_accuracy": 0.19390682876110077,
      "num_tokens": 35088647.0,
      "step": 15720
    },
    {
      "entropy": 5.4145777225494385,
      "epoch": 1.7661032178356826,
      "grad_norm": 1.125,
      "learning_rate": 0.00046964526450124164,
      "loss": 5.1801,
      "mean_token_accuracy": 0.19774106591939927,
      "num_tokens": 35099372.0,
      "step": 15725
    },
    {
      "entropy": 5.4915234565734865,
      "epoch": 1.7666647947436402,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046962512213995,
      "loss": 5.3307,
      "mean_token_accuracy": 0.18832096308469773,
      "num_tokens": 35110663.0,
      "step": 15730
    },
    {
      "entropy": 5.602528238296509,
      "epoch": 1.7672263716515977,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046960497358171913,
      "loss": 5.4226,
      "mean_token_accuracy": 0.18312822729349137,
      "num_tokens": 35122406.0,
      "step": 15735
    },
    {
      "entropy": 5.389739370346069,
      "epoch": 1.7677879485595551,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046958481882719035,
      "loss": 5.1976,
      "mean_token_accuracy": 0.19650693535804747,
      "num_tokens": 35133444.0,
      "step": 15740
    },
    {
      "entropy": 5.367168855667114,
      "epoch": 1.7683495254675128,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046956465787700554,
      "loss": 5.2169,
      "mean_token_accuracy": 0.19533205926418304,
      "num_tokens": 35143840.0,
      "step": 15745
    },
    {
      "entropy": 5.542659187316895,
      "epoch": 1.7689111023754704,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046954449073180644,
      "loss": 5.389,
      "mean_token_accuracy": 0.18226466327905655,
      "num_tokens": 35155050.0,
      "step": 15750
    },
    {
      "entropy": 5.499002504348755,
      "epoch": 1.7694726792834279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004695243173922354,
      "loss": 5.2972,
      "mean_token_accuracy": 0.19458663314580918,
      "num_tokens": 35167649.0,
      "step": 15755
    },
    {
      "entropy": 5.429230070114135,
      "epoch": 1.7700342561913853,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046950413785893465,
      "loss": 5.2253,
      "mean_token_accuracy": 0.19821490347385406,
      "num_tokens": 35179238.0,
      "step": 15760
    },
    {
      "entropy": 5.4547093391418455,
      "epoch": 1.770595833099343,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046948395213254667,
      "loss": 5.2908,
      "mean_token_accuracy": 0.19794457107782365,
      "num_tokens": 35189510.0,
      "step": 15765
    },
    {
      "entropy": 5.4824127674102785,
      "epoch": 1.7711574100073006,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004694637602137143,
      "loss": 5.2617,
      "mean_token_accuracy": 0.1932627946138382,
      "num_tokens": 35199727.0,
      "step": 15770
    },
    {
      "entropy": 5.4371294498443605,
      "epoch": 1.771718986915258,
      "grad_norm": 1.28125,
      "learning_rate": 0.0004694435621030803,
      "loss": 5.2235,
      "mean_token_accuracy": 0.20334824472665786,
      "num_tokens": 35210311.0,
      "step": 15775
    },
    {
      "entropy": 5.450145435333252,
      "epoch": 1.7722805638232155,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046942335780128785,
      "loss": 5.2369,
      "mean_token_accuracy": 0.20194853097200394,
      "num_tokens": 35222755.0,
      "step": 15780
    },
    {
      "entropy": 5.425462627410889,
      "epoch": 1.7728421407311732,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046940314730898035,
      "loss": 5.2641,
      "mean_token_accuracy": 0.19422116428613662,
      "num_tokens": 35234337.0,
      "step": 15785
    },
    {
      "entropy": 5.394010353088379,
      "epoch": 1.7734037176391306,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004693829306268012,
      "loss": 5.2319,
      "mean_token_accuracy": 0.19464054703712463,
      "num_tokens": 35245736.0,
      "step": 15790
    },
    {
      "entropy": 5.433611869812012,
      "epoch": 1.773965294547088,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004693627077553942,
      "loss": 5.1845,
      "mean_token_accuracy": 0.20585317462682723,
      "num_tokens": 35256368.0,
      "step": 15795
    },
    {
      "entropy": 5.387801694869995,
      "epoch": 1.7745268714550457,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004693424786954031,
      "loss": 5.2728,
      "mean_token_accuracy": 0.1972307786345482,
      "num_tokens": 35267238.0,
      "step": 15800
    },
    {
      "entropy": 5.397465515136719,
      "epoch": 1.7750884483630034,
      "grad_norm": 1.3671875,
      "learning_rate": 0.00046932224344747224,
      "loss": 5.1852,
      "mean_token_accuracy": 0.1980837494134903,
      "num_tokens": 35277113.0,
      "step": 15805
    },
    {
      "entropy": 5.396743965148926,
      "epoch": 1.7756500252709608,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046930200201224573,
      "loss": 5.1412,
      "mean_token_accuracy": 0.19477415829896927,
      "num_tokens": 35288504.0,
      "step": 15810
    },
    {
      "entropy": 5.359449863433838,
      "epoch": 1.7762116021789183,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004692817543903681,
      "loss": 5.183,
      "mean_token_accuracy": 0.2036059856414795,
      "num_tokens": 35301215.0,
      "step": 15815
    },
    {
      "entropy": 5.367363214492798,
      "epoch": 1.776773179086876,
      "grad_norm": 1.234375,
      "learning_rate": 0.000469261500582484,
      "loss": 5.1739,
      "mean_token_accuracy": 0.1997649997472763,
      "num_tokens": 35311997.0,
      "step": 15820
    },
    {
      "entropy": 5.432251119613648,
      "epoch": 1.7773347559948336,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046924124058923843,
      "loss": 5.2651,
      "mean_token_accuracy": 0.19380033761262894,
      "num_tokens": 35323217.0,
      "step": 15825
    },
    {
      "entropy": 5.441873693466187,
      "epoch": 1.777896332902791,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004692209744112765,
      "loss": 5.2813,
      "mean_token_accuracy": 0.199746210873127,
      "num_tokens": 35333102.0,
      "step": 15830
    },
    {
      "entropy": 5.475516653060913,
      "epoch": 1.7784579098107485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004692007020492433,
      "loss": 5.3012,
      "mean_token_accuracy": 0.19270191639661788,
      "num_tokens": 35346394.0,
      "step": 15835
    },
    {
      "entropy": 5.339158821105957,
      "epoch": 1.7790194867187061,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004691804235037845,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20481023639440538,
      "num_tokens": 35357827.0,
      "step": 15840
    },
    {
      "entropy": 5.399376058578492,
      "epoch": 1.7795810636266638,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004691601387755456,
      "loss": 5.2325,
      "mean_token_accuracy": 0.1978169083595276,
      "num_tokens": 35370517.0,
      "step": 15845
    },
    {
      "entropy": 5.414783763885498,
      "epoch": 1.7801426405346212,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046913984786517275,
      "loss": 5.2694,
      "mean_token_accuracy": 0.19282466322183608,
      "num_tokens": 35381830.0,
      "step": 15850
    },
    {
      "entropy": 5.539386320114136,
      "epoch": 1.7807042174425787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046911955077331177,
      "loss": 5.3501,
      "mean_token_accuracy": 0.1839522033929825,
      "num_tokens": 35393161.0,
      "step": 15855
    },
    {
      "entropy": 5.452949810028076,
      "epoch": 1.7812657943505363,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046909924750060893,
      "loss": 5.1887,
      "mean_token_accuracy": 0.20163364410400392,
      "num_tokens": 35403338.0,
      "step": 15860
    },
    {
      "entropy": 5.432959079742432,
      "epoch": 1.781827371258494,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004690789380477109,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18954213857650756,
      "num_tokens": 35415862.0,
      "step": 15865
    },
    {
      "entropy": 5.396900463104248,
      "epoch": 1.7823889481664514,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004690586224152641,
      "loss": 5.1882,
      "mean_token_accuracy": 0.19624126106500625,
      "num_tokens": 35426696.0,
      "step": 15870
    },
    {
      "entropy": 5.419425201416016,
      "epoch": 1.7829505250744089,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004690383006039156,
      "loss": 5.2412,
      "mean_token_accuracy": 0.1966450333595276,
      "num_tokens": 35438624.0,
      "step": 15875
    },
    {
      "entropy": 5.487663078308105,
      "epoch": 1.7835121019823665,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004690179726143123,
      "loss": 5.3604,
      "mean_token_accuracy": 0.19170273691415787,
      "num_tokens": 35449841.0,
      "step": 15880
    },
    {
      "entropy": 5.626764297485352,
      "epoch": 1.784073678890324,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004689976384471014,
      "loss": 5.4457,
      "mean_token_accuracy": 0.18559928238391876,
      "num_tokens": 35462081.0,
      "step": 15885
    },
    {
      "entropy": 5.399773502349854,
      "epoch": 1.7846352557982814,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004689772981029306,
      "loss": 5.1498,
      "mean_token_accuracy": 0.20385131239891052,
      "num_tokens": 35472834.0,
      "step": 15890
    },
    {
      "entropy": 5.374467897415161,
      "epoch": 1.785196832706239,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004689569515824473,
      "loss": 5.1058,
      "mean_token_accuracy": 0.20000360161066055,
      "num_tokens": 35482975.0,
      "step": 15895
    },
    {
      "entropy": 5.321172904968262,
      "epoch": 1.7857584096141967,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004689365988862995,
      "loss": 5.2246,
      "mean_token_accuracy": 0.1946693480014801,
      "num_tokens": 35494470.0,
      "step": 15900
    },
    {
      "entropy": 5.345484781265259,
      "epoch": 1.7863199865221542,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046891624001513515,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19519539624452592,
      "num_tokens": 35505464.0,
      "step": 15905
    },
    {
      "entropy": 5.373464345932007,
      "epoch": 1.7868815634301116,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046889587496960257,
      "loss": 5.2349,
      "mean_token_accuracy": 0.19590731114149093,
      "num_tokens": 35517609.0,
      "step": 15910
    },
    {
      "entropy": 5.416655254364014,
      "epoch": 1.7874431403380693,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046887550375034997,
      "loss": 5.131,
      "mean_token_accuracy": 0.20046807080507278,
      "num_tokens": 35528994.0,
      "step": 15915
    },
    {
      "entropy": 5.454743146896362,
      "epoch": 1.788004717246027,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046885512635802627,
      "loss": 5.326,
      "mean_token_accuracy": 0.18871060162782669,
      "num_tokens": 35541044.0,
      "step": 15920
    },
    {
      "entropy": 5.381872320175171,
      "epoch": 1.7885662941539844,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004688347427932801,
      "loss": 5.153,
      "mean_token_accuracy": 0.1989303395152092,
      "num_tokens": 35552660.0,
      "step": 15925
    },
    {
      "entropy": 5.473754739761352,
      "epoch": 1.7891278710619418,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046881435305676055,
      "loss": 5.2502,
      "mean_token_accuracy": 0.19069281816482545,
      "num_tokens": 35565144.0,
      "step": 15930
    },
    {
      "entropy": 5.464682102203369,
      "epoch": 1.7896894479698995,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046879395714911686,
      "loss": 5.2159,
      "mean_token_accuracy": 0.19815842658281327,
      "num_tokens": 35577067.0,
      "step": 15935
    },
    {
      "entropy": 5.372298288345337,
      "epoch": 1.7902510248778571,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046877355507099846,
      "loss": 5.1995,
      "mean_token_accuracy": 0.19727577418088912,
      "num_tokens": 35588712.0,
      "step": 15940
    },
    {
      "entropy": 5.427581834793091,
      "epoch": 1.7908126017858146,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004687531468230549,
      "loss": 5.1538,
      "mean_token_accuracy": 0.19600952416658401,
      "num_tokens": 35599212.0,
      "step": 15945
    },
    {
      "entropy": 5.436887454986572,
      "epoch": 1.791374178693772,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004687327324059359,
      "loss": 5.198,
      "mean_token_accuracy": 0.1981391802430153,
      "num_tokens": 35609397.0,
      "step": 15950
    },
    {
      "entropy": 5.403420162200928,
      "epoch": 1.7919357556017297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004687123118202916,
      "loss": 5.2033,
      "mean_token_accuracy": 0.19730148017406463,
      "num_tokens": 35620615.0,
      "step": 15955
    },
    {
      "entropy": 5.4530126571655275,
      "epoch": 1.7924973325096873,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004686918850667722,
      "loss": 5.2346,
      "mean_token_accuracy": 0.19698902666568757,
      "num_tokens": 35632684.0,
      "step": 15960
    },
    {
      "entropy": 5.42054762840271,
      "epoch": 1.7930589094176448,
      "grad_norm": 1.2265625,
      "learning_rate": 0.000468671452146028,
      "loss": 5.2387,
      "mean_token_accuracy": 0.2018766760826111,
      "num_tokens": 35643896.0,
      "step": 15965
    },
    {
      "entropy": 5.442661619186401,
      "epoch": 1.7936204863256022,
      "grad_norm": 1.125,
      "learning_rate": 0.00046865101305870965,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19904266595840453,
      "num_tokens": 35655752.0,
      "step": 15970
    },
    {
      "entropy": 5.347767305374146,
      "epoch": 1.7941820632335599,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000468630567805468,
      "loss": 5.1172,
      "mean_token_accuracy": 0.19516800791025163,
      "num_tokens": 35665870.0,
      "step": 15975
    },
    {
      "entropy": 5.370148134231568,
      "epoch": 1.7947436401415173,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000468610116386954,
      "loss": 5.2878,
      "mean_token_accuracy": 0.19392186850309373,
      "num_tokens": 35677792.0,
      "step": 15980
    },
    {
      "entropy": 5.435721588134766,
      "epoch": 1.7953052170494748,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004685896588038187,
      "loss": 5.1888,
      "mean_token_accuracy": 0.19626725018024443,
      "num_tokens": 35689059.0,
      "step": 15985
    },
    {
      "entropy": 5.461448717117309,
      "epoch": 1.7958667939574324,
      "grad_norm": 1.125,
      "learning_rate": 0.0004685691950567136,
      "loss": 5.2684,
      "mean_token_accuracy": 0.19455415159463882,
      "num_tokens": 35699953.0,
      "step": 15990
    },
    {
      "entropy": 5.4490350723266605,
      "epoch": 1.79642837086539,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046854872514629035,
      "loss": 5.19,
      "mean_token_accuracy": 0.1966267004609108,
      "num_tokens": 35710894.0,
      "step": 15995
    },
    {
      "entropy": 5.362541675567627,
      "epoch": 1.7969899477733475,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004685282490732005,
      "loss": 5.2439,
      "mean_token_accuracy": 0.19514892548322677,
      "num_tokens": 35722440.0,
      "step": 16000
    },
    {
      "entropy": 5.411489534378052,
      "epoch": 1.797551524681305,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046850776683809623,
      "loss": 5.2376,
      "mean_token_accuracy": 0.20047878324985505,
      "num_tokens": 35733768.0,
      "step": 16005
    },
    {
      "entropy": 5.407706832885742,
      "epoch": 1.7981131015892626,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004684872784416295,
      "loss": 5.2088,
      "mean_token_accuracy": 0.19769827276468277,
      "num_tokens": 35744211.0,
      "step": 16010
    },
    {
      "entropy": 5.378367185592651,
      "epoch": 1.7986746784972203,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004684667838844529,
      "loss": 5.2224,
      "mean_token_accuracy": 0.19222935289144516,
      "num_tokens": 35755498.0,
      "step": 16015
    },
    {
      "entropy": 5.48088321685791,
      "epoch": 1.7992362554051777,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004684462831672188,
      "loss": 5.2387,
      "mean_token_accuracy": 0.20067128688097,
      "num_tokens": 35767442.0,
      "step": 16020
    },
    {
      "entropy": 5.390842151641846,
      "epoch": 1.7997978323131352,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046842577629058006,
      "loss": 5.1946,
      "mean_token_accuracy": 0.19865343421697618,
      "num_tokens": 35778040.0,
      "step": 16025
    },
    {
      "entropy": 5.334991598129273,
      "epoch": 1.8003594092210928,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004684052632551896,
      "loss": 5.1545,
      "mean_token_accuracy": 0.21034828275442125,
      "num_tokens": 35788139.0,
      "step": 16030
    },
    {
      "entropy": 5.357537508010864,
      "epoch": 1.8009209861290505,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046838474406170046,
      "loss": 5.2072,
      "mean_token_accuracy": 0.1967114180326462,
      "num_tokens": 35799077.0,
      "step": 16035
    },
    {
      "entropy": 5.45209813117981,
      "epoch": 1.801482563037008,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046836421871076616,
      "loss": 5.2341,
      "mean_token_accuracy": 0.1972969353199005,
      "num_tokens": 35809748.0,
      "step": 16040
    },
    {
      "entropy": 5.481454086303711,
      "epoch": 1.8020441399449654,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004683436872030401,
      "loss": 5.2686,
      "mean_token_accuracy": 0.19425856918096543,
      "num_tokens": 35821621.0,
      "step": 16045
    },
    {
      "entropy": 5.411822891235351,
      "epoch": 1.802605716852923,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046832314953917605,
      "loss": 5.2193,
      "mean_token_accuracy": 0.1981479972600937,
      "num_tokens": 35833497.0,
      "step": 16050
    },
    {
      "entropy": 5.526373815536499,
      "epoch": 1.8031672937608807,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046830260571982795,
      "loss": 5.3266,
      "mean_token_accuracy": 0.19589103758335114,
      "num_tokens": 35844266.0,
      "step": 16055
    },
    {
      "entropy": 5.5067931652069095,
      "epoch": 1.8037288706688381,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004682820557456499,
      "loss": 5.3237,
      "mean_token_accuracy": 0.19288819134235383,
      "num_tokens": 35855260.0,
      "step": 16060
    },
    {
      "entropy": 5.445611000061035,
      "epoch": 1.8042904475767956,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046826149961729625,
      "loss": 5.2537,
      "mean_token_accuracy": 0.19727153331041336,
      "num_tokens": 35865950.0,
      "step": 16065
    },
    {
      "entropy": 5.429213666915894,
      "epoch": 1.8048520244847532,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046824093733542144,
      "loss": 5.4191,
      "mean_token_accuracy": 0.18584830164909363,
      "num_tokens": 35877028.0,
      "step": 16070
    },
    {
      "entropy": 5.449163722991943,
      "epoch": 1.8054136013927107,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046822036890068033,
      "loss": 5.2563,
      "mean_token_accuracy": 0.19551195353269576,
      "num_tokens": 35887831.0,
      "step": 16075
    },
    {
      "entropy": 5.519545078277588,
      "epoch": 1.8059751783006681,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046819979431372773,
      "loss": 5.2315,
      "mean_token_accuracy": 0.1970939189195633,
      "num_tokens": 35897946.0,
      "step": 16080
    },
    {
      "entropy": 5.5011695384979244,
      "epoch": 1.8065367552086258,
      "grad_norm": 1.28125,
      "learning_rate": 0.00046817921357521865,
      "loss": 5.3654,
      "mean_token_accuracy": 0.19246957898139955,
      "num_tokens": 35910973.0,
      "step": 16085
    },
    {
      "entropy": 5.449217128753662,
      "epoch": 1.8070983321165834,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00046815862668580854,
      "loss": 5.3101,
      "mean_token_accuracy": 0.1929725542664528,
      "num_tokens": 35921478.0,
      "step": 16090
    },
    {
      "entropy": 5.43829607963562,
      "epoch": 1.8076599090245409,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004681380336461528,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20171307027339935,
      "num_tokens": 35932211.0,
      "step": 16095
    },
    {
      "entropy": 5.346833610534668,
      "epoch": 1.8082214859324983,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046811743445690724,
      "loss": 5.3429,
      "mean_token_accuracy": 0.1936837390065193,
      "num_tokens": 35944350.0,
      "step": 16100
    },
    {
      "entropy": 5.531755590438843,
      "epoch": 1.808783062840456,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004680968291187276,
      "loss": 5.3193,
      "mean_token_accuracy": 0.19187294095754623,
      "num_tokens": 35955590.0,
      "step": 16105
    },
    {
      "entropy": 5.414097785949707,
      "epoch": 1.8093446397484136,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004680762176322701,
      "loss": 5.0929,
      "mean_token_accuracy": 0.20915471613407136,
      "num_tokens": 35965872.0,
      "step": 16110
    },
    {
      "entropy": 5.3265704154968265,
      "epoch": 1.809906216656371,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046805559999819093,
      "loss": 5.2083,
      "mean_token_accuracy": 0.19323351234197617,
      "num_tokens": 35976851.0,
      "step": 16115
    },
    {
      "entropy": 5.428417348861695,
      "epoch": 1.8104677935643285,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004680349762171465,
      "loss": 5.1765,
      "mean_token_accuracy": 0.19633649587631224,
      "num_tokens": 35986984.0,
      "step": 16120
    },
    {
      "entropy": 5.439968013763428,
      "epoch": 1.8110293704722862,
      "grad_norm": 1.515625,
      "learning_rate": 0.0004680143462897936,
      "loss": 5.2413,
      "mean_token_accuracy": 0.1974828749895096,
      "num_tokens": 35999349.0,
      "step": 16125
    },
    {
      "entropy": 5.395186901092529,
      "epoch": 1.8115909473802438,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004679937102167891,
      "loss": 5.2621,
      "mean_token_accuracy": 0.19741955846548082,
      "num_tokens": 36011122.0,
      "step": 16130
    },
    {
      "entropy": 5.446416997909546,
      "epoch": 1.8121525242882013,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046797306799879,
      "loss": 5.1862,
      "mean_token_accuracy": 0.1961023762822151,
      "num_tokens": 36022729.0,
      "step": 16135
    },
    {
      "entropy": 5.400855541229248,
      "epoch": 1.8127141011961587,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046795241963645356,
      "loss": 5.1795,
      "mean_token_accuracy": 0.20251315087080002,
      "num_tokens": 36034042.0,
      "step": 16140
    },
    {
      "entropy": 5.383280420303345,
      "epoch": 1.8132756781041164,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004679317651304373,
      "loss": 5.2275,
      "mean_token_accuracy": 0.19339942932128906,
      "num_tokens": 36045357.0,
      "step": 16145
    },
    {
      "entropy": 5.297131681442261,
      "epoch": 1.813837255012074,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004679111044813987,
      "loss": 4.9895,
      "mean_token_accuracy": 0.20970333963632584,
      "num_tokens": 36055180.0,
      "step": 16150
    },
    {
      "entropy": 5.328434419631958,
      "epoch": 1.8143988319200315,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046789043768999573,
      "loss": 5.1448,
      "mean_token_accuracy": 0.19819593280553818,
      "num_tokens": 36065574.0,
      "step": 16155
    },
    {
      "entropy": 5.338675117492675,
      "epoch": 1.814960408827989,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004678697647568865,
      "loss": 5.224,
      "mean_token_accuracy": 0.19773971289396286,
      "num_tokens": 36077316.0,
      "step": 16160
    },
    {
      "entropy": 5.3838958740234375,
      "epoch": 1.8155219857359466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046784908568272903,
      "loss": 5.1959,
      "mean_token_accuracy": 0.19575451761484147,
      "num_tokens": 36088735.0,
      "step": 16165
    },
    {
      "entropy": 5.4344563484191895,
      "epoch": 1.8160835626439042,
      "grad_norm": 1.125,
      "learning_rate": 0.000467828400468182,
      "loss": 5.2375,
      "mean_token_accuracy": 0.19635221511125564,
      "num_tokens": 36098806.0,
      "step": 16170
    },
    {
      "entropy": 5.446918487548828,
      "epoch": 1.8166451395518615,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046780770911390387,
      "loss": 5.2676,
      "mean_token_accuracy": 0.19193698316812516,
      "num_tokens": 36110333.0,
      "step": 16175
    },
    {
      "entropy": 5.443539142608643,
      "epoch": 1.8172067164598191,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0004677870116205534,
      "loss": 5.2175,
      "mean_token_accuracy": 0.19319263547658921,
      "num_tokens": 36121719.0,
      "step": 16180
    },
    {
      "entropy": 5.397157335281372,
      "epoch": 1.8177682933677768,
      "grad_norm": 1.34375,
      "learning_rate": 0.00046776630798878985,
      "loss": 5.1299,
      "mean_token_accuracy": 0.20170477479696275,
      "num_tokens": 36132129.0,
      "step": 16185
    },
    {
      "entropy": 5.491793060302735,
      "epoch": 1.8183298702757342,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004677455982192723,
      "loss": 5.4243,
      "mean_token_accuracy": 0.1893936961889267,
      "num_tokens": 36145189.0,
      "step": 16190
    },
    {
      "entropy": 5.39800877571106,
      "epoch": 1.8188914471836917,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046772488231266004,
      "loss": 5.1679,
      "mean_token_accuracy": 0.19427197873592378,
      "num_tokens": 36157896.0,
      "step": 16195
    },
    {
      "entropy": 5.462921524047852,
      "epoch": 1.8194530240916493,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046770416026961284,
      "loss": 5.282,
      "mean_token_accuracy": 0.19151925593614577,
      "num_tokens": 36168842.0,
      "step": 16200
    },
    {
      "entropy": 5.343754625320434,
      "epoch": 1.820014600999607,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004676834320907904,
      "loss": 5.118,
      "mean_token_accuracy": 0.2015329271554947,
      "num_tokens": 36180075.0,
      "step": 16205
    },
    {
      "entropy": 5.408979320526123,
      "epoch": 1.8205761779075644,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004676626977768528,
      "loss": 5.1722,
      "mean_token_accuracy": 0.2047710195183754,
      "num_tokens": 36190589.0,
      "step": 16210
    },
    {
      "entropy": 5.305890989303589,
      "epoch": 1.8211377548155219,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046764195732846013,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2026120111346245,
      "num_tokens": 36201145.0,
      "step": 16215
    },
    {
      "entropy": 5.363481521606445,
      "epoch": 1.8216993317234795,
      "grad_norm": 1.125,
      "learning_rate": 0.00046762121074627285,
      "loss": 5.1957,
      "mean_token_accuracy": 0.19830980896949768,
      "num_tokens": 36213243.0,
      "step": 16220
    },
    {
      "entropy": 5.324520444869995,
      "epoch": 1.8222609086314372,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046760045803095153,
      "loss": 5.1662,
      "mean_token_accuracy": 0.20360076874494554,
      "num_tokens": 36224491.0,
      "step": 16225
    },
    {
      "entropy": 5.388520765304565,
      "epoch": 1.8228224855393946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004675796991831569,
      "loss": 5.2188,
      "mean_token_accuracy": 0.1954011857509613,
      "num_tokens": 36236006.0,
      "step": 16230
    },
    {
      "entropy": 5.5579915046691895,
      "epoch": 1.823384062447352,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046755893420355,
      "loss": 5.302,
      "mean_token_accuracy": 0.19135032892227172,
      "num_tokens": 36246925.0,
      "step": 16235
    },
    {
      "entropy": 5.448173904418946,
      "epoch": 1.8239456393553097,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004675381630927919,
      "loss": 5.2191,
      "mean_token_accuracy": 0.19228349775075912,
      "num_tokens": 36258449.0,
      "step": 16240
    },
    {
      "entropy": 5.431148290634155,
      "epoch": 1.8245072162632674,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046751738585154405,
      "loss": 5.232,
      "mean_token_accuracy": 0.19366778433322906,
      "num_tokens": 36270820.0,
      "step": 16245
    },
    {
      "entropy": 5.519060325622559,
      "epoch": 1.8250687931712248,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046749660248046794,
      "loss": 5.2935,
      "mean_token_accuracy": 0.19157375246286393,
      "num_tokens": 36282346.0,
      "step": 16250
    },
    {
      "entropy": 5.448248291015625,
      "epoch": 1.8256303700791823,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046747581298022537,
      "loss": 5.3134,
      "mean_token_accuracy": 0.18495482951402664,
      "num_tokens": 36294111.0,
      "step": 16255
    },
    {
      "entropy": 5.389816093444824,
      "epoch": 1.82619194698714,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004674550173514782,
      "loss": 5.1886,
      "mean_token_accuracy": 0.1975169986486435,
      "num_tokens": 36304899.0,
      "step": 16260
    },
    {
      "entropy": 5.462480306625366,
      "epoch": 1.8267535238950976,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046743421559488875,
      "loss": 5.3276,
      "mean_token_accuracy": 0.19281600117683412,
      "num_tokens": 36316272.0,
      "step": 16265
    },
    {
      "entropy": 5.4049155712127686,
      "epoch": 1.8273151008030548,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046741340771111917,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20524371415376663,
      "num_tokens": 36326624.0,
      "step": 16270
    },
    {
      "entropy": 5.350335550308228,
      "epoch": 1.8278766777110125,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046739259370083205,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20373691171407698,
      "num_tokens": 36338178.0,
      "step": 16275
    },
    {
      "entropy": 5.473524522781372,
      "epoch": 1.8284382546189701,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004673717735646902,
      "loss": 5.227,
      "mean_token_accuracy": 0.19841630309820174,
      "num_tokens": 36349993.0,
      "step": 16280
    },
    {
      "entropy": 5.4660194396972654,
      "epoch": 1.8289998315269276,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004673509473033564,
      "loss": 5.3026,
      "mean_token_accuracy": 0.1954042434692383,
      "num_tokens": 36361002.0,
      "step": 16285
    },
    {
      "entropy": 5.429429960250855,
      "epoch": 1.829561408434885,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046733011491749383,
      "loss": 5.2251,
      "mean_token_accuracy": 0.1964532494544983,
      "num_tokens": 36371688.0,
      "step": 16290
    },
    {
      "entropy": 5.387039279937744,
      "epoch": 1.8301229853428427,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004673092764077658,
      "loss": 5.1673,
      "mean_token_accuracy": 0.20498823672533034,
      "num_tokens": 36382683.0,
      "step": 16295
    },
    {
      "entropy": 5.52043833732605,
      "epoch": 1.8306845622508003,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004672884317748359,
      "loss": 5.3167,
      "mean_token_accuracy": 0.19046375751495362,
      "num_tokens": 36394170.0,
      "step": 16300
    },
    {
      "entropy": 5.475243091583252,
      "epoch": 1.8312461391587578,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004672675810193677,
      "loss": 5.2479,
      "mean_token_accuracy": 0.18889773339033128,
      "num_tokens": 36405613.0,
      "step": 16305
    },
    {
      "entropy": 5.452319765090943,
      "epoch": 1.8318077160667152,
      "grad_norm": 1.125,
      "learning_rate": 0.00046724672414202514,
      "loss": 5.2967,
      "mean_token_accuracy": 0.19847092032432556,
      "num_tokens": 36416282.0,
      "step": 16310
    },
    {
      "entropy": 5.4885488986969,
      "epoch": 1.8323692929746729,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004672258611434724,
      "loss": 5.3218,
      "mean_token_accuracy": 0.18818721622228624,
      "num_tokens": 36427624.0,
      "step": 16315
    },
    {
      "entropy": 5.367805671691895,
      "epoch": 1.8329308698826305,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046720499202437354,
      "loss": 5.097,
      "mean_token_accuracy": 0.2120461642742157,
      "num_tokens": 36438230.0,
      "step": 16320
    },
    {
      "entropy": 5.3712578296661375,
      "epoch": 1.833492446790588,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004671841167853933,
      "loss": 5.2479,
      "mean_token_accuracy": 0.1934431478381157,
      "num_tokens": 36448854.0,
      "step": 16325
    },
    {
      "entropy": 5.409632205963135,
      "epoch": 1.8340540236985454,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004671632354271962,
      "loss": 5.2392,
      "mean_token_accuracy": 0.19226710945367814,
      "num_tokens": 36460426.0,
      "step": 16330
    },
    {
      "entropy": 5.513731479644775,
      "epoch": 1.834615600606503,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046714234795044724,
      "loss": 5.2483,
      "mean_token_accuracy": 0.1921834632754326,
      "num_tokens": 36470914.0,
      "step": 16335
    },
    {
      "entropy": 5.411918687820434,
      "epoch": 1.8351771775144607,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004671214543558113,
      "loss": 5.1534,
      "mean_token_accuracy": 0.20135027915239334,
      "num_tokens": 36480753.0,
      "step": 16340
    },
    {
      "entropy": 5.3375931739807125,
      "epoch": 1.8357387544224182,
      "grad_norm": 1.25,
      "learning_rate": 0.0004671005546439538,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20335187762975693,
      "num_tokens": 36492858.0,
      "step": 16345
    },
    {
      "entropy": 5.309221267700195,
      "epoch": 1.8363003313303756,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046707964881554005,
      "loss": 5.0971,
      "mean_token_accuracy": 0.20305445194244384,
      "num_tokens": 36503912.0,
      "step": 16350
    },
    {
      "entropy": 5.3653724193573,
      "epoch": 1.8368619082383333,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046705873687123586,
      "loss": 5.1728,
      "mean_token_accuracy": 0.1952134370803833,
      "num_tokens": 36515410.0,
      "step": 16355
    },
    {
      "entropy": 5.43466968536377,
      "epoch": 1.837423485146291,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046703781881170703,
      "loss": 5.1379,
      "mean_token_accuracy": 0.2026953876018524,
      "num_tokens": 36526310.0,
      "step": 16360
    },
    {
      "entropy": 5.33543119430542,
      "epoch": 1.8379850620542482,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046701689463761947,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2013400301337242,
      "num_tokens": 36537081.0,
      "step": 16365
    },
    {
      "entropy": 5.320647716522217,
      "epoch": 1.8385466389622058,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046699596434963953,
      "loss": 5.1667,
      "mean_token_accuracy": 0.19930822551250457,
      "num_tokens": 36549419.0,
      "step": 16370
    },
    {
      "entropy": 5.317944478988648,
      "epoch": 1.8391082158701635,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004669750279484337,
      "loss": 5.1704,
      "mean_token_accuracy": 0.2060430243611336,
      "num_tokens": 36560842.0,
      "step": 16375
    },
    {
      "entropy": 5.378369617462158,
      "epoch": 1.839669792778121,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004669540854346685,
      "loss": 5.1556,
      "mean_token_accuracy": 0.20058241933584214,
      "num_tokens": 36571717.0,
      "step": 16380
    },
    {
      "entropy": 5.427543592453003,
      "epoch": 1.8402313696860784,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046693313680901073,
      "loss": 5.1618,
      "mean_token_accuracy": 0.1978237047791481,
      "num_tokens": 36581709.0,
      "step": 16385
    },
    {
      "entropy": 5.379416370391846,
      "epoch": 1.840792946594036,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046691218207212744,
      "loss": 5.2078,
      "mean_token_accuracy": 0.19435210675001144,
      "num_tokens": 36593033.0,
      "step": 16390
    },
    {
      "entropy": 5.449182081222534,
      "epoch": 1.8413545235019937,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004668912212246859,
      "loss": 5.286,
      "mean_token_accuracy": 0.19717612415552138,
      "num_tokens": 36603605.0,
      "step": 16395
    },
    {
      "entropy": 5.4512779712677,
      "epoch": 1.8419161004099511,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046687025426735335,
      "loss": 5.3112,
      "mean_token_accuracy": 0.19360125809907913,
      "num_tokens": 36616219.0,
      "step": 16400
    },
    {
      "entropy": 5.4823061466217045,
      "epoch": 1.8424776773179086,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004668492812007976,
      "loss": 5.2529,
      "mean_token_accuracy": 0.19652436077594757,
      "num_tokens": 36627777.0,
      "step": 16405
    },
    {
      "entropy": 5.447732639312744,
      "epoch": 1.8430392542258662,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004668283020256863,
      "loss": 5.2117,
      "mean_token_accuracy": 0.19509063810110092,
      "num_tokens": 36638792.0,
      "step": 16410
    },
    {
      "entropy": 5.411299753189087,
      "epoch": 1.843600831133824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046680731674268744,
      "loss": 5.2236,
      "mean_token_accuracy": 0.19851649850606917,
      "num_tokens": 36650218.0,
      "step": 16415
    },
    {
      "entropy": 5.419660758972168,
      "epoch": 1.8441624080417813,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004667863253524692,
      "loss": 5.203,
      "mean_token_accuracy": 0.19474465847015382,
      "num_tokens": 36660987.0,
      "step": 16420
    },
    {
      "entropy": 5.392714929580689,
      "epoch": 1.8447239849497388,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046676532785569995,
      "loss": 5.1661,
      "mean_token_accuracy": 0.19210840463638307,
      "num_tokens": 36671270.0,
      "step": 16425
    },
    {
      "entropy": 5.394438791275024,
      "epoch": 1.8452855618576964,
      "grad_norm": 1.4296875,
      "learning_rate": 0.0004667443242530484,
      "loss": 5.182,
      "mean_token_accuracy": 0.20304467976093293,
      "num_tokens": 36682484.0,
      "step": 16430
    },
    {
      "entropy": 5.437289762496948,
      "epoch": 1.845847138765654,
      "grad_norm": 1.125,
      "learning_rate": 0.00046672331454518314,
      "loss": 5.2756,
      "mean_token_accuracy": 0.2002399519085884,
      "num_tokens": 36693723.0,
      "step": 16435
    },
    {
      "entropy": 5.345190000534058,
      "epoch": 1.8464087156736115,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004667022987327732,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20894381105899812,
      "num_tokens": 36704166.0,
      "step": 16440
    },
    {
      "entropy": 5.362553596496582,
      "epoch": 1.846970292581569,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004666812768164877,
      "loss": 5.1363,
      "mean_token_accuracy": 0.19936540722846985,
      "num_tokens": 36714139.0,
      "step": 16445
    },
    {
      "entropy": 5.414290142059326,
      "epoch": 1.8475318694895266,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046666024879699606,
      "loss": 5.1613,
      "mean_token_accuracy": 0.20467677116394042,
      "num_tokens": 36724273.0,
      "step": 16450
    },
    {
      "entropy": 5.438724088668823,
      "epoch": 1.8480934463974843,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046663921467496766,
      "loss": 5.244,
      "mean_token_accuracy": 0.1896618753671646,
      "num_tokens": 36735176.0,
      "step": 16455
    },
    {
      "entropy": 5.346801519393921,
      "epoch": 1.8486550233054415,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004666181744510724,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20139126479625702,
      "num_tokens": 36746170.0,
      "step": 16460
    },
    {
      "entropy": 5.503177833557129,
      "epoch": 1.8492166002133992,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046659712812598017,
      "loss": 5.3389,
      "mean_token_accuracy": 0.1917860135436058,
      "num_tokens": 36756969.0,
      "step": 16465
    },
    {
      "entropy": 5.478408908843994,
      "epoch": 1.8497781771213568,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046657607570036103,
      "loss": 5.1857,
      "mean_token_accuracy": 0.2032455861568451,
      "num_tokens": 36767892.0,
      "step": 16470
    },
    {
      "entropy": 5.349558639526367,
      "epoch": 1.8503397540293143,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046655501717488533,
      "loss": 5.1838,
      "mean_token_accuracy": 0.20176382213830948,
      "num_tokens": 36778967.0,
      "step": 16475
    },
    {
      "entropy": 5.292702960968017,
      "epoch": 1.8509013309372717,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004665339525502236,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20440674871206282,
      "num_tokens": 36789204.0,
      "step": 16480
    },
    {
      "entropy": 5.384573984146118,
      "epoch": 1.8514629078452294,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046651288182704653,
      "loss": 5.1379,
      "mean_token_accuracy": 0.20209928303956987,
      "num_tokens": 36798990.0,
      "step": 16485
    },
    {
      "entropy": 5.478635978698731,
      "epoch": 1.852024484753187,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004664918050060251,
      "loss": 5.2516,
      "mean_token_accuracy": 0.19567494094371796,
      "num_tokens": 36809988.0,
      "step": 16490
    },
    {
      "entropy": 5.376132869720459,
      "epoch": 1.8525860616611445,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004664707220878302,
      "loss": 5.17,
      "mean_token_accuracy": 0.203729347884655,
      "num_tokens": 36821579.0,
      "step": 16495
    },
    {
      "entropy": 5.483326768875122,
      "epoch": 1.853147638569102,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004664496330731334,
      "loss": 5.2886,
      "mean_token_accuracy": 0.1986663594841957,
      "num_tokens": 36833141.0,
      "step": 16500
    },
    {
      "entropy": 5.422681713104248,
      "epoch": 1.8537092154770596,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004664285379626059,
      "loss": 5.302,
      "mean_token_accuracy": 0.1993295043706894,
      "num_tokens": 36843347.0,
      "step": 16505
    },
    {
      "entropy": 5.487172603607178,
      "epoch": 1.8542707923850172,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046640743675691954,
      "loss": 5.327,
      "mean_token_accuracy": 0.18590908348560334,
      "num_tokens": 36854459.0,
      "step": 16510
    },
    {
      "entropy": 5.381793642044068,
      "epoch": 1.8548323692929747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004663863294567462,
      "loss": 5.034,
      "mean_token_accuracy": 0.1997492402791977,
      "num_tokens": 36864761.0,
      "step": 16515
    },
    {
      "entropy": 5.513497447967529,
      "epoch": 1.8553939462009321,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004663652160627579,
      "loss": 5.3467,
      "mean_token_accuracy": 0.18831229954957962,
      "num_tokens": 36875413.0,
      "step": 16520
    },
    {
      "entropy": 5.403191375732422,
      "epoch": 1.8559555231088898,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046634409657562683,
      "loss": 5.2119,
      "mean_token_accuracy": 0.20070861130952836,
      "num_tokens": 36885350.0,
      "step": 16525
    },
    {
      "entropy": 5.40773491859436,
      "epoch": 1.8565171000168474,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004663229709960255,
      "loss": 5.1132,
      "mean_token_accuracy": 0.2031973570585251,
      "num_tokens": 36895355.0,
      "step": 16530
    },
    {
      "entropy": 5.4129698276519775,
      "epoch": 1.8570786769248049,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046630183932462664,
      "loss": 5.2141,
      "mean_token_accuracy": 0.20361922532320023,
      "num_tokens": 36907454.0,
      "step": 16535
    },
    {
      "entropy": 5.40431022644043,
      "epoch": 1.8576402538327623,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004662807015621031,
      "loss": 5.2254,
      "mean_token_accuracy": 0.19160055965185166,
      "num_tokens": 36919199.0,
      "step": 16540
    },
    {
      "entropy": 5.4786779403686525,
      "epoch": 1.85820183074072,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046625955770912764,
      "loss": 5.2994,
      "mean_token_accuracy": 0.19370331317186357,
      "num_tokens": 36930499.0,
      "step": 16545
    },
    {
      "entropy": 5.40087571144104,
      "epoch": 1.8587634076486776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046623840776637384,
      "loss": 5.2385,
      "mean_token_accuracy": 0.19194958359003067,
      "num_tokens": 36943082.0,
      "step": 16550
    },
    {
      "entropy": 5.338045120239258,
      "epoch": 1.859324984556635,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004662172517345149,
      "loss": 5.1414,
      "mean_token_accuracy": 0.20065291970968246,
      "num_tokens": 36954591.0,
      "step": 16555
    },
    {
      "entropy": 5.375982046127319,
      "epoch": 1.8598865614645925,
      "grad_norm": 1.125,
      "learning_rate": 0.0004661960896142245,
      "loss": 5.2415,
      "mean_token_accuracy": 0.1968989372253418,
      "num_tokens": 36967028.0,
      "step": 16560
    },
    {
      "entropy": 5.400326538085937,
      "epoch": 1.8604481383725502,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046617492140617654,
      "loss": 5.2561,
      "mean_token_accuracy": 0.19282017052173614,
      "num_tokens": 36978716.0,
      "step": 16565
    },
    {
      "entropy": 5.401040172576904,
      "epoch": 1.8610097152805076,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004661537471110448,
      "loss": 5.1401,
      "mean_token_accuracy": 0.20016276687383652,
      "num_tokens": 36989416.0,
      "step": 16570
    },
    {
      "entropy": 5.414360237121582,
      "epoch": 1.861571292188465,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004661325667295037,
      "loss": 5.1946,
      "mean_token_accuracy": 0.2035980686545372,
      "num_tokens": 37002197.0,
      "step": 16575
    },
    {
      "entropy": 5.52215895652771,
      "epoch": 1.8621328690964227,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004661113802622275,
      "loss": 5.309,
      "mean_token_accuracy": 0.18793494403362274,
      "num_tokens": 37012977.0,
      "step": 16580
    },
    {
      "entropy": 5.4043375015258786,
      "epoch": 1.8626944460043804,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004660901877098908,
      "loss": 5.2083,
      "mean_token_accuracy": 0.1987544357776642,
      "num_tokens": 37024141.0,
      "step": 16585
    },
    {
      "entropy": 5.346583032608033,
      "epoch": 1.8632560229123378,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004660689890731685,
      "loss": 5.1305,
      "mean_token_accuracy": 0.20212703943252563,
      "num_tokens": 37034471.0,
      "step": 16590
    },
    {
      "entropy": 5.38755612373352,
      "epoch": 1.8638175998202953,
      "grad_norm": 1.5703125,
      "learning_rate": 0.00046604778435273546,
      "loss": 5.2029,
      "mean_token_accuracy": 0.19776857644319534,
      "num_tokens": 37046375.0,
      "step": 16595
    },
    {
      "entropy": 5.435866022109986,
      "epoch": 1.864379176728253,
      "grad_norm": 1.6328125,
      "learning_rate": 0.00046602657354926686,
      "loss": 5.2395,
      "mean_token_accuracy": 0.20145386457443237,
      "num_tokens": 37056617.0,
      "step": 16600
    },
    {
      "entropy": 5.368656063079834,
      "epoch": 1.8649407536362106,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000466005356663438,
      "loss": 5.0541,
      "mean_token_accuracy": 0.2074560582637787,
      "num_tokens": 37066477.0,
      "step": 16605
    },
    {
      "entropy": 5.369606256484985,
      "epoch": 1.865502330544168,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046598413369592463,
      "loss": 5.2386,
      "mean_token_accuracy": 0.19350991398096085,
      "num_tokens": 37077150.0,
      "step": 16610
    },
    {
      "entropy": 5.442261409759522,
      "epoch": 1.8660639074521255,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004659629046474023,
      "loss": 5.2824,
      "mean_token_accuracy": 0.19620876610279084,
      "num_tokens": 37089961.0,
      "step": 16615
    },
    {
      "entropy": 5.472710704803466,
      "epoch": 1.8666254843600831,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000465941669518547,
      "loss": 5.2748,
      "mean_token_accuracy": 0.19373660236597062,
      "num_tokens": 37100779.0,
      "step": 16620
    },
    {
      "entropy": 5.407883501052856,
      "epoch": 1.8671870612680408,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004659204283100349,
      "loss": 5.1434,
      "mean_token_accuracy": 0.20432212501764296,
      "num_tokens": 37111293.0,
      "step": 16625
    },
    {
      "entropy": 5.4227265357971195,
      "epoch": 1.8677486381759982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046589918102254237,
      "loss": 5.2475,
      "mean_token_accuracy": 0.20140852481126786,
      "num_tokens": 37123087.0,
      "step": 16630
    },
    {
      "entropy": 5.372471714019776,
      "epoch": 1.8683102150839557,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004658779276567458,
      "loss": 5.2153,
      "mean_token_accuracy": 0.19963955730199814,
      "num_tokens": 37135227.0,
      "step": 16635
    },
    {
      "entropy": 5.3944511890411375,
      "epoch": 1.8688717919919133,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00046585666821332203,
      "loss": 5.1862,
      "mean_token_accuracy": 0.2015415459871292,
      "num_tokens": 37146970.0,
      "step": 16640
    },
    {
      "entropy": 5.4518050193786625,
      "epoch": 1.869433368899871,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046583540269294794,
      "loss": 5.1808,
      "mean_token_accuracy": 0.1937746822834015,
      "num_tokens": 37158144.0,
      "step": 16645
    },
    {
      "entropy": 5.444436979293823,
      "epoch": 1.8699949458078284,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004658141310963006,
      "loss": 5.3122,
      "mean_token_accuracy": 0.19287445694208144,
      "num_tokens": 37169265.0,
      "step": 16650
    },
    {
      "entropy": 5.461561822891236,
      "epoch": 1.8705565227157859,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004657928534240573,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1957921341061592,
      "num_tokens": 37179186.0,
      "step": 16655
    },
    {
      "entropy": 5.511565780639648,
      "epoch": 1.8711180996237435,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004657715696768955,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20423247665166855,
      "num_tokens": 37188906.0,
      "step": 16660
    },
    {
      "entropy": 5.3206470012664795,
      "epoch": 1.871679676531701,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000465750279855493,
      "loss": 5.1174,
      "mean_token_accuracy": 0.20478984266519545,
      "num_tokens": 37200879.0,
      "step": 16665
    },
    {
      "entropy": 5.378474617004395,
      "epoch": 1.8722412534396584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004657289839605276,
      "loss": 5.1279,
      "mean_token_accuracy": 0.2045281335711479,
      "num_tokens": 37211648.0,
      "step": 16670
    },
    {
      "entropy": 5.498213911056519,
      "epoch": 1.872802830347616,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004657076819926774,
      "loss": 5.3377,
      "mean_token_accuracy": 0.19406127482652663,
      "num_tokens": 37223164.0,
      "step": 16675
    },
    {
      "entropy": 5.414129161834717,
      "epoch": 1.8733644072555737,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004656863739526206,
      "loss": 5.1976,
      "mean_token_accuracy": 0.1987188756465912,
      "num_tokens": 37233860.0,
      "step": 16680
    },
    {
      "entropy": 5.491499280929565,
      "epoch": 1.8739259841635312,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046566505984103575,
      "loss": 5.151,
      "mean_token_accuracy": 0.19732064604759217,
      "num_tokens": 37245431.0,
      "step": 16685
    },
    {
      "entropy": 5.423554944992065,
      "epoch": 1.8744875610714886,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004656437396586014,
      "loss": 5.2946,
      "mean_token_accuracy": 0.1955429270863533,
      "num_tokens": 37256394.0,
      "step": 16690
    },
    {
      "entropy": 5.3890684127807615,
      "epoch": 1.8750491379794463,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004656224134059964,
      "loss": 5.2666,
      "mean_token_accuracy": 0.19405095279216766,
      "num_tokens": 37268367.0,
      "step": 16695
    },
    {
      "entropy": 5.4331347942352295,
      "epoch": 1.875610714887404,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046560108108389985,
      "loss": 5.2284,
      "mean_token_accuracy": 0.19831867814064025,
      "num_tokens": 37279019.0,
      "step": 16700
    },
    {
      "entropy": 5.510738325119019,
      "epoch": 1.8761722917953614,
      "grad_norm": 1.1953125,
      "learning_rate": 0.000465579742692991,
      "loss": 5.1867,
      "mean_token_accuracy": 0.1970874160528183,
      "num_tokens": 37289806.0,
      "step": 16705
    },
    {
      "entropy": 5.318840646743775,
      "epoch": 1.8767338687033188,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004655583982339492,
      "loss": 5.2084,
      "mean_token_accuracy": 0.19491557031869888,
      "num_tokens": 37300607.0,
      "step": 16710
    },
    {
      "entropy": 5.417339706420899,
      "epoch": 1.8772954456112765,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004655370477074541,
      "loss": 5.261,
      "mean_token_accuracy": 0.19406976103782653,
      "num_tokens": 37311966.0,
      "step": 16715
    },
    {
      "entropy": 5.341329336166382,
      "epoch": 1.8778570225192341,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004655156911141855,
      "loss": 5.1866,
      "mean_token_accuracy": 0.20028994977474213,
      "num_tokens": 37324459.0,
      "step": 16720
    },
    {
      "entropy": 5.443198490142822,
      "epoch": 1.8784185994271916,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004654943284548234,
      "loss": 5.188,
      "mean_token_accuracy": 0.19851246625185012,
      "num_tokens": 37335276.0,
      "step": 16725
    },
    {
      "entropy": 5.435481595993042,
      "epoch": 1.878980176335149,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046547295973004807,
      "loss": 5.1227,
      "mean_token_accuracy": 0.20124844014644622,
      "num_tokens": 37347274.0,
      "step": 16730
    },
    {
      "entropy": 5.380018091201782,
      "epoch": 1.8795417532431067,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046545158494053966,
      "loss": 5.235,
      "mean_token_accuracy": 0.1896946534514427,
      "num_tokens": 37359789.0,
      "step": 16735
    },
    {
      "entropy": 5.414810371398926,
      "epoch": 1.8801033301510643,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046543020408697905,
      "loss": 5.1826,
      "mean_token_accuracy": 0.20242450684309005,
      "num_tokens": 37369851.0,
      "step": 16740
    },
    {
      "entropy": 5.522116231918335,
      "epoch": 1.8806649070590218,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046540881717004685,
      "loss": 5.3355,
      "mean_token_accuracy": 0.1905204564332962,
      "num_tokens": 37381008.0,
      "step": 16745
    },
    {
      "entropy": 5.395009899139405,
      "epoch": 1.8812264839669792,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004653874241904241,
      "loss": 5.194,
      "mean_token_accuracy": 0.19708289504051207,
      "num_tokens": 37392110.0,
      "step": 16750
    },
    {
      "entropy": 5.481320810317993,
      "epoch": 1.8817880608749369,
      "grad_norm": 1.125,
      "learning_rate": 0.0004653660251487919,
      "loss": 5.2169,
      "mean_token_accuracy": 0.2028021737933159,
      "num_tokens": 37402995.0,
      "step": 16755
    },
    {
      "entropy": 5.436915063858033,
      "epoch": 1.8823496377828943,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046534462004583154,
      "loss": 5.2065,
      "mean_token_accuracy": 0.20292257964611055,
      "num_tokens": 37415072.0,
      "step": 16760
    },
    {
      "entropy": 5.4732091426849365,
      "epoch": 1.8829112146908518,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004653232088822248,
      "loss": 5.3618,
      "mean_token_accuracy": 0.19079849123954773,
      "num_tokens": 37427375.0,
      "step": 16765
    },
    {
      "entropy": 5.3222729682922365,
      "epoch": 1.8834727915988094,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004653017916586532,
      "loss": 5.1413,
      "mean_token_accuracy": 0.20021533370018005,
      "num_tokens": 37436829.0,
      "step": 16770
    },
    {
      "entropy": 5.461451292037964,
      "epoch": 1.884034368506767,
      "grad_norm": 1.125,
      "learning_rate": 0.0004652803683757987,
      "loss": 5.2378,
      "mean_token_accuracy": 0.1894003853201866,
      "num_tokens": 37447875.0,
      "step": 16775
    },
    {
      "entropy": 5.355460834503174,
      "epoch": 1.8845959454147245,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004652589390343436,
      "loss": 5.1162,
      "mean_token_accuracy": 0.20405496209859847,
      "num_tokens": 37458408.0,
      "step": 16780
    },
    {
      "entropy": 5.373467922210693,
      "epoch": 1.885157522322682,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046523750363497,
      "loss": 5.1729,
      "mean_token_accuracy": 0.1977369636297226,
      "num_tokens": 37468437.0,
      "step": 16785
    },
    {
      "entropy": 5.540972948074341,
      "epoch": 1.8857190992306396,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004652160621783605,
      "loss": 5.273,
      "mean_token_accuracy": 0.20020643174648284,
      "num_tokens": 37479359.0,
      "step": 16790
    },
    {
      "entropy": 5.388713026046753,
      "epoch": 1.8862806761385973,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046519461466519785,
      "loss": 5.2025,
      "mean_token_accuracy": 0.20322736501693725,
      "num_tokens": 37491321.0,
      "step": 16795
    },
    {
      "entropy": 5.437159013748169,
      "epoch": 1.8868422530465547,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004651731610961649,
      "loss": 5.225,
      "mean_token_accuracy": 0.19454448968172072,
      "num_tokens": 37501828.0,
      "step": 16800
    },
    {
      "entropy": 5.55532021522522,
      "epoch": 1.8874038299545122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004651517014719446,
      "loss": 5.3446,
      "mean_token_accuracy": 0.19017530977725983,
      "num_tokens": 37514525.0,
      "step": 16805
    },
    {
      "entropy": 5.424213600158692,
      "epoch": 1.8879654068624698,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004651302357932206,
      "loss": 5.1253,
      "mean_token_accuracy": 0.19491153508424758,
      "num_tokens": 37525588.0,
      "step": 16810
    },
    {
      "entropy": 5.365662574768066,
      "epoch": 1.8885269837704275,
      "grad_norm": 1.1796875,
      "learning_rate": 0.000465108764060676,
      "loss": 5.1889,
      "mean_token_accuracy": 0.19527479112148285,
      "num_tokens": 37537941.0,
      "step": 16815
    },
    {
      "entropy": 5.433779859542847,
      "epoch": 1.889088560678385,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046508728627499464,
      "loss": 5.2212,
      "mean_token_accuracy": 0.19074545055627823,
      "num_tokens": 37549663.0,
      "step": 16820
    },
    {
      "entropy": 5.4865823745727536,
      "epoch": 1.8896501375863424,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004650658024368604,
      "loss": 5.2821,
      "mean_token_accuracy": 0.19766873270273208,
      "num_tokens": 37559793.0,
      "step": 16825
    },
    {
      "entropy": 5.39378662109375,
      "epoch": 1.8902117144943,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004650443125469573,
      "loss": 5.1389,
      "mean_token_accuracy": 0.1975836381316185,
      "num_tokens": 37570203.0,
      "step": 16830
    },
    {
      "entropy": 5.349493026733398,
      "epoch": 1.8907732914022577,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004650228166059695,
      "loss": 5.2314,
      "mean_token_accuracy": 0.19433011561632157,
      "num_tokens": 37581858.0,
      "step": 16835
    },
    {
      "entropy": 5.4001881122589115,
      "epoch": 1.8913348683102151,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004650013146145816,
      "loss": 5.1723,
      "mean_token_accuracy": 0.19971663802862166,
      "num_tokens": 37592311.0,
      "step": 16840
    },
    {
      "entropy": 5.516441679000854,
      "epoch": 1.8918964452181726,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004649798065734781,
      "loss": 5.3253,
      "mean_token_accuracy": 0.1901743397116661,
      "num_tokens": 37604721.0,
      "step": 16845
    },
    {
      "entropy": 5.530319356918335,
      "epoch": 1.8924580221261302,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046495829248334393,
      "loss": 5.4187,
      "mean_token_accuracy": 0.1902414932847023,
      "num_tokens": 37616044.0,
      "step": 16850
    },
    {
      "entropy": 5.429763746261597,
      "epoch": 1.8930195990340877,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004649367723448641,
      "loss": 5.2712,
      "mean_token_accuracy": 0.1990755096077919,
      "num_tokens": 37626501.0,
      "step": 16855
    },
    {
      "entropy": 5.4841306686401365,
      "epoch": 1.8935811759420451,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046491524615872366,
      "loss": 5.2341,
      "mean_token_accuracy": 0.1968056306242943,
      "num_tokens": 37637861.0,
      "step": 16860
    },
    {
      "entropy": 5.354612064361572,
      "epoch": 1.8941427528500028,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046489371392560816,
      "loss": 5.0979,
      "mean_token_accuracy": 0.20413348227739334,
      "num_tokens": 37649288.0,
      "step": 16865
    },
    {
      "entropy": 5.319094657897949,
      "epoch": 1.8947043297579604,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046487217564620317,
      "loss": 5.1096,
      "mean_token_accuracy": 0.20416274964809417,
      "num_tokens": 37659700.0,
      "step": 16870
    },
    {
      "entropy": 5.399422311782837,
      "epoch": 1.8952659066659179,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004648506313211945,
      "loss": 5.2687,
      "mean_token_accuracy": 0.1933228462934494,
      "num_tokens": 37671739.0,
      "step": 16875
    },
    {
      "entropy": 5.444066429138184,
      "epoch": 1.8958274835738753,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046482908095126803,
      "loss": 5.1499,
      "mean_token_accuracy": 0.21089623868465424,
      "num_tokens": 37682250.0,
      "step": 16880
    },
    {
      "entropy": 5.392901420593262,
      "epoch": 1.896389060481833,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046480752453711,
      "loss": 5.2132,
      "mean_token_accuracy": 0.20136189460754395,
      "num_tokens": 37694052.0,
      "step": 16885
    },
    {
      "entropy": 5.497559261322022,
      "epoch": 1.8969506373897906,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004647859620794068,
      "loss": 5.2838,
      "mean_token_accuracy": 0.19946751445531846,
      "num_tokens": 37704818.0,
      "step": 16890
    },
    {
      "entropy": 5.560021591186524,
      "epoch": 1.897512214297748,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046476439357884497,
      "loss": 5.3474,
      "mean_token_accuracy": 0.18836625069379806,
      "num_tokens": 37717679.0,
      "step": 16895
    },
    {
      "entropy": 5.380611228942871,
      "epoch": 1.8980737912057055,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046474281903611126,
      "loss": 5.1625,
      "mean_token_accuracy": 0.2029123291373253,
      "num_tokens": 37728711.0,
      "step": 16900
    },
    {
      "entropy": 5.40647029876709,
      "epoch": 1.8986353681136632,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046472123845189253,
      "loss": 5.1947,
      "mean_token_accuracy": 0.19565970152616502,
      "num_tokens": 37740082.0,
      "step": 16905
    },
    {
      "entropy": 5.488197374343872,
      "epoch": 1.8991969450216208,
      "grad_norm": 1.4296875,
      "learning_rate": 0.000464699651826876,
      "loss": 5.3079,
      "mean_token_accuracy": 0.19070151150226594,
      "num_tokens": 37750715.0,
      "step": 16910
    },
    {
      "entropy": 5.4412421703338625,
      "epoch": 1.8997585219295783,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046467805916174907,
      "loss": 5.2801,
      "mean_token_accuracy": 0.18811442703008652,
      "num_tokens": 37763230.0,
      "step": 16915
    },
    {
      "entropy": 5.4420081615448,
      "epoch": 1.9003200988375357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046465646045719915,
      "loss": 5.1568,
      "mean_token_accuracy": 0.19407744258642196,
      "num_tokens": 37773621.0,
      "step": 16920
    },
    {
      "entropy": 5.438135099411011,
      "epoch": 1.9008816757454934,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004646348557139139,
      "loss": 5.2312,
      "mean_token_accuracy": 0.19549254029989244,
      "num_tokens": 37785068.0,
      "step": 16925
    },
    {
      "entropy": 5.419169569015503,
      "epoch": 1.901443252653451,
      "grad_norm": 1.9140625,
      "learning_rate": 0.00046461324493258133,
      "loss": 5.2437,
      "mean_token_accuracy": 0.19779633730649948,
      "num_tokens": 37796925.0,
      "step": 16930
    },
    {
      "entropy": 5.3959566116333,
      "epoch": 1.9020048295614085,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046459162811388955,
      "loss": 5.2494,
      "mean_token_accuracy": 0.19175779372453688,
      "num_tokens": 37808291.0,
      "step": 16935
    },
    {
      "entropy": 5.331517219543457,
      "epoch": 1.902566406469366,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004645700052585267,
      "loss": 5.0925,
      "mean_token_accuracy": 0.2124789223074913,
      "num_tokens": 37819337.0,
      "step": 16940
    },
    {
      "entropy": 5.433816146850586,
      "epoch": 1.9031279833773236,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004645483763671815,
      "loss": 5.2942,
      "mean_token_accuracy": 0.1935405507683754,
      "num_tokens": 37831832.0,
      "step": 16945
    },
    {
      "entropy": 5.433073949813843,
      "epoch": 1.903689560285281,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046452674144054236,
      "loss": 5.1916,
      "mean_token_accuracy": 0.19832090586423873,
      "num_tokens": 37842586.0,
      "step": 16950
    },
    {
      "entropy": 5.386325168609619,
      "epoch": 1.9042511371932385,
      "grad_norm": 1.234375,
      "learning_rate": 0.00046450510047929826,
      "loss": 5.1308,
      "mean_token_accuracy": 0.2030602440237999,
      "num_tokens": 37852198.0,
      "step": 16955
    },
    {
      "entropy": 5.409019184112549,
      "epoch": 1.9048127141011961,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00046448345348413833,
      "loss": 5.1831,
      "mean_token_accuracy": 0.19747235924005507,
      "num_tokens": 37862967.0,
      "step": 16960
    },
    {
      "entropy": 5.438821792602539,
      "epoch": 1.9053742910091538,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004644618004557517,
      "loss": 5.2898,
      "mean_token_accuracy": 0.18842438459396363,
      "num_tokens": 37874501.0,
      "step": 16965
    },
    {
      "entropy": 5.368289422988892,
      "epoch": 1.9059358679171112,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004644401413948279,
      "loss": 5.0575,
      "mean_token_accuracy": 0.20515141189098357,
      "num_tokens": 37886404.0,
      "step": 16970
    },
    {
      "entropy": 5.456409025192261,
      "epoch": 1.9064974448250687,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004644184763020565,
      "loss": 5.2653,
      "mean_token_accuracy": 0.19490793496370315,
      "num_tokens": 37897344.0,
      "step": 16975
    },
    {
      "entropy": 5.367180013656617,
      "epoch": 1.9070590217330263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004643968051781273,
      "loss": 5.1148,
      "mean_token_accuracy": 0.20911724418401717,
      "num_tokens": 37909226.0,
      "step": 16980
    },
    {
      "entropy": 5.426237010955811,
      "epoch": 1.907620598640984,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004643751280237305,
      "loss": 5.2539,
      "mean_token_accuracy": 0.20015060007572175,
      "num_tokens": 37920578.0,
      "step": 16985
    },
    {
      "entropy": 5.447788715362549,
      "epoch": 1.9081821755489414,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004643534448395561,
      "loss": 5.152,
      "mean_token_accuracy": 0.19414484351873398,
      "num_tokens": 37931814.0,
      "step": 16990
    },
    {
      "entropy": 5.37963604927063,
      "epoch": 1.9087437524568989,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004643317556262946,
      "loss": 5.191,
      "mean_token_accuracy": 0.1984264522790909,
      "num_tokens": 37942345.0,
      "step": 16995
    },
    {
      "entropy": 5.3659200191497805,
      "epoch": 1.9093053293648565,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046431006038463653,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20038702487945556,
      "num_tokens": 37953027.0,
      "step": 17000
    },
    {
      "entropy": 5.430207681655884,
      "epoch": 1.9098669062728142,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004642883591152727,
      "loss": 5.2556,
      "mean_token_accuracy": 0.19772350490093232,
      "num_tokens": 37963290.0,
      "step": 17005
    },
    {
      "entropy": 5.483037948608398,
      "epoch": 1.9104284831807716,
      "grad_norm": 1.3515625,
      "learning_rate": 0.0004642666518188942,
      "loss": 5.3366,
      "mean_token_accuracy": 0.18888563215732573,
      "num_tokens": 37974424.0,
      "step": 17010
    },
    {
      "entropy": 5.518302965164184,
      "epoch": 1.910990060088729,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046424493849619205,
      "loss": 5.3373,
      "mean_token_accuracy": 0.19278511703014373,
      "num_tokens": 37987993.0,
      "step": 17015
    },
    {
      "entropy": 5.390953922271729,
      "epoch": 1.9115516369966867,
      "grad_norm": 1.4296875,
      "learning_rate": 0.00046422321914785757,
      "loss": 5.1215,
      "mean_token_accuracy": 0.20266088098287582,
      "num_tokens": 37999340.0,
      "step": 17020
    },
    {
      "entropy": 5.377292728424072,
      "epoch": 1.9121132139046444,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046420149377458255,
      "loss": 5.1764,
      "mean_token_accuracy": 0.19506048262119294,
      "num_tokens": 38009737.0,
      "step": 17025
    },
    {
      "entropy": 5.427719306945801,
      "epoch": 1.9126747908126018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046417976237705846,
      "loss": 5.1715,
      "mean_token_accuracy": 0.20062230974435807,
      "num_tokens": 38021283.0,
      "step": 17030
    },
    {
      "entropy": 5.503713083267212,
      "epoch": 1.9132363677205593,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046415802495597744,
      "loss": 5.3194,
      "mean_token_accuracy": 0.19133986979722978,
      "num_tokens": 38034044.0,
      "step": 17035
    },
    {
      "entropy": 5.380450296401977,
      "epoch": 1.913797944628517,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046413628151203154,
      "loss": 5.1667,
      "mean_token_accuracy": 0.20291347205638885,
      "num_tokens": 38044585.0,
      "step": 17040
    },
    {
      "entropy": 5.308675956726074,
      "epoch": 1.9143595215364744,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046411453204591303,
      "loss": 5.1153,
      "mean_token_accuracy": 0.1982468530535698,
      "num_tokens": 38055771.0,
      "step": 17045
    },
    {
      "entropy": 5.372910022735596,
      "epoch": 1.9149210984444318,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046409277655831455,
      "loss": 5.207,
      "mean_token_accuracy": 0.19540495723485946,
      "num_tokens": 38066344.0,
      "step": 17050
    },
    {
      "entropy": 5.450403022766113,
      "epoch": 1.9154826753523895,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004640710150499287,
      "loss": 5.145,
      "mean_token_accuracy": 0.20422522276639937,
      "num_tokens": 38077785.0,
      "step": 17055
    },
    {
      "entropy": 5.38496470451355,
      "epoch": 1.9160442522603471,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046404924752144834,
      "loss": 5.2535,
      "mean_token_accuracy": 0.19590739607810975,
      "num_tokens": 38089023.0,
      "step": 17060
    },
    {
      "entropy": 5.384890604019165,
      "epoch": 1.9166058291683046,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046402747397356666,
      "loss": 5.0878,
      "mean_token_accuracy": 0.20458854883909225,
      "num_tokens": 38099428.0,
      "step": 17065
    },
    {
      "entropy": 5.32054967880249,
      "epoch": 1.917167406076262,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004640056944069768,
      "loss": 5.1392,
      "mean_token_accuracy": 0.19842455238103868,
      "num_tokens": 38110757.0,
      "step": 17070
    },
    {
      "entropy": 5.404375886917114,
      "epoch": 1.9177289829842197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004639839088223725,
      "loss": 5.2106,
      "mean_token_accuracy": 0.19488186985254288,
      "num_tokens": 38121442.0,
      "step": 17075
    },
    {
      "entropy": 5.415630197525024,
      "epoch": 1.9182905598921773,
      "grad_norm": 1.2421875,
      "learning_rate": 0.00046396211722044716,
      "loss": 5.1934,
      "mean_token_accuracy": 0.19628413915634155,
      "num_tokens": 38132917.0,
      "step": 17080
    },
    {
      "entropy": 5.311100578308105,
      "epoch": 1.9188521368001348,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046394031960189473,
      "loss": 5.1199,
      "mean_token_accuracy": 0.19692344665527345,
      "num_tokens": 38143188.0,
      "step": 17085
    },
    {
      "entropy": 5.466765832901001,
      "epoch": 1.9194137137080922,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004639185159674093,
      "loss": 5.3062,
      "mean_token_accuracy": 0.19594579488039016,
      "num_tokens": 38154300.0,
      "step": 17090
    },
    {
      "entropy": 5.470481491088867,
      "epoch": 1.9199752906160499,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046389670631768494,
      "loss": 5.2864,
      "mean_token_accuracy": 0.19543361365795137,
      "num_tokens": 38165697.0,
      "step": 17095
    },
    {
      "entropy": 5.439120054244995,
      "epoch": 1.9205368675240075,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004638748906534162,
      "loss": 5.2018,
      "mean_token_accuracy": 0.20053154081106186,
      "num_tokens": 38176601.0,
      "step": 17100
    },
    {
      "entropy": 5.47201099395752,
      "epoch": 1.921098444431965,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004638530689752977,
      "loss": 5.2824,
      "mean_token_accuracy": 0.2009994864463806,
      "num_tokens": 38188334.0,
      "step": 17105
    },
    {
      "entropy": 5.4503021240234375,
      "epoch": 1.9216600213399224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00046383124128402433,
      "loss": 5.2256,
      "mean_token_accuracy": 0.1980147510766983,
      "num_tokens": 38199483.0,
      "step": 17110
    },
    {
      "entropy": 5.436794185638428,
      "epoch": 1.92222159824788,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046380940758029095,
      "loss": 5.2973,
      "mean_token_accuracy": 0.19464406818151475,
      "num_tokens": 38211391.0,
      "step": 17115
    },
    {
      "entropy": 5.446100664138794,
      "epoch": 1.9227831751558377,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00046378756786479274,
      "loss": 5.191,
      "mean_token_accuracy": 0.19831823706626892,
      "num_tokens": 38223415.0,
      "step": 17120
    },
    {
      "entropy": 5.36629285812378,
      "epoch": 1.9233447520637952,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004637657221382252,
      "loss": 5.153,
      "mean_token_accuracy": 0.20241188406944274,
      "num_tokens": 38235093.0,
      "step": 17125
    },
    {
      "entropy": 5.407430028915405,
      "epoch": 1.9239063289717526,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046374387040128384,
      "loss": 5.2741,
      "mean_token_accuracy": 0.190486279129982,
      "num_tokens": 38247192.0,
      "step": 17130
    },
    {
      "entropy": 5.481328344345092,
      "epoch": 1.9244679058797103,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046372201265466444,
      "loss": 5.2547,
      "mean_token_accuracy": 0.19432902783155442,
      "num_tokens": 38257965.0,
      "step": 17135
    },
    {
      "entropy": 5.511951351165772,
      "epoch": 1.9250294827876677,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000463700148899063,
      "loss": 5.3319,
      "mean_token_accuracy": 0.1925298050045967,
      "num_tokens": 38268529.0,
      "step": 17140
    },
    {
      "entropy": 5.347607183456421,
      "epoch": 1.9255910596956252,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004636782791351756,
      "loss": 5.2053,
      "mean_token_accuracy": 0.19116217494010926,
      "num_tokens": 38279480.0,
      "step": 17145
    },
    {
      "entropy": 5.4459727764129635,
      "epoch": 1.9261526366035828,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004636564033636987,
      "loss": 5.266,
      "mean_token_accuracy": 0.19454890489578247,
      "num_tokens": 38291670.0,
      "step": 17150
    },
    {
      "entropy": 5.397248125076294,
      "epoch": 1.9267142135115405,
      "grad_norm": 1.125,
      "learning_rate": 0.00046363452158532857,
      "loss": 5.0571,
      "mean_token_accuracy": 0.21293195933103562,
      "num_tokens": 38303073.0,
      "step": 17155
    },
    {
      "entropy": 5.372452068328857,
      "epoch": 1.927275790419498,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004636126338007622,
      "loss": 5.2215,
      "mean_token_accuracy": 0.19101249128580094,
      "num_tokens": 38313915.0,
      "step": 17160
    },
    {
      "entropy": 5.362490606307984,
      "epoch": 1.9278373673274554,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004635907400106965,
      "loss": 5.1791,
      "mean_token_accuracy": 0.19504186064004897,
      "num_tokens": 38324845.0,
      "step": 17165
    },
    {
      "entropy": 5.433828639984131,
      "epoch": 1.928398944235413,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004635688402158285,
      "loss": 5.2032,
      "mean_token_accuracy": 0.19939032942056656,
      "num_tokens": 38335096.0,
      "step": 17170
    },
    {
      "entropy": 5.421277475357056,
      "epoch": 1.9289605211433707,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046354693441685537,
      "loss": 5.1837,
      "mean_token_accuracy": 0.20032843798398972,
      "num_tokens": 38345559.0,
      "step": 17175
    },
    {
      "entropy": 5.317073822021484,
      "epoch": 1.9295220980513281,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046352502261447475,
      "loss": 5.1112,
      "mean_token_accuracy": 0.212502221763134,
      "num_tokens": 38356318.0,
      "step": 17180
    },
    {
      "entropy": 5.338948488235474,
      "epoch": 1.9300836749592856,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004635031048093843,
      "loss": 5.1752,
      "mean_token_accuracy": 0.19731987863779069,
      "num_tokens": 38367246.0,
      "step": 17185
    },
    {
      "entropy": 5.3979713916778564,
      "epoch": 1.9306452518672432,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004634811810022819,
      "loss": 5.204,
      "mean_token_accuracy": 0.1964675322175026,
      "num_tokens": 38378767.0,
      "step": 17190
    },
    {
      "entropy": 5.454441738128662,
      "epoch": 1.931206828775201,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046345925119386556,
      "loss": 5.1507,
      "mean_token_accuracy": 0.20487340092658995,
      "num_tokens": 38390414.0,
      "step": 17195
    },
    {
      "entropy": 5.3831071853637695,
      "epoch": 1.9317684056831583,
      "grad_norm": 1.125,
      "learning_rate": 0.0004634373153848336,
      "loss": 5.1607,
      "mean_token_accuracy": 0.20038486272096634,
      "num_tokens": 38401569.0,
      "step": 17200
    },
    {
      "entropy": 5.431592178344727,
      "epoch": 1.9323299825911158,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004634153735758845,
      "loss": 5.2905,
      "mean_token_accuracy": 0.19502657651901245,
      "num_tokens": 38413561.0,
      "step": 17205
    },
    {
      "entropy": 5.431154775619507,
      "epoch": 1.9328915594990734,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004633934257677167,
      "loss": 5.2304,
      "mean_token_accuracy": 0.20406914949417115,
      "num_tokens": 38423965.0,
      "step": 17210
    },
    {
      "entropy": 5.37981276512146,
      "epoch": 1.933453136407031,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004633714719610292,
      "loss": 5.1391,
      "mean_token_accuracy": 0.20097376108169557,
      "num_tokens": 38434158.0,
      "step": 17215
    },
    {
      "entropy": 5.3468122482299805,
      "epoch": 1.9340147133149885,
      "grad_norm": 1.125,
      "learning_rate": 0.00046334951215652097,
      "loss": 5.1797,
      "mean_token_accuracy": 0.20062789618968963,
      "num_tokens": 38445449.0,
      "step": 17220
    },
    {
      "entropy": 5.45343861579895,
      "epoch": 1.934576290222946,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046332754635489126,
      "loss": 5.2818,
      "mean_token_accuracy": 0.1988828495144844,
      "num_tokens": 38457601.0,
      "step": 17225
    },
    {
      "entropy": 5.421912717819214,
      "epoch": 1.9351378671309036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004633055745568394,
      "loss": 5.1721,
      "mean_token_accuracy": 0.20002431571483612,
      "num_tokens": 38468631.0,
      "step": 17230
    },
    {
      "entropy": 5.333572435379028,
      "epoch": 1.935699444038861,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046328359676306506,
      "loss": 5.2158,
      "mean_token_accuracy": 0.2017907351255417,
      "num_tokens": 38479567.0,
      "step": 17235
    },
    {
      "entropy": 5.357053756713867,
      "epoch": 1.9362610209468185,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000463261612974268,
      "loss": 5.1739,
      "mean_token_accuracy": 0.2001163214445114,
      "num_tokens": 38489934.0,
      "step": 17240
    },
    {
      "entropy": 5.523341989517212,
      "epoch": 1.9368225978547762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004632396231911481,
      "loss": 5.3599,
      "mean_token_accuracy": 0.1907963290810585,
      "num_tokens": 38502214.0,
      "step": 17245
    },
    {
      "entropy": 5.411765670776367,
      "epoch": 1.9373841747627338,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004632176274144056,
      "loss": 5.1807,
      "mean_token_accuracy": 0.19869184195995332,
      "num_tokens": 38513915.0,
      "step": 17250
    },
    {
      "entropy": 5.38779935836792,
      "epoch": 1.9379457516706913,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046319562564474083,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20419521778821945,
      "num_tokens": 38525039.0,
      "step": 17255
    },
    {
      "entropy": 5.359553384780884,
      "epoch": 1.9385073285786487,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046317361788285436,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20672918260097503,
      "num_tokens": 38537264.0,
      "step": 17260
    },
    {
      "entropy": 5.4351897716522215,
      "epoch": 1.9390689054866064,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004631516041294469,
      "loss": 5.2375,
      "mean_token_accuracy": 0.20133037865161896,
      "num_tokens": 38547926.0,
      "step": 17265
    },
    {
      "entropy": 5.385799360275269,
      "epoch": 1.939630482394564,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004631295843852194,
      "loss": 5.2088,
      "mean_token_accuracy": 0.19948341995477675,
      "num_tokens": 38558745.0,
      "step": 17270
    },
    {
      "entropy": 5.418465662002563,
      "epoch": 1.9401920593025215,
      "grad_norm": 1.203125,
      "learning_rate": 0.000463107558650873,
      "loss": 5.1552,
      "mean_token_accuracy": 0.20313900113105773,
      "num_tokens": 38568190.0,
      "step": 17275
    },
    {
      "entropy": 5.479466152191162,
      "epoch": 1.940753636210479,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046308552692710895,
      "loss": 5.2578,
      "mean_token_accuracy": 0.1956784322857857,
      "num_tokens": 38578932.0,
      "step": 17280
    },
    {
      "entropy": 5.407718563079834,
      "epoch": 1.9413152131184366,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046306348921462876,
      "loss": 5.2142,
      "mean_token_accuracy": 0.19778369069099427,
      "num_tokens": 38590501.0,
      "step": 17285
    },
    {
      "entropy": 5.396119689941406,
      "epoch": 1.9418767900263942,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004630414455141342,
      "loss": 5.1194,
      "mean_token_accuracy": 0.20853587687015535,
      "num_tokens": 38600498.0,
      "step": 17290
    },
    {
      "entropy": 5.355353927612304,
      "epoch": 1.9424383669343517,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000463019395826327,
      "loss": 5.1355,
      "mean_token_accuracy": 0.19830868542194366,
      "num_tokens": 38611991.0,
      "step": 17295
    },
    {
      "entropy": 5.314302825927735,
      "epoch": 1.9429999438423091,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046299734015190934,
      "loss": 5.1562,
      "mean_token_accuracy": 0.20204446017742156,
      "num_tokens": 38622284.0,
      "step": 17300
    },
    {
      "entropy": 5.417598009109497,
      "epoch": 1.9435615207502668,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046297527849158344,
      "loss": 5.2519,
      "mean_token_accuracy": 0.19245603084564208,
      "num_tokens": 38633777.0,
      "step": 17305
    },
    {
      "entropy": 5.49086594581604,
      "epoch": 1.9441230976582244,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004629532108460519,
      "loss": 5.216,
      "mean_token_accuracy": 0.19272633492946625,
      "num_tokens": 38644360.0,
      "step": 17310
    },
    {
      "entropy": 5.371915626525879,
      "epoch": 1.9446846745661819,
      "grad_norm": 1.125,
      "learning_rate": 0.00046293113721601706,
      "loss": 5.2244,
      "mean_token_accuracy": 0.199888913333416,
      "num_tokens": 38655242.0,
      "step": 17315
    },
    {
      "entropy": 5.360963535308838,
      "epoch": 1.9452462514741393,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046290905760218193,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20644943416118622,
      "num_tokens": 38667792.0,
      "step": 17320
    },
    {
      "entropy": 5.3178328514099125,
      "epoch": 1.945807828382097,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046288697200524964,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20916850566864015,
      "num_tokens": 38677735.0,
      "step": 17325
    },
    {
      "entropy": 5.406602430343628,
      "epoch": 1.9463694052900544,
      "grad_norm": 1.015625,
      "learning_rate": 0.00046286488042592325,
      "loss": 5.1883,
      "mean_token_accuracy": 0.19762401580810546,
      "num_tokens": 38689556.0,
      "step": 17330
    },
    {
      "entropy": 5.344314289093018,
      "epoch": 1.9469309821980119,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004628427828649062,
      "loss": 5.1784,
      "mean_token_accuracy": 0.203419092297554,
      "num_tokens": 38700556.0,
      "step": 17335
    },
    {
      "entropy": 5.388719844818115,
      "epoch": 1.9474925591059695,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046282067932290204,
      "loss": 5.1611,
      "mean_token_accuracy": 0.19951771199703217,
      "num_tokens": 38711019.0,
      "step": 17340
    },
    {
      "entropy": 5.417884778976441,
      "epoch": 1.9480541360139272,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004627985698006146,
      "loss": 5.1813,
      "mean_token_accuracy": 0.19998494982719422,
      "num_tokens": 38722228.0,
      "step": 17345
    },
    {
      "entropy": 5.345585346221924,
      "epoch": 1.9486157129218846,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000462776454298748,
      "loss": 5.1088,
      "mean_token_accuracy": 0.20702529549598694,
      "num_tokens": 38732183.0,
      "step": 17350
    },
    {
      "entropy": 5.278182601928711,
      "epoch": 1.949177289829842,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046275433281800613,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20451745986938477,
      "num_tokens": 38743659.0,
      "step": 17355
    },
    {
      "entropy": 5.444448709487915,
      "epoch": 1.9497388667377997,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004627322053590935,
      "loss": 5.2493,
      "mean_token_accuracy": 0.19802791625261307,
      "num_tokens": 38755032.0,
      "step": 17360
    },
    {
      "entropy": 5.422481870651245,
      "epoch": 1.9503004436457574,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004627100719227147,
      "loss": 5.1518,
      "mean_token_accuracy": 0.2032081589102745,
      "num_tokens": 38766126.0,
      "step": 17365
    },
    {
      "entropy": 5.456494092941284,
      "epoch": 1.9508620205537148,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046268793250957437,
      "loss": 5.2982,
      "mean_token_accuracy": 0.1951813280582428,
      "num_tokens": 38776921.0,
      "step": 17370
    },
    {
      "entropy": 5.358982753753662,
      "epoch": 1.9514235974616723,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046266578712037747,
      "loss": 5.1546,
      "mean_token_accuracy": 0.20297640711069107,
      "num_tokens": 38788884.0,
      "step": 17375
    },
    {
      "entropy": 5.465472459793091,
      "epoch": 1.95198517436963,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046264363575582916,
      "loss": 5.336,
      "mean_token_accuracy": 0.18892173320055008,
      "num_tokens": 38800505.0,
      "step": 17380
    },
    {
      "entropy": 5.429839324951172,
      "epoch": 1.9525467512775876,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004626214784166347,
      "loss": 5.11,
      "mean_token_accuracy": 0.19924159049987794,
      "num_tokens": 38811132.0,
      "step": 17385
    },
    {
      "entropy": 5.433992052078247,
      "epoch": 1.953108328185545,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004625993151034995,
      "loss": 5.2135,
      "mean_token_accuracy": 0.20528505593538285,
      "num_tokens": 38821349.0,
      "step": 17390
    },
    {
      "entropy": 5.4114098072052,
      "epoch": 1.9536699050935025,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004625771458171294,
      "loss": 5.2357,
      "mean_token_accuracy": 0.1972706362605095,
      "num_tokens": 38832970.0,
      "step": 17395
    },
    {
      "entropy": 5.381783103942871,
      "epoch": 1.9542314820014601,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046255497055823025,
      "loss": 5.1896,
      "mean_token_accuracy": 0.20002585649490356,
      "num_tokens": 38844220.0,
      "step": 17400
    },
    {
      "entropy": 5.340125513076782,
      "epoch": 1.9547930589094178,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004625327893275082,
      "loss": 5.1694,
      "mean_token_accuracy": 0.19232664704322816,
      "num_tokens": 38854848.0,
      "step": 17405
    },
    {
      "entropy": 5.436822366714478,
      "epoch": 1.9553546358173752,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046251060212566927,
      "loss": 5.2973,
      "mean_token_accuracy": 0.1868386447429657,
      "num_tokens": 38864827.0,
      "step": 17410
    },
    {
      "entropy": 5.432698488235474,
      "epoch": 1.9559162127253327,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004624884089534201,
      "loss": 5.1538,
      "mean_token_accuracy": 0.19674716293811798,
      "num_tokens": 38875914.0,
      "step": 17415
    },
    {
      "entropy": 5.345088100433349,
      "epoch": 1.9564777896332903,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004624662098114672,
      "loss": 5.1189,
      "mean_token_accuracy": 0.19966707229614258,
      "num_tokens": 38886565.0,
      "step": 17420
    },
    {
      "entropy": 5.493402528762817,
      "epoch": 1.9570393665412478,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046244400470051754,
      "loss": 5.3763,
      "mean_token_accuracy": 0.18935890644788742,
      "num_tokens": 38897319.0,
      "step": 17425
    },
    {
      "entropy": 5.440078210830689,
      "epoch": 1.9576009434492052,
      "grad_norm": 1.140625,
      "learning_rate": 0.000462421793621278,
      "loss": 5.149,
      "mean_token_accuracy": 0.2018355533480644,
      "num_tokens": 38908824.0,
      "step": 17430
    },
    {
      "entropy": 5.370447111129761,
      "epoch": 1.9581625203571629,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000462399576574456,
      "loss": 5.2297,
      "mean_token_accuracy": 0.1998927965760231,
      "num_tokens": 38921800.0,
      "step": 17435
    },
    {
      "entropy": 5.385359859466552,
      "epoch": 1.9587240972651205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004623773535607587,
      "loss": 5.1844,
      "mean_token_accuracy": 0.19819723814725876,
      "num_tokens": 38932284.0,
      "step": 17440
    },
    {
      "entropy": 5.386744737625122,
      "epoch": 1.959285674173078,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004623551245808938,
      "loss": 5.1893,
      "mean_token_accuracy": 0.19746399521827698,
      "num_tokens": 38943632.0,
      "step": 17445
    },
    {
      "entropy": 5.406161737442017,
      "epoch": 1.9598472510810354,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004623328896355691,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20212339907884597,
      "num_tokens": 38954701.0,
      "step": 17450
    },
    {
      "entropy": 5.498396682739258,
      "epoch": 1.960408827988993,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046231064872549244,
      "loss": 5.2268,
      "mean_token_accuracy": 0.19593794792890548,
      "num_tokens": 38965625.0,
      "step": 17455
    },
    {
      "entropy": 5.3784421443939205,
      "epoch": 1.9609704048969507,
      "grad_norm": 1.03125,
      "learning_rate": 0.00046228840185137215,
      "loss": 5.1829,
      "mean_token_accuracy": 0.199091400206089,
      "num_tokens": 38976989.0,
      "step": 17460
    },
    {
      "entropy": 5.344245386123657,
      "epoch": 1.9615319818049082,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004622661490139165,
      "loss": 5.2207,
      "mean_token_accuracy": 0.20736266523599625,
      "num_tokens": 38987661.0,
      "step": 17465
    },
    {
      "entropy": 5.3767828941345215,
      "epoch": 1.9620935587128656,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004622438902138341,
      "loss": 5.1921,
      "mean_token_accuracy": 0.20350518226623535,
      "num_tokens": 38998014.0,
      "step": 17470
    },
    {
      "entropy": 5.35342116355896,
      "epoch": 1.9626551356208233,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046222162545183357,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20728033930063247,
      "num_tokens": 39008843.0,
      "step": 17475
    },
    {
      "entropy": 5.398371315002441,
      "epoch": 1.963216712528781,
      "grad_norm": 1.265625,
      "learning_rate": 0.00046219935472862386,
      "loss": 5.255,
      "mean_token_accuracy": 0.19220672100782393,
      "num_tokens": 39020173.0,
      "step": 17480
    },
    {
      "entropy": 5.450340270996094,
      "epoch": 1.9637782894367384,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00046217707804491404,
      "loss": 5.2676,
      "mean_token_accuracy": 0.18816238790750503,
      "num_tokens": 39033087.0,
      "step": 17485
    },
    {
      "entropy": 5.526058149337769,
      "epoch": 1.9643398663446958,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004621547954014135,
      "loss": 5.2099,
      "mean_token_accuracy": 0.20282547175884247,
      "num_tokens": 39045252.0,
      "step": 17490
    },
    {
      "entropy": 5.4031744480133055,
      "epoch": 1.9649014432526535,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004621325067988317,
      "loss": 5.3409,
      "mean_token_accuracy": 0.19465623646974564,
      "num_tokens": 39056755.0,
      "step": 17495
    },
    {
      "entropy": 5.352791500091553,
      "epoch": 1.9654630201606111,
      "grad_norm": 1.078125,
      "learning_rate": 0.00046211021223787827,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20625587105751036,
      "num_tokens": 39066993.0,
      "step": 17500
    },
    {
      "entropy": 5.427899932861328,
      "epoch": 1.9660245970685686,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046208791171926315,
      "loss": 5.1997,
      "mean_token_accuracy": 0.20257920324802398,
      "num_tokens": 39078671.0,
      "step": 17505
    },
    {
      "entropy": 5.394921541213989,
      "epoch": 1.966586173976526,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046206560524369636,
      "loss": 5.1917,
      "mean_token_accuracy": 0.19805443435907363,
      "num_tokens": 39089563.0,
      "step": 17510
    },
    {
      "entropy": 5.444635581970215,
      "epoch": 1.9671477508844837,
      "grad_norm": 1.140625,
      "learning_rate": 0.00046204329281188816,
      "loss": 5.2434,
      "mean_token_accuracy": 0.19480003118515016,
      "num_tokens": 39099837.0,
      "step": 17515
    },
    {
      "entropy": 5.483411645889282,
      "epoch": 1.9677093277924411,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004620209744245489,
      "loss": 5.2395,
      "mean_token_accuracy": 0.1955772340297699,
      "num_tokens": 39112061.0,
      "step": 17520
    },
    {
      "entropy": 5.391260814666748,
      "epoch": 1.9682709047003986,
      "grad_norm": 1.21875,
      "learning_rate": 0.00046199865008238937,
      "loss": 5.1874,
      "mean_token_accuracy": 0.20410646200180055,
      "num_tokens": 39122777.0,
      "step": 17525
    },
    {
      "entropy": 5.291329956054687,
      "epoch": 1.9688324816083562,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004619763197861203,
      "loss": 5.1402,
      "mean_token_accuracy": 0.20243419259786605,
      "num_tokens": 39134827.0,
      "step": 17530
    },
    {
      "entropy": 5.39892578125,
      "epoch": 1.9693940585163139,
      "grad_norm": 1.1875,
      "learning_rate": 0.00046195398353645256,
      "loss": 5.1759,
      "mean_token_accuracy": 0.20454369932413102,
      "num_tokens": 39144352.0,
      "step": 17535
    },
    {
      "entropy": 5.396446371078492,
      "epoch": 1.9699556354242713,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004619316413340976,
      "loss": 5.12,
      "mean_token_accuracy": 0.2012733682990074,
      "num_tokens": 39155610.0,
      "step": 17540
    },
    {
      "entropy": 5.309647274017334,
      "epoch": 1.9705172123322288,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00046190929317976664,
      "loss": 5.179,
      "mean_token_accuracy": 0.20145184695720672,
      "num_tokens": 39167058.0,
      "step": 17545
    },
    {
      "entropy": 5.433630418777466,
      "epoch": 1.9710787892401864,
      "grad_norm": 1.0,
      "learning_rate": 0.0004618869390741713,
      "loss": 5.2772,
      "mean_token_accuracy": 0.19650811851024627,
      "num_tokens": 39178797.0,
      "step": 17550
    },
    {
      "entropy": 5.424329137802124,
      "epoch": 1.971640366148144,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004618645790180233,
      "loss": 5.1341,
      "mean_token_accuracy": 0.20124213248491288,
      "num_tokens": 39188980.0,
      "step": 17555
    },
    {
      "entropy": 5.350632381439209,
      "epoch": 1.9722019430561015,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004618422130120347,
      "loss": 5.1369,
      "mean_token_accuracy": 0.19740630835294723,
      "num_tokens": 39200058.0,
      "step": 17560
    },
    {
      "entropy": 5.378911447525025,
      "epoch": 1.972763519964059,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004618198410569175,
      "loss": 5.211,
      "mean_token_accuracy": 0.19966285675764084,
      "num_tokens": 39211369.0,
      "step": 17565
    },
    {
      "entropy": 5.396366119384766,
      "epoch": 1.9733250968720166,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046179746315338414,
      "loss": 5.2563,
      "mean_token_accuracy": 0.19893464893102647,
      "num_tokens": 39222438.0,
      "step": 17570
    },
    {
      "entropy": 5.4548264503479,
      "epoch": 1.9738866737799743,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00046177507930214714,
      "loss": 5.2332,
      "mean_token_accuracy": 0.19236407577991485,
      "num_tokens": 39233047.0,
      "step": 17575
    },
    {
      "entropy": 5.430233335494995,
      "epoch": 1.9744482506879317,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004617526895039191,
      "loss": 5.1524,
      "mean_token_accuracy": 0.2003133103251457,
      "num_tokens": 39244547.0,
      "step": 17580
    },
    {
      "entropy": 5.351201963424683,
      "epoch": 1.9750098275958892,
      "grad_norm": 1.203125,
      "learning_rate": 0.00046173029375941305,
      "loss": 5.1103,
      "mean_token_accuracy": 0.19972345530986785,
      "num_tokens": 39255549.0,
      "step": 17585
    },
    {
      "entropy": 5.2948401927947994,
      "epoch": 1.9755714045038468,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000461707892069342,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20270635485649108,
      "num_tokens": 39266527.0,
      "step": 17590
    },
    {
      "entropy": 5.375566625595093,
      "epoch": 1.9761329814118045,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046168548443441926,
      "loss": 5.2167,
      "mean_token_accuracy": 0.20037177801132203,
      "num_tokens": 39278013.0,
      "step": 17595
    },
    {
      "entropy": 5.389305830001831,
      "epoch": 1.976694558319762,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004616630708553583,
      "loss": 5.1702,
      "mean_token_accuracy": 0.19923675656318665,
      "num_tokens": 39289349.0,
      "step": 17600
    },
    {
      "entropy": 5.361451482772827,
      "epoch": 1.9772561352277194,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046164065133287276,
      "loss": 5.1602,
      "mean_token_accuracy": 0.20103998482227325,
      "num_tokens": 39301213.0,
      "step": 17605
    },
    {
      "entropy": 5.33935580253601,
      "epoch": 1.977817712135677,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004616182258676765,
      "loss": 5.065,
      "mean_token_accuracy": 0.20336418449878693,
      "num_tokens": 39310960.0,
      "step": 17610
    },
    {
      "entropy": 5.4179588794708256,
      "epoch": 1.9783792890436345,
      "grad_norm": 1.5546875,
      "learning_rate": 0.0004615957944604835,
      "loss": 5.236,
      "mean_token_accuracy": 0.19934232383966446,
      "num_tokens": 39322603.0,
      "step": 17615
    },
    {
      "entropy": 5.4213399410247805,
      "epoch": 1.978940865951592,
      "grad_norm": 1.125,
      "learning_rate": 0.0004615733571120081,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20278631001710892,
      "num_tokens": 39333811.0,
      "step": 17620
    },
    {
      "entropy": 5.3991844177246096,
      "epoch": 1.9795024428595496,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004615509138229646,
      "loss": 5.203,
      "mean_token_accuracy": 0.1973966807126999,
      "num_tokens": 39345218.0,
      "step": 17625
    },
    {
      "entropy": 5.312092971801758,
      "epoch": 1.9800640197675072,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046152846459406765,
      "loss": 5.0362,
      "mean_token_accuracy": 0.20124004930257797,
      "num_tokens": 39356777.0,
      "step": 17630
    },
    {
      "entropy": 5.35697751045227,
      "epoch": 1.9806255966754647,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004615060094260321,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20007307678461075,
      "num_tokens": 39368288.0,
      "step": 17635
    },
    {
      "entropy": 5.431741094589233,
      "epoch": 1.9811871735834221,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046148354831957285,
      "loss": 5.1725,
      "mean_token_accuracy": 0.2054842382669449,
      "num_tokens": 39379398.0,
      "step": 17640
    },
    {
      "entropy": 5.303338575363159,
      "epoch": 1.9817487504913798,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046146108127540506,
      "loss": 5.1762,
      "mean_token_accuracy": 0.20564125329256058,
      "num_tokens": 39391129.0,
      "step": 17645
    },
    {
      "entropy": 5.426952219009399,
      "epoch": 1.9823103273993374,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004614386082942442,
      "loss": 5.288,
      "mean_token_accuracy": 0.18791701048612594,
      "num_tokens": 39402442.0,
      "step": 17650
    },
    {
      "entropy": 5.532159280776978,
      "epoch": 1.9828719043072949,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004614161293768057,
      "loss": 5.2148,
      "mean_token_accuracy": 0.19993433952331544,
      "num_tokens": 39413044.0,
      "step": 17655
    },
    {
      "entropy": 5.416688776016235,
      "epoch": 1.9834334812152523,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00046139364452380536,
      "loss": 5.1782,
      "mean_token_accuracy": 0.2016997903585434,
      "num_tokens": 39423739.0,
      "step": 17660
    },
    {
      "entropy": 5.410736751556397,
      "epoch": 1.98399505812321,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000461371153735959,
      "loss": 5.3112,
      "mean_token_accuracy": 0.19203440546989442,
      "num_tokens": 39436199.0,
      "step": 17665
    },
    {
      "entropy": 5.454751062393188,
      "epoch": 1.9845566350311676,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046134865701398287,
      "loss": 5.1929,
      "mean_token_accuracy": 0.2083341136574745,
      "num_tokens": 39449302.0,
      "step": 17670
    },
    {
      "entropy": 5.420571994781494,
      "epoch": 1.985118211939125,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00046132615435859326,
      "loss": 5.2465,
      "mean_token_accuracy": 0.19248813837766648,
      "num_tokens": 39460723.0,
      "step": 17675
    },
    {
      "entropy": 5.3681779384613035,
      "epoch": 1.9856797888470825,
      "grad_norm": 1.109375,
      "learning_rate": 0.00046130364577050655,
      "loss": 5.1505,
      "mean_token_accuracy": 0.20501491129398347,
      "num_tokens": 39471251.0,
      "step": 17680
    },
    {
      "entropy": 5.391210460662842,
      "epoch": 1.9862413657550402,
      "grad_norm": 1.046875,
      "learning_rate": 0.00046128113125043957,
      "loss": 5.1725,
      "mean_token_accuracy": 0.20591398775577546,
      "num_tokens": 39482613.0,
      "step": 17685
    },
    {
      "entropy": 5.390169143676758,
      "epoch": 1.9868029426629978,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046125861079910913,
      "loss": 5.1746,
      "mean_token_accuracy": 0.2132398158311844,
      "num_tokens": 39494626.0,
      "step": 17690
    },
    {
      "entropy": 5.347628784179688,
      "epoch": 1.9873645195709553,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004612360844172322,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20022881776094437,
      "num_tokens": 39505302.0,
      "step": 17695
    },
    {
      "entropy": 5.3758728981018065,
      "epoch": 1.9879260964789127,
      "grad_norm": 1.171875,
      "learning_rate": 0.00046121355210552626,
      "loss": 5.1011,
      "mean_token_accuracy": 0.20194316506385804,
      "num_tokens": 39515854.0,
      "step": 17700
    },
    {
      "entropy": 5.4112317085266115,
      "epoch": 1.9884876733868704,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004611910138647085,
      "loss": 5.2611,
      "mean_token_accuracy": 0.19290996789932252,
      "num_tokens": 39527504.0,
      "step": 17705
    },
    {
      "entropy": 5.424066734313965,
      "epoch": 1.989049250294828,
      "grad_norm": 1.0,
      "learning_rate": 0.00046116846969549666,
      "loss": 5.1711,
      "mean_token_accuracy": 0.19952830374240876,
      "num_tokens": 39538885.0,
      "step": 17710
    },
    {
      "entropy": 5.383217430114746,
      "epoch": 1.9896108272027853,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004611459195986085,
      "loss": 5.1471,
      "mean_token_accuracy": 0.20784952640533447,
      "num_tokens": 39549401.0,
      "step": 17715
    },
    {
      "entropy": 5.349120044708252,
      "epoch": 1.990172404110743,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00046112336357476217,
      "loss": 5.1962,
      "mean_token_accuracy": 0.19637592285871505,
      "num_tokens": 39560814.0,
      "step": 17720
    },
    {
      "entropy": 5.394489383697509,
      "epoch": 1.9907339810187006,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004611008016246757,
      "loss": 5.1774,
      "mean_token_accuracy": 0.19825324565172195,
      "num_tokens": 39571490.0,
      "step": 17725
    },
    {
      "entropy": 5.324645709991455,
      "epoch": 1.991295557926658,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00046107823374906754,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20534008443355561,
      "num_tokens": 39582815.0,
      "step": 17730
    },
    {
      "entropy": 5.382453584671021,
      "epoch": 1.9918571348346155,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00046105565994865614,
      "loss": 5.1838,
      "mean_token_accuracy": 0.19903729110956192,
      "num_tokens": 39593712.0,
      "step": 17735
    },
    {
      "entropy": 5.399639987945557,
      "epoch": 1.9924187117425731,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004610330802241605,
      "loss": 5.1699,
      "mean_token_accuracy": 0.2017083689570427,
      "num_tokens": 39604504.0,
      "step": 17740
    },
    {
      "entropy": 5.4152350425720215,
      "epoch": 1.9929802886505308,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004610104945762994,
      "loss": 5.1862,
      "mean_token_accuracy": 0.20468616783618926,
      "num_tokens": 39614971.0,
      "step": 17745
    },
    {
      "entropy": 5.37033486366272,
      "epoch": 1.9935418655584882,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000460987903005792,
      "loss": 5.1153,
      "mean_token_accuracy": 0.20931664407253264,
      "num_tokens": 39626647.0,
      "step": 17750
    },
    {
      "entropy": 5.4290016174316404,
      "epoch": 1.9941034424664457,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004609653055133576,
      "loss": 5.2488,
      "mean_token_accuracy": 0.19639914035797118,
      "num_tokens": 39636432.0,
      "step": 17755
    },
    {
      "entropy": 5.428470945358276,
      "epoch": 1.9946650193744033,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046094270209971576,
      "loss": 5.2075,
      "mean_token_accuracy": 0.19905703216791154,
      "num_tokens": 39647055.0,
      "step": 17760
    },
    {
      "entropy": 5.398600625991821,
      "epoch": 1.995226596282361,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004609200927655862,
      "loss": 5.1752,
      "mean_token_accuracy": 0.20248110294342042,
      "num_tokens": 39657891.0,
      "step": 17765
    },
    {
      "entropy": 5.284284782409668,
      "epoch": 1.9957881731903184,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004608974775116888,
      "loss": 5.0584,
      "mean_token_accuracy": 0.20982797741889953,
      "num_tokens": 39668023.0,
      "step": 17770
    },
    {
      "entropy": 5.39661226272583,
      "epoch": 1.9963497500982759,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004608748563387436,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20422269254922867,
      "num_tokens": 39678398.0,
      "step": 17775
    },
    {
      "entropy": 5.368674325942993,
      "epoch": 1.9969113270062335,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004608522292474708,
      "loss": 5.1802,
      "mean_token_accuracy": 0.197450190782547,
      "num_tokens": 39690619.0,
      "step": 17780
    },
    {
      "entropy": 5.28471417427063,
      "epoch": 1.9974729039141912,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004608295962385911,
      "loss": 5.0961,
      "mean_token_accuracy": 0.20618446469306945,
      "num_tokens": 39701690.0,
      "step": 17785
    },
    {
      "entropy": 5.372359991073608,
      "epoch": 1.9980344808221486,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00046080695731282484,
      "loss": 5.2454,
      "mean_token_accuracy": 0.1928040489554405,
      "num_tokens": 39713878.0,
      "step": 17790
    },
    {
      "entropy": 5.466627359390259,
      "epoch": 1.998596057730106,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000460784312470893,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19878810346126558,
      "num_tokens": 39725164.0,
      "step": 17795
    },
    {
      "entropy": 5.38539605140686,
      "epoch": 1.9991576346380637,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046076166171351673,
      "loss": 5.1248,
      "mean_token_accuracy": 0.19812840074300767,
      "num_tokens": 39735354.0,
      "step": 17800
    },
    {
      "entropy": 5.487515163421631,
      "epoch": 1.9997192115460214,
      "grad_norm": 1.046875,
      "learning_rate": 0.000460739005041417,
      "loss": 5.2615,
      "mean_token_accuracy": 0.19964937418699263,
      "num_tokens": 39746618.0,
      "step": 17805
    },
    {
      "entropy": 5.40585274166531,
      "epoch": 2.000224630763183,
      "grad_norm": 1.0,
      "learning_rate": 0.0004607163424553153,
      "loss": 5.2104,
      "mean_token_accuracy": 0.19742655257383981,
      "num_tokens": 39755697.0,
      "step": 17810
    },
    {
      "entropy": 5.27701096534729,
      "epoch": 2.0007862076711405,
      "grad_norm": 1.125,
      "learning_rate": 0.00046069367395593325,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20144225656986237,
      "num_tokens": 39766311.0,
      "step": 17815
    },
    {
      "entropy": 5.414301013946533,
      "epoch": 2.001347784579098,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004606709995439926,
      "loss": 5.207,
      "mean_token_accuracy": 0.19515834152698516,
      "num_tokens": 39777504.0,
      "step": 17820
    },
    {
      "entropy": 5.443689632415771,
      "epoch": 2.001909361487056,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00046064831922021534,
      "loss": 5.078,
      "mean_token_accuracy": 0.20121817886829377,
      "num_tokens": 39789190.0,
      "step": 17825
    },
    {
      "entropy": 5.365042304992675,
      "epoch": 2.002470938395013,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004606256329853235,
      "loss": 5.0698,
      "mean_token_accuracy": 0.2003874510526657,
      "num_tokens": 39801534.0,
      "step": 17830
    },
    {
      "entropy": 5.408135986328125,
      "epoch": 2.0030325153029707,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004606029408400396,
      "loss": 5.1472,
      "mean_token_accuracy": 0.19824248254299165,
      "num_tokens": 39812353.0,
      "step": 17835
    },
    {
      "entropy": 5.337577962875367,
      "epoch": 2.0035940922109283,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004605802427850861,
      "loss": 5.0457,
      "mean_token_accuracy": 0.20930266380310059,
      "num_tokens": 39823313.0,
      "step": 17840
    },
    {
      "entropy": 5.266083669662476,
      "epoch": 2.004155669118886,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046055753882118564,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21112132370471953,
      "num_tokens": 39833966.0,
      "step": 17845
    },
    {
      "entropy": 5.355870008468628,
      "epoch": 2.004717246026843,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004605348289490611,
      "loss": 5.1241,
      "mean_token_accuracy": 0.20649746656417847,
      "num_tokens": 39843963.0,
      "step": 17850
    },
    {
      "entropy": 5.441735792160034,
      "epoch": 2.005278822934801,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004605121131694358,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20493690818548202,
      "num_tokens": 39855237.0,
      "step": 17855
    },
    {
      "entropy": 5.417480850219727,
      "epoch": 2.0058403998427585,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004604893914830328,
      "loss": 5.2294,
      "mean_token_accuracy": 0.19613784551620483,
      "num_tokens": 39867270.0,
      "step": 17860
    },
    {
      "entropy": 5.450803279876709,
      "epoch": 2.006401976750716,
      "grad_norm": 1.25,
      "learning_rate": 0.00046046666389057563,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20359700471162795,
      "num_tokens": 39878103.0,
      "step": 17865
    },
    {
      "entropy": 5.345403432846069,
      "epoch": 2.0069635536586734,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000460443930392788,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20274411588907243,
      "num_tokens": 39888945.0,
      "step": 17870
    },
    {
      "entropy": 5.481605434417725,
      "epoch": 2.007525130566631,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00046042119099039367,
      "loss": 5.2676,
      "mean_token_accuracy": 0.19336591511964799,
      "num_tokens": 39899848.0,
      "step": 17875
    },
    {
      "entropy": 5.432954025268555,
      "epoch": 2.0080867074745887,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00046039844568411667,
      "loss": 5.1348,
      "mean_token_accuracy": 0.20050491988658906,
      "num_tokens": 39911824.0,
      "step": 17880
    },
    {
      "entropy": 5.306520605087281,
      "epoch": 2.0086482843825464,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004603756944746813,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20528721511363984,
      "num_tokens": 39923257.0,
      "step": 17885
    },
    {
      "entropy": 5.301839542388916,
      "epoch": 2.0092098612905036,
      "grad_norm": 1.09375,
      "learning_rate": 0.00046035293736281185,
      "loss": 5.0498,
      "mean_token_accuracy": 0.2077257066965103,
      "num_tokens": 39935312.0,
      "step": 17890
    },
    {
      "entropy": 5.313592386245728,
      "epoch": 2.0097714381984613,
      "grad_norm": 1.265625,
      "learning_rate": 0.0004603301743492331,
      "loss": 5.0739,
      "mean_token_accuracy": 0.2079662263393402,
      "num_tokens": 39946468.0,
      "step": 17895
    },
    {
      "entropy": 5.254870319366455,
      "epoch": 2.010333015106419,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004603074054346696,
      "loss": 4.9239,
      "mean_token_accuracy": 0.21702035367488862,
      "num_tokens": 39957517.0,
      "step": 17900
    },
    {
      "entropy": 5.266244840621948,
      "epoch": 2.0108945920143766,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004602846306198465,
      "loss": 5.0594,
      "mean_token_accuracy": 0.20753604173660278,
      "num_tokens": 39968385.0,
      "step": 17905
    },
    {
      "entropy": 5.346811199188233,
      "epoch": 2.011456168922334,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004602618499054889,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2050669953227043,
      "num_tokens": 39979393.0,
      "step": 17910
    },
    {
      "entropy": 5.470460510253906,
      "epoch": 2.0120177458302915,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004602390632923221,
      "loss": 5.0961,
      "mean_token_accuracy": 0.2004488468170166,
      "num_tokens": 39989709.0,
      "step": 17915
    },
    {
      "entropy": 5.3384950160980225,
      "epoch": 2.012579322738249,
      "grad_norm": 1.328125,
      "learning_rate": 0.00046021627078107177,
      "loss": 5.0415,
      "mean_token_accuracy": 0.19882239699363707,
      "num_tokens": 40000472.0,
      "step": 17920
    },
    {
      "entropy": 5.182809495925904,
      "epoch": 2.0131408996462063,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00046019347237246343,
      "loss": 4.9177,
      "mean_token_accuracy": 0.2127896025776863,
      "num_tokens": 40011062.0,
      "step": 17925
    },
    {
      "entropy": 5.399270009994507,
      "epoch": 2.013702476554164,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004601706680672232,
      "loss": 5.1835,
      "mean_token_accuracy": 0.193230938911438,
      "num_tokens": 40022430.0,
      "step": 17930
    },
    {
      "entropy": 5.3236083984375,
      "epoch": 2.0142640534621217,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004601478578660771,
      "loss": 5.0738,
      "mean_token_accuracy": 0.20267102271318435,
      "num_tokens": 40033077.0,
      "step": 17935
    },
    {
      "entropy": 5.401332378387451,
      "epoch": 2.0148256303700793,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004601250417697513,
      "loss": 5.0951,
      "mean_token_accuracy": 0.20817828923463821,
      "num_tokens": 40044258.0,
      "step": 17940
    },
    {
      "entropy": 5.366886138916016,
      "epoch": 2.0153872072780366,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004601022197789724,
      "loss": 5.117,
      "mean_token_accuracy": 0.196757972240448,
      "num_tokens": 40055876.0,
      "step": 17945
    },
    {
      "entropy": 5.296723318099976,
      "epoch": 2.015948784185994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004600793918944671,
      "loss": 5.0017,
      "mean_token_accuracy": 0.21432248651981353,
      "num_tokens": 40066691.0,
      "step": 17950
    },
    {
      "entropy": 5.293151569366455,
      "epoch": 2.016510361093952,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004600565581169621,
      "loss": 5.0025,
      "mean_token_accuracy": 0.2028060734272003,
      "num_tokens": 40077335.0,
      "step": 17955
    },
    {
      "entropy": 5.323266410827637,
      "epoch": 2.0170719380019095,
      "grad_norm": 1.0625,
      "learning_rate": 0.00046003371844718455,
      "loss": 5.1456,
      "mean_token_accuracy": 0.19561377465724944,
      "num_tokens": 40089053.0,
      "step": 17960
    },
    {
      "entropy": 5.364812850952148,
      "epoch": 2.0176335149098668,
      "grad_norm": 1.15625,
      "learning_rate": 0.00046001087288586164,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21388910561800004,
      "num_tokens": 40099371.0,
      "step": 17965
    },
    {
      "entropy": 5.394143438339233,
      "epoch": 2.0181950918178244,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004599880214337208,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20136368423700332,
      "num_tokens": 40110274.0,
      "step": 17970
    },
    {
      "entropy": 5.379863023757935,
      "epoch": 2.018756668725782,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004599651640914896,
      "loss": 5.1629,
      "mean_token_accuracy": 0.1986452594399452,
      "num_tokens": 40120364.0,
      "step": 17975
    },
    {
      "entropy": 5.328201150894165,
      "epoch": 2.0193182456337397,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004599423008598957,
      "loss": 5.1083,
      "mean_token_accuracy": 0.20367644876241683,
      "num_tokens": 40132400.0,
      "step": 17980
    },
    {
      "entropy": 5.40523328781128,
      "epoch": 2.019879822541697,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045991943173966744,
      "loss": 5.1514,
      "mean_token_accuracy": 0.19722880423069,
      "num_tokens": 40142820.0,
      "step": 17985
    },
    {
      "entropy": 5.415780782699585,
      "epoch": 2.0204413994496546,
      "grad_norm": 1.125,
      "learning_rate": 0.0004598965567315326,
      "loss": 5.1595,
      "mean_token_accuracy": 0.19911278784275055,
      "num_tokens": 40153620.0,
      "step": 17990
    },
    {
      "entropy": 5.35541296005249,
      "epoch": 2.0210029763576123,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004598736758362198,
      "loss": 5.0815,
      "mean_token_accuracy": 0.20393969416618346,
      "num_tokens": 40164250.0,
      "step": 17995
    },
    {
      "entropy": 5.331550407409668,
      "epoch": 2.02156455326557,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045985078905445734,
      "loss": 5.0304,
      "mean_token_accuracy": 0.2031165361404419,
      "num_tokens": 40175809.0,
      "step": 18000
    },
    {
      "epoch": 2.02156455326557,
      "eval_entropy": 5.2089060325120755,
      "eval_loss": 5.279140949249268,
      "eval_mean_token_accuracy": 0.20285280997033028,
      "eval_num_tokens": 40175809.0,
      "eval_runtime": 23.9172,
      "eval_samples_per_second": 1296.681,
      "eval_steps_per_second": 162.101,
      "step": 18000
    },
    {
      "entropy": 5.231017589569092,
      "epoch": 2.022126130173527,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045982789638697414,
      "loss": 4.9884,
      "mean_token_accuracy": 0.20771982669830322,
      "num_tokens": 40187855.0,
      "step": 18005
    },
    {
      "entropy": 5.323897790908814,
      "epoch": 2.022687707081485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045980499783449897,
      "loss": 5.05,
      "mean_token_accuracy": 0.20788765400648118,
      "num_tokens": 40198802.0,
      "step": 18010
    },
    {
      "entropy": 5.471941041946411,
      "epoch": 2.0232492839894425,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045978209339776104,
      "loss": 5.2491,
      "mean_token_accuracy": 0.19387288391590118,
      "num_tokens": 40210636.0,
      "step": 18015
    },
    {
      "entropy": 5.384461879730225,
      "epoch": 2.0238108608973997,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045975918307748965,
      "loss": 5.1238,
      "mean_token_accuracy": 0.2028246581554413,
      "num_tokens": 40221216.0,
      "step": 18020
    },
    {
      "entropy": 5.310945081710815,
      "epoch": 2.0243724378053574,
      "grad_norm": 1.5078125,
      "learning_rate": 0.0004597362668744141,
      "loss": 4.9813,
      "mean_token_accuracy": 0.21240756511688233,
      "num_tokens": 40231988.0,
      "step": 18025
    },
    {
      "entropy": 5.415923833847046,
      "epoch": 2.024934014713315,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004597133447892642,
      "loss": 5.2914,
      "mean_token_accuracy": 0.19551550596952438,
      "num_tokens": 40243789.0,
      "step": 18030
    },
    {
      "entropy": 5.3948770523071286,
      "epoch": 2.0254955916212727,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004596904168227698,
      "loss": 5.1045,
      "mean_token_accuracy": 0.20624535530805588,
      "num_tokens": 40254771.0,
      "step": 18035
    },
    {
      "entropy": 5.344694757461548,
      "epoch": 2.02605716852923,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045966748297566085,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20568991154432298,
      "num_tokens": 40265444.0,
      "step": 18040
    },
    {
      "entropy": 5.329793310165405,
      "epoch": 2.0266187454371876,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045964454324866767,
      "loss": 5.0966,
      "mean_token_accuracy": 0.19761816263198853,
      "num_tokens": 40277146.0,
      "step": 18045
    },
    {
      "entropy": 5.34061975479126,
      "epoch": 2.0271803223451452,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045962159764252055,
      "loss": 5.1557,
      "mean_token_accuracy": 0.2072695553302765,
      "num_tokens": 40286493.0,
      "step": 18050
    },
    {
      "entropy": 5.412021255493164,
      "epoch": 2.027741899253103,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004595986461579502,
      "loss": 5.1057,
      "mean_token_accuracy": 0.20423293858766556,
      "num_tokens": 40297209.0,
      "step": 18055
    },
    {
      "entropy": 5.32928204536438,
      "epoch": 2.02830347616106,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045957568879568736,
      "loss": 5.0659,
      "mean_token_accuracy": 0.2051330626010895,
      "num_tokens": 40308666.0,
      "step": 18060
    },
    {
      "entropy": 5.414534378051758,
      "epoch": 2.0288650530690178,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045955272555646306,
      "loss": 5.1802,
      "mean_token_accuracy": 0.19476450979709625,
      "num_tokens": 40321145.0,
      "step": 18065
    },
    {
      "entropy": 5.399529027938843,
      "epoch": 2.0294266299769754,
      "grad_norm": 1.390625,
      "learning_rate": 0.0004595297564410083,
      "loss": 5.1122,
      "mean_token_accuracy": 0.20350463837385177,
      "num_tokens": 40332418.0,
      "step": 18070
    },
    {
      "entropy": 5.278768014907837,
      "epoch": 2.029988206884933,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045950678145005456,
      "loss": 5.0682,
      "mean_token_accuracy": 0.20518109798431397,
      "num_tokens": 40342838.0,
      "step": 18075
    },
    {
      "entropy": 5.335705184936524,
      "epoch": 2.0305497837928903,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004594838005843333,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21170784533023834,
      "num_tokens": 40352940.0,
      "step": 18080
    },
    {
      "entropy": 5.44862551689148,
      "epoch": 2.031111360700848,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045946081384457633,
      "loss": 5.1432,
      "mean_token_accuracy": 0.19848093390464783,
      "num_tokens": 40364273.0,
      "step": 18085
    },
    {
      "entropy": 5.33009934425354,
      "epoch": 2.0316729376088056,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045943782123151556,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2162682056427002,
      "num_tokens": 40374617.0,
      "step": 18090
    },
    {
      "entropy": 5.3156637191772464,
      "epoch": 2.0322345145167633,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004594148227458829,
      "loss": 5.1446,
      "mean_token_accuracy": 0.20275897830724715,
      "num_tokens": 40385398.0,
      "step": 18095
    },
    {
      "entropy": 5.517345190048218,
      "epoch": 2.0327960914247205,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004593918183884108,
      "loss": 5.2289,
      "mean_token_accuracy": 0.19807093739509582,
      "num_tokens": 40397030.0,
      "step": 18100
    },
    {
      "entropy": 5.441907215118408,
      "epoch": 2.033357668332678,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004593688081598318,
      "loss": 5.1882,
      "mean_token_accuracy": 0.19929959923028945,
      "num_tokens": 40408027.0,
      "step": 18105
    },
    {
      "entropy": 5.401389932632446,
      "epoch": 2.033919245240636,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045934579206087837,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19932475835084915,
      "num_tokens": 40418755.0,
      "step": 18110
    },
    {
      "entropy": 5.445931339263916,
      "epoch": 2.034480822148593,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045932277009228336,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20645360946655272,
      "num_tokens": 40430694.0,
      "step": 18115
    },
    {
      "entropy": 5.411109733581543,
      "epoch": 2.0350423990565507,
      "grad_norm": 1.34375,
      "learning_rate": 0.00045929974225477986,
      "loss": 5.1127,
      "mean_token_accuracy": 0.20078421384096146,
      "num_tokens": 40442019.0,
      "step": 18120
    },
    {
      "entropy": 5.390072917938232,
      "epoch": 2.0356039759645084,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045927670854910117,
      "loss": 5.1199,
      "mean_token_accuracy": 0.19206283390522003,
      "num_tokens": 40454120.0,
      "step": 18125
    },
    {
      "entropy": 5.432745456695557,
      "epoch": 2.036165552872466,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004592536689759806,
      "loss": 5.2569,
      "mean_token_accuracy": 0.19687375724315642,
      "num_tokens": 40467445.0,
      "step": 18130
    },
    {
      "entropy": 5.345901441574097,
      "epoch": 2.0367271297804233,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045923062353615167,
      "loss": 5.0402,
      "mean_token_accuracy": 0.2021577164530754,
      "num_tokens": 40477512.0,
      "step": 18135
    },
    {
      "entropy": 5.266358041763306,
      "epoch": 2.037288706688381,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045920757223034833,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20638511776924134,
      "num_tokens": 40488209.0,
      "step": 18140
    },
    {
      "entropy": 5.440270566940308,
      "epoch": 2.0378502835963386,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004591845150593044,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20350110828876494,
      "num_tokens": 40499793.0,
      "step": 18145
    },
    {
      "entropy": 5.444608974456787,
      "epoch": 2.0384118605042962,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045916145202375407,
      "loss": 5.2119,
      "mean_token_accuracy": 0.19534004777669906,
      "num_tokens": 40510598.0,
      "step": 18150
    },
    {
      "entropy": 5.470892858505249,
      "epoch": 2.0389734374122535,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004591383831244317,
      "loss": 5.1489,
      "mean_token_accuracy": 0.20659288316965102,
      "num_tokens": 40521289.0,
      "step": 18155
    },
    {
      "entropy": 5.4251727104187015,
      "epoch": 2.039535014320211,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045911530836207173,
      "loss": 5.171,
      "mean_token_accuracy": 0.2007632151246071,
      "num_tokens": 40532048.0,
      "step": 18160
    },
    {
      "entropy": 5.376459169387817,
      "epoch": 2.0400965912281688,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045909222773740904,
      "loss": 5.1625,
      "mean_token_accuracy": 0.20368960201740266,
      "num_tokens": 40543699.0,
      "step": 18165
    },
    {
      "entropy": 5.482229900360108,
      "epoch": 2.0406581681361264,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045906914125117833,
      "loss": 5.2023,
      "mean_token_accuracy": 0.20113713890314103,
      "num_tokens": 40554626.0,
      "step": 18170
    },
    {
      "entropy": 5.429351139068603,
      "epoch": 2.0412197450440837,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045904604890411476,
      "loss": 5.1478,
      "mean_token_accuracy": 0.19965285956859588,
      "num_tokens": 40567046.0,
      "step": 18175
    },
    {
      "entropy": 5.514778137207031,
      "epoch": 2.0417813219520413,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045902295069695373,
      "loss": 5.2087,
      "mean_token_accuracy": 0.1921456053853035,
      "num_tokens": 40578472.0,
      "step": 18180
    },
    {
      "entropy": 5.407752656936646,
      "epoch": 2.042342898859999,
      "grad_norm": 1.5,
      "learning_rate": 0.0004589998466304305,
      "loss": 5.1264,
      "mean_token_accuracy": 0.2012196496129036,
      "num_tokens": 40589012.0,
      "step": 18185
    },
    {
      "entropy": 5.306635141372681,
      "epoch": 2.0429044757679566,
      "grad_norm": 1.03125,
      "learning_rate": 0.00045897673670528086,
      "loss": 5.0615,
      "mean_token_accuracy": 0.20474885255098343,
      "num_tokens": 40600807.0,
      "step": 18190
    },
    {
      "entropy": 5.433145427703858,
      "epoch": 2.043466052675914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004589536209222404,
      "loss": 5.2202,
      "mean_token_accuracy": 0.1994292303919792,
      "num_tokens": 40612578.0,
      "step": 18195
    },
    {
      "entropy": 5.361132526397705,
      "epoch": 2.0440276295838715,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004589304992820454,
      "loss": 4.9731,
      "mean_token_accuracy": 0.20831649601459504,
      "num_tokens": 40623504.0,
      "step": 18200
    },
    {
      "entropy": 5.384540271759033,
      "epoch": 2.044589206491829,
      "grad_norm": 1.140625,
      "learning_rate": 0.000458907371785432,
      "loss": 5.1375,
      "mean_token_accuracy": 0.20204037129879,
      "num_tokens": 40633676.0,
      "step": 18205
    },
    {
      "entropy": 5.4194262504577635,
      "epoch": 2.0451507833997864,
      "grad_norm": 1.234375,
      "learning_rate": 0.0004588842384331366,
      "loss": 5.2102,
      "mean_token_accuracy": 0.20009807795286177,
      "num_tokens": 40645991.0,
      "step": 18210
    },
    {
      "entropy": 5.3448771953582765,
      "epoch": 2.045712360307744,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045886109922589564,
      "loss": 5.027,
      "mean_token_accuracy": 0.2073945090174675,
      "num_tokens": 40656617.0,
      "step": 18215
    },
    {
      "entropy": 5.3701536655426025,
      "epoch": 2.0462739372157017,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045883795416444594,
      "loss": 5.1501,
      "mean_token_accuracy": 0.20787471085786818,
      "num_tokens": 40669708.0,
      "step": 18220
    },
    {
      "entropy": 5.392904472351074,
      "epoch": 2.0468355141236594,
      "grad_norm": 1.0,
      "learning_rate": 0.00045881480324952455,
      "loss": 5.1166,
      "mean_token_accuracy": 0.1989521250128746,
      "num_tokens": 40680248.0,
      "step": 18225
    },
    {
      "entropy": 5.448878192901612,
      "epoch": 2.0473970910316166,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045879164648186853,
      "loss": 5.2031,
      "mean_token_accuracy": 0.19653677195310593,
      "num_tokens": 40692704.0,
      "step": 18230
    },
    {
      "entropy": 5.406700992584229,
      "epoch": 2.0479586679395743,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045876848386221513,
      "loss": 5.1102,
      "mean_token_accuracy": 0.2030825674533844,
      "num_tokens": 40703156.0,
      "step": 18235
    },
    {
      "entropy": 5.356577587127686,
      "epoch": 2.048520244847532,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045874531539130204,
      "loss": 5.0771,
      "mean_token_accuracy": 0.20506429076194763,
      "num_tokens": 40713996.0,
      "step": 18240
    },
    {
      "entropy": 5.36669430732727,
      "epoch": 2.0490818217554896,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045872214106986684,
      "loss": 5.0023,
      "mean_token_accuracy": 0.2052008956670761,
      "num_tokens": 40724532.0,
      "step": 18245
    },
    {
      "entropy": 5.43301796913147,
      "epoch": 2.049643398663447,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045869896089864733,
      "loss": 5.196,
      "mean_token_accuracy": 0.19586007297039032,
      "num_tokens": 40736045.0,
      "step": 18250
    },
    {
      "entropy": 5.26068902015686,
      "epoch": 2.0502049755714045,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004586757748783817,
      "loss": 4.991,
      "mean_token_accuracy": 0.2045486867427826,
      "num_tokens": 40747398.0,
      "step": 18255
    },
    {
      "entropy": 5.249208450317383,
      "epoch": 2.050766552479362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045865258300980817,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21433522850275039,
      "num_tokens": 40758219.0,
      "step": 18260
    },
    {
      "entropy": 5.402108955383301,
      "epoch": 2.05132812938732,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045862938529366514,
      "loss": 5.0593,
      "mean_token_accuracy": 0.20787545293569565,
      "num_tokens": 40770002.0,
      "step": 18265
    },
    {
      "entropy": 5.276144313812256,
      "epoch": 2.051889706295277,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004586061817306913,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20612627565860747,
      "num_tokens": 40780087.0,
      "step": 18270
    },
    {
      "entropy": 5.30702428817749,
      "epoch": 2.0524512832032347,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045858297232162536,
      "loss": 5.181,
      "mean_token_accuracy": 0.1962374210357666,
      "num_tokens": 40792225.0,
      "step": 18275
    },
    {
      "entropy": 5.451559114456177,
      "epoch": 2.0530128601111923,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045855975706720644,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19323976784944535,
      "num_tokens": 40804437.0,
      "step": 18280
    },
    {
      "entropy": 5.347412014007569,
      "epoch": 2.05357443701915,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004585365359681736,
      "loss": 4.9853,
      "mean_token_accuracy": 0.2207135334610939,
      "num_tokens": 40814625.0,
      "step": 18285
    },
    {
      "entropy": 5.254509592056275,
      "epoch": 2.054136013927107,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045851330902526633,
      "loss": 5.0336,
      "mean_token_accuracy": 0.21061470210552216,
      "num_tokens": 40825843.0,
      "step": 18290
    },
    {
      "entropy": 5.398763799667359,
      "epoch": 2.054697590835065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045849007623922406,
      "loss": 5.2107,
      "mean_token_accuracy": 0.19434541761875151,
      "num_tokens": 40837146.0,
      "step": 18295
    },
    {
      "entropy": 5.5239674091339115,
      "epoch": 2.0552591677430225,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004584668376107867,
      "loss": 5.2814,
      "mean_token_accuracy": 0.1863650470972061,
      "num_tokens": 40849109.0,
      "step": 18300
    },
    {
      "entropy": 5.3860982894897464,
      "epoch": 2.0558207446509797,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000458443593140694,
      "loss": 5.0571,
      "mean_token_accuracy": 0.20215950757265091,
      "num_tokens": 40861168.0,
      "step": 18305
    },
    {
      "entropy": 5.405079174041748,
      "epoch": 2.0563823215589374,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00045842034282968615,
      "loss": 5.15,
      "mean_token_accuracy": 0.19916045069694518,
      "num_tokens": 40873403.0,
      "step": 18310
    },
    {
      "entropy": 5.34844012260437,
      "epoch": 2.056943898466895,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004583970866785035,
      "loss": 5.0624,
      "mean_token_accuracy": 0.1991749584674835,
      "num_tokens": 40883434.0,
      "step": 18315
    },
    {
      "entropy": 5.319121026992798,
      "epoch": 2.0575054753748527,
      "grad_norm": 1.421875,
      "learning_rate": 0.00045837382468788643,
      "loss": 5.088,
      "mean_token_accuracy": 0.20082665085792542,
      "num_tokens": 40894742.0,
      "step": 18320
    },
    {
      "entropy": 5.315261507034302,
      "epoch": 2.05806705228281,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004583505568585757,
      "loss": 5.0734,
      "mean_token_accuracy": 0.2064606487751007,
      "num_tokens": 40905233.0,
      "step": 18325
    },
    {
      "entropy": 5.379388618469238,
      "epoch": 2.0586286291907676,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045832728319131205,
      "loss": 5.0548,
      "mean_token_accuracy": 0.212684328854084,
      "num_tokens": 40915897.0,
      "step": 18330
    },
    {
      "entropy": 5.309212779998779,
      "epoch": 2.0591902060987253,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004583040036868367,
      "loss": 5.0658,
      "mean_token_accuracy": 0.20224380642175674,
      "num_tokens": 40928178.0,
      "step": 18335
    },
    {
      "entropy": 5.326982307434082,
      "epoch": 2.059751783006683,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004582807183458907,
      "loss": 5.1446,
      "mean_token_accuracy": 0.2008953258395195,
      "num_tokens": 40938988.0,
      "step": 18340
    },
    {
      "entropy": 5.346106958389282,
      "epoch": 2.06031335991464,
      "grad_norm": 1.125,
      "learning_rate": 0.00045825742716921553,
      "loss": 5.0175,
      "mean_token_accuracy": 0.20857040584087372,
      "num_tokens": 40949671.0,
      "step": 18345
    },
    {
      "entropy": 5.231579685211182,
      "epoch": 2.060874936822598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004582341301575529,
      "loss": 4.9863,
      "mean_token_accuracy": 0.20516614019870758,
      "num_tokens": 40960818.0,
      "step": 18350
    },
    {
      "entropy": 5.441619920730591,
      "epoch": 2.0614365137305555,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045821082731164445,
      "loss": 5.233,
      "mean_token_accuracy": 0.1916434124112129,
      "num_tokens": 40973866.0,
      "step": 18355
    },
    {
      "entropy": 5.355815076828003,
      "epoch": 2.061998090638513,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004581875186322322,
      "loss": 5.104,
      "mean_token_accuracy": 0.20263177752494813,
      "num_tokens": 40983355.0,
      "step": 18360
    },
    {
      "entropy": 5.306634187698364,
      "epoch": 2.0625596675464704,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045816420412005826,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2068729281425476,
      "num_tokens": 40994045.0,
      "step": 18365
    },
    {
      "entropy": 5.290441942214966,
      "epoch": 2.063121244454428,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045814088377586505,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2117360755801201,
      "num_tokens": 41003974.0,
      "step": 18370
    },
    {
      "entropy": 5.373724889755249,
      "epoch": 2.0636828213623857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045811755760039513,
      "loss": 5.1761,
      "mean_token_accuracy": 0.19960088282823563,
      "num_tokens": 41015262.0,
      "step": 18375
    },
    {
      "entropy": 5.4907402992248535,
      "epoch": 2.0642443982703433,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045809422559439113,
      "loss": 5.2048,
      "mean_token_accuracy": 0.1991424024105072,
      "num_tokens": 41027103.0,
      "step": 18380
    },
    {
      "entropy": 5.421656847000122,
      "epoch": 2.0648059751783006,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045807088775859595,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20648439973592758,
      "num_tokens": 41037240.0,
      "step": 18385
    },
    {
      "entropy": 5.399942350387573,
      "epoch": 2.065367552086258,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004580475440937527,
      "loss": 5.0924,
      "mean_token_accuracy": 0.208098141849041,
      "num_tokens": 41048258.0,
      "step": 18390
    },
    {
      "entropy": 5.413048028945923,
      "epoch": 2.065929128994216,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045802419460060464,
      "loss": 5.1853,
      "mean_token_accuracy": 0.19917104691267012,
      "num_tokens": 41058204.0,
      "step": 18395
    },
    {
      "entropy": 5.4817554473876955,
      "epoch": 2.066490705902173,
      "grad_norm": 1.25,
      "learning_rate": 0.0004580008392798953,
      "loss": 5.1663,
      "mean_token_accuracy": 0.19782832115888596,
      "num_tokens": 41070019.0,
      "step": 18400
    },
    {
      "entropy": 5.414575529098511,
      "epoch": 2.0670522828101308,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045797747813236815,
      "loss": 5.1399,
      "mean_token_accuracy": 0.2044190913438797,
      "num_tokens": 41081503.0,
      "step": 18405
    },
    {
      "entropy": 5.400751256942749,
      "epoch": 2.0676138597180884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045795411115876725,
      "loss": 5.1053,
      "mean_token_accuracy": 0.19786039739847183,
      "num_tokens": 41091959.0,
      "step": 18410
    },
    {
      "entropy": 5.416269731521607,
      "epoch": 2.068175436626046,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045793073835983644,
      "loss": 5.1529,
      "mean_token_accuracy": 0.19674252569675446,
      "num_tokens": 41103193.0,
      "step": 18415
    },
    {
      "entropy": 5.43084020614624,
      "epoch": 2.0687370135340033,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045790735973631996,
      "loss": 5.1932,
      "mean_token_accuracy": 0.20695868581533433,
      "num_tokens": 41115886.0,
      "step": 18420
    },
    {
      "entropy": 5.340741348266602,
      "epoch": 2.069298590441961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045788397528896224,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20246489942073823,
      "num_tokens": 41126868.0,
      "step": 18425
    },
    {
      "entropy": 5.4531485080719,
      "epoch": 2.0698601673499186,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045786058501850774,
      "loss": 5.1027,
      "mean_token_accuracy": 0.20663734823465346,
      "num_tokens": 41136417.0,
      "step": 18430
    },
    {
      "entropy": 5.380668640136719,
      "epoch": 2.0704217442578763,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045783718892570136,
      "loss": 5.1953,
      "mean_token_accuracy": 0.1978719040751457,
      "num_tokens": 41149515.0,
      "step": 18435
    },
    {
      "entropy": 5.421310567855835,
      "epoch": 2.0709833211658335,
      "grad_norm": 1.078125,
      "learning_rate": 0.000457813787011288,
      "loss": 5.2008,
      "mean_token_accuracy": 0.20366082191467286,
      "num_tokens": 41160023.0,
      "step": 18440
    },
    {
      "entropy": 5.2876317501068115,
      "epoch": 2.071544898073791,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004577903792760127,
      "loss": 4.9833,
      "mean_token_accuracy": 0.21394117176532745,
      "num_tokens": 41170909.0,
      "step": 18445
    },
    {
      "entropy": 5.44946403503418,
      "epoch": 2.072106474981749,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004577669657206209,
      "loss": 5.1911,
      "mean_token_accuracy": 0.1966727778315544,
      "num_tokens": 41181960.0,
      "step": 18450
    },
    {
      "entropy": 5.409207010269165,
      "epoch": 2.0726680518897065,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004577435463458579,
      "loss": 5.1856,
      "mean_token_accuracy": 0.20080088824033737,
      "num_tokens": 41192292.0,
      "step": 18455
    },
    {
      "entropy": 5.4297956943511965,
      "epoch": 2.0732296287976637,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004577201211524696,
      "loss": 5.1264,
      "mean_token_accuracy": 0.20194948315620423,
      "num_tokens": 41203826.0,
      "step": 18460
    },
    {
      "entropy": 5.352336168289185,
      "epoch": 2.0737912057056214,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004576966901412018,
      "loss": 5.1572,
      "mean_token_accuracy": 0.19956033527851105,
      "num_tokens": 41215095.0,
      "step": 18465
    },
    {
      "entropy": 5.37502236366272,
      "epoch": 2.074352782613579,
      "grad_norm": 1.2578125,
      "learning_rate": 0.00045767325331280044,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20302596390247346,
      "num_tokens": 41224953.0,
      "step": 18470
    },
    {
      "entropy": 5.397208547592163,
      "epoch": 2.0749143595215367,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004576498106680119,
      "loss": 5.1488,
      "mean_token_accuracy": 0.19366004467010497,
      "num_tokens": 41235479.0,
      "step": 18475
    },
    {
      "entropy": 5.402877426147461,
      "epoch": 2.075475936429494,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004576263622075826,
      "loss": 5.1527,
      "mean_token_accuracy": 0.20090683549642563,
      "num_tokens": 41248520.0,
      "step": 18480
    },
    {
      "entropy": 5.500815200805664,
      "epoch": 2.0760375133374516,
      "grad_norm": 1.0625,
      "learning_rate": 0.000457602907932259,
      "loss": 5.2538,
      "mean_token_accuracy": 0.1929857462644577,
      "num_tokens": 41259112.0,
      "step": 18485
    },
    {
      "entropy": 5.387196826934814,
      "epoch": 2.0765990902454092,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045757944784278803,
      "loss": 5.0805,
      "mean_token_accuracy": 0.20559409856796265,
      "num_tokens": 41269554.0,
      "step": 18490
    },
    {
      "entropy": 5.350020933151245,
      "epoch": 2.0771606671533664,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575559819399166,
      "loss": 5.1274,
      "mean_token_accuracy": 0.197537362575531,
      "num_tokens": 41281339.0,
      "step": 18495
    },
    {
      "entropy": 5.308217620849609,
      "epoch": 2.077722244061324,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045753251022439195,
      "loss": 5.0241,
      "mean_token_accuracy": 0.21587809324264526,
      "num_tokens": 41291629.0,
      "step": 18500
    },
    {
      "entropy": 5.428149032592773,
      "epoch": 2.0782838209692818,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004575090326969614,
      "loss": 5.1676,
      "mean_token_accuracy": 0.20216035842895508,
      "num_tokens": 41302162.0,
      "step": 18505
    },
    {
      "entropy": 5.326787662506104,
      "epoch": 2.0788453978772394,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004574855493583723,
      "loss": 5.1127,
      "mean_token_accuracy": 0.20899973213672637,
      "num_tokens": 41313667.0,
      "step": 18510
    },
    {
      "entropy": 5.330275297164917,
      "epoch": 2.0794069747851966,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045746206020937276,
      "loss": 5.0794,
      "mean_token_accuracy": 0.20010633319616317,
      "num_tokens": 41325063.0,
      "step": 18515
    },
    {
      "entropy": 5.3897974491119385,
      "epoch": 2.0799685516931543,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045743856525071034,
      "loss": 5.1491,
      "mean_token_accuracy": 0.20574633926153182,
      "num_tokens": 41335974.0,
      "step": 18520
    },
    {
      "entropy": 5.333673191070557,
      "epoch": 2.080530128601112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045741506448313327,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20108962059020996,
      "num_tokens": 41347469.0,
      "step": 18525
    },
    {
      "entropy": 5.412943696975708,
      "epoch": 2.0810917055090696,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045739155790738975,
      "loss": 5.16,
      "mean_token_accuracy": 0.2078002616763115,
      "num_tokens": 41358418.0,
      "step": 18530
    },
    {
      "entropy": 5.394091653823852,
      "epoch": 2.081653282417027,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00045736804552422836,
      "loss": 5.1646,
      "mean_token_accuracy": 0.20038521736860276,
      "num_tokens": 41368859.0,
      "step": 18535
    },
    {
      "entropy": 5.384805345535279,
      "epoch": 2.0822148593249845,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004573445273343976,
      "loss": 5.2097,
      "mean_token_accuracy": 0.1947450578212738,
      "num_tokens": 41380929.0,
      "step": 18540
    },
    {
      "entropy": 5.3965669631958,
      "epoch": 2.082776436232942,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004573210033386464,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20930348634719848,
      "num_tokens": 41392190.0,
      "step": 18545
    },
    {
      "entropy": 5.469496488571167,
      "epoch": 2.0833380131409,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004572974735377238,
      "loss": 5.0803,
      "mean_token_accuracy": 0.2099890947341919,
      "num_tokens": 41402646.0,
      "step": 18550
    },
    {
      "entropy": 5.353345632553101,
      "epoch": 2.083899590048857,
      "grad_norm": 1.125,
      "learning_rate": 0.0004572739379323788,
      "loss": 5.1065,
      "mean_token_accuracy": 0.20224815607070923,
      "num_tokens": 41415179.0,
      "step": 18555
    },
    {
      "entropy": 5.3554931640625,
      "epoch": 2.0844611669568147,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045725039652336105,
      "loss": 5.1671,
      "mean_token_accuracy": 0.1957186356186867,
      "num_tokens": 41426695.0,
      "step": 18560
    },
    {
      "entropy": 5.3380695343017575,
      "epoch": 2.0850227438647724,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004572268493114199,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2004076734185219,
      "num_tokens": 41438078.0,
      "step": 18565
    },
    {
      "entropy": 5.415233850479126,
      "epoch": 2.08558432077273,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004572032962973052,
      "loss": 5.1766,
      "mean_token_accuracy": 0.19689043015241622,
      "num_tokens": 41448862.0,
      "step": 18570
    },
    {
      "entropy": 5.316043472290039,
      "epoch": 2.0861458976806873,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045717973748176695,
      "loss": 4.9543,
      "mean_token_accuracy": 0.20849559158086778,
      "num_tokens": 41458528.0,
      "step": 18575
    },
    {
      "entropy": 5.32478666305542,
      "epoch": 2.086707474588645,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045715617286555516,
      "loss": 5.0658,
      "mean_token_accuracy": 0.19791805744171143,
      "num_tokens": 41469553.0,
      "step": 18580
    },
    {
      "entropy": 5.460769891738892,
      "epoch": 2.0872690514966026,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045713260244942017,
      "loss": 5.2235,
      "mean_token_accuracy": 0.1977940708398819,
      "num_tokens": 41482327.0,
      "step": 18585
    },
    {
      "entropy": 5.338142585754395,
      "epoch": 2.08783062840456,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004571090262341125,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21440523117780685,
      "num_tokens": 41493126.0,
      "step": 18590
    },
    {
      "entropy": 5.379748153686523,
      "epoch": 2.0883922053125175,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045708544422038284,
      "loss": 5.1868,
      "mean_token_accuracy": 0.19389020651578903,
      "num_tokens": 41505597.0,
      "step": 18595
    },
    {
      "entropy": 5.433377027511597,
      "epoch": 2.088953782220475,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004570618564089819,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19524456709623336,
      "num_tokens": 41517296.0,
      "step": 18600
    },
    {
      "entropy": 5.4231782913208,
      "epoch": 2.089515359128433,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045703826280066095,
      "loss": 5.1031,
      "mean_token_accuracy": 0.20402015447616578,
      "num_tokens": 41529900.0,
      "step": 18605
    },
    {
      "entropy": 5.363110685348511,
      "epoch": 2.09007693603639,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004570146633961711,
      "loss": 5.1519,
      "mean_token_accuracy": 0.20295602530241014,
      "num_tokens": 41540641.0,
      "step": 18610
    },
    {
      "entropy": 5.326742315292359,
      "epoch": 2.0906385129443477,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045699105819626374,
      "loss": 5.1093,
      "mean_token_accuracy": 0.19371536374092102,
      "num_tokens": 41551335.0,
      "step": 18615
    },
    {
      "entropy": 5.368008279800415,
      "epoch": 2.0912000898523053,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004569674472016904,
      "loss": 5.0845,
      "mean_token_accuracy": 0.2052995130419731,
      "num_tokens": 41562265.0,
      "step": 18620
    },
    {
      "entropy": 5.3541703701019285,
      "epoch": 2.091761666760263,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004569438304132031,
      "loss": 4.9993,
      "mean_token_accuracy": 0.2083893120288849,
      "num_tokens": 41574758.0,
      "step": 18625
    },
    {
      "entropy": 5.299733972549438,
      "epoch": 2.09232324366822,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045692020783155376,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20870408415794373,
      "num_tokens": 41585007.0,
      "step": 18630
    },
    {
      "entropy": 5.291616344451905,
      "epoch": 2.092884820576178,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004568965794574943,
      "loss": 5.0271,
      "mean_token_accuracy": 0.20842434465885162,
      "num_tokens": 41596026.0,
      "step": 18635
    },
    {
      "entropy": 5.408572673797607,
      "epoch": 2.0934463974841355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004568729452917772,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20587016344070436,
      "num_tokens": 41608440.0,
      "step": 18640
    },
    {
      "entropy": 5.3973188400268555,
      "epoch": 2.094007974392093,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045684930533515506,
      "loss": 5.064,
      "mean_token_accuracy": 0.20384204536676406,
      "num_tokens": 41620467.0,
      "step": 18645
    },
    {
      "entropy": 5.279742813110351,
      "epoch": 2.0945695513000504,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045682565958838045,
      "loss": 5.0603,
      "mean_token_accuracy": 0.20455557852983475,
      "num_tokens": 41630918.0,
      "step": 18650
    },
    {
      "entropy": 5.326931762695312,
      "epoch": 2.095131128208008,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004568020080522063,
      "loss": 5.0636,
      "mean_token_accuracy": 0.20253249406814575,
      "num_tokens": 41641148.0,
      "step": 18655
    },
    {
      "entropy": 5.380511903762818,
      "epoch": 2.0956927051159657,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004567783507273858,
      "loss": 5.119,
      "mean_token_accuracy": 0.20608083605766297,
      "num_tokens": 41652148.0,
      "step": 18660
    },
    {
      "entropy": 5.301640319824219,
      "epoch": 2.0962542820239234,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045675468761467204,
      "loss": 5.095,
      "mean_token_accuracy": 0.20120594650506973,
      "num_tokens": 41662275.0,
      "step": 18665
    },
    {
      "entropy": 5.404917478561401,
      "epoch": 2.0968158589318806,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045673101871481853,
      "loss": 5.1488,
      "mean_token_accuracy": 0.1964876964688301,
      "num_tokens": 41673141.0,
      "step": 18670
    },
    {
      "entropy": 5.534079265594483,
      "epoch": 2.0973774358398383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004567073440285789,
      "loss": 5.2555,
      "mean_token_accuracy": 0.19630328863859176,
      "num_tokens": 41683473.0,
      "step": 18675
    },
    {
      "entropy": 5.288400363922119,
      "epoch": 2.097939012747796,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045668366355670697,
      "loss": 4.8994,
      "mean_token_accuracy": 0.20954634994268417,
      "num_tokens": 41693847.0,
      "step": 18680
    },
    {
      "entropy": 5.337480258941651,
      "epoch": 2.098500589655753,
      "grad_norm": 1.125,
      "learning_rate": 0.0004566599772999567,
      "loss": 5.1697,
      "mean_token_accuracy": 0.20514702945947647,
      "num_tokens": 41705810.0,
      "step": 18685
    },
    {
      "entropy": 5.353377151489258,
      "epoch": 2.099062166563711,
      "grad_norm": 1.125,
      "learning_rate": 0.00045663628525908237,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20367245376110077,
      "num_tokens": 41717160.0,
      "step": 18690
    },
    {
      "entropy": 5.419959783554077,
      "epoch": 2.0996237434716685,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004566125874348382,
      "loss": 5.1794,
      "mean_token_accuracy": 0.2020849347114563,
      "num_tokens": 41728631.0,
      "step": 18695
    },
    {
      "entropy": 5.465868520736694,
      "epoch": 2.100185320379626,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045658888382797886,
      "loss": 5.2293,
      "mean_token_accuracy": 0.19605599194765091,
      "num_tokens": 41738933.0,
      "step": 18700
    },
    {
      "entropy": 5.477331066131592,
      "epoch": 2.1007468972875833,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045656517443925897,
      "loss": 5.2248,
      "mean_token_accuracy": 0.20296594500541687,
      "num_tokens": 41750386.0,
      "step": 18705
    },
    {
      "entropy": 5.372495508193969,
      "epoch": 2.101308474195541,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004565414592694336,
      "loss": 5.124,
      "mean_token_accuracy": 0.20496938973665238,
      "num_tokens": 41760711.0,
      "step": 18710
    },
    {
      "entropy": 5.356407165527344,
      "epoch": 2.1018700511034987,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004565177383192577,
      "loss": 5.1734,
      "mean_token_accuracy": 0.19895849674940108,
      "num_tokens": 41772147.0,
      "step": 18715
    },
    {
      "entropy": 5.429410314559936,
      "epoch": 2.1024316280114563,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045649401158948664,
      "loss": 5.0564,
      "mean_token_accuracy": 0.20853542387485505,
      "num_tokens": 41783265.0,
      "step": 18720
    },
    {
      "entropy": 5.361446762084961,
      "epoch": 2.1029932049194136,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004564702790808758,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20600181668996811,
      "num_tokens": 41793977.0,
      "step": 18725
    },
    {
      "entropy": 5.330615711212158,
      "epoch": 2.103554781827371,
      "grad_norm": 1.3046875,
      "learning_rate": 0.0004564465407941809,
      "loss": 5.054,
      "mean_token_accuracy": 0.2093093529343605,
      "num_tokens": 41804790.0,
      "step": 18730
    },
    {
      "entropy": 5.2915812015533445,
      "epoch": 2.104116358735329,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045642279673015787,
      "loss": 5.0566,
      "mean_token_accuracy": 0.20036528259515762,
      "num_tokens": 41815611.0,
      "step": 18735
    },
    {
      "entropy": 5.310164451599121,
      "epoch": 2.1046779356432865,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004563990468895625,
      "loss": 5.0792,
      "mean_token_accuracy": 0.19866398125886917,
      "num_tokens": 41826935.0,
      "step": 18740
    },
    {
      "entropy": 5.288794374465942,
      "epoch": 2.1052395125512438,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004563752912731512,
      "loss": 5.033,
      "mean_token_accuracy": 0.20669914335012435,
      "num_tokens": 41838932.0,
      "step": 18745
    },
    {
      "entropy": 5.310239505767822,
      "epoch": 2.1058010894592014,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004563515298816803,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20065125077962875,
      "num_tokens": 41850771.0,
      "step": 18750
    },
    {
      "entropy": 5.396411371231079,
      "epoch": 2.106362666367159,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045632776271590624,
      "loss": 5.04,
      "mean_token_accuracy": 0.20902392715215684,
      "num_tokens": 41861515.0,
      "step": 18755
    },
    {
      "entropy": 5.3220374584198,
      "epoch": 2.1069242432751167,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004563039897765859,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21872160732746124,
      "num_tokens": 41872195.0,
      "step": 18760
    },
    {
      "entropy": 5.236818265914917,
      "epoch": 2.107485820183074,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004562802110644764,
      "loss": 5.0574,
      "mean_token_accuracy": 0.20958322584629058,
      "num_tokens": 41882496.0,
      "step": 18765
    },
    {
      "entropy": 5.460976600646973,
      "epoch": 2.1080473970910316,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045625642658033447,
      "loss": 5.2739,
      "mean_token_accuracy": 0.19815288633108138,
      "num_tokens": 41893078.0,
      "step": 18770
    },
    {
      "entropy": 5.477948522567749,
      "epoch": 2.1086089739989893,
      "grad_norm": 1.25,
      "learning_rate": 0.00045623263632491766,
      "loss": 5.146,
      "mean_token_accuracy": 0.20062417089939116,
      "num_tokens": 41905062.0,
      "step": 18775
    },
    {
      "entropy": 5.300647926330567,
      "epoch": 2.1091705509069465,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045620884029898344,
      "loss": 5.1045,
      "mean_token_accuracy": 0.20420591682195663,
      "num_tokens": 41916086.0,
      "step": 18780
    },
    {
      "entropy": 5.398499488830566,
      "epoch": 2.109732127814904,
      "grad_norm": 1.2890625,
      "learning_rate": 0.00045618503850328954,
      "loss": 5.1729,
      "mean_token_accuracy": 0.20482241958379746,
      "num_tokens": 41925466.0,
      "step": 18785
    },
    {
      "entropy": 5.432051134109497,
      "epoch": 2.110293704722862,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004561612309385936,
      "loss": 5.2037,
      "mean_token_accuracy": 0.19592470228672026,
      "num_tokens": 41936567.0,
      "step": 18790
    },
    {
      "entropy": 5.337327241897583,
      "epoch": 2.1108552816308195,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045613741760565385,
      "loss": 5.0387,
      "mean_token_accuracy": 0.20835209637880325,
      "num_tokens": 41947021.0,
      "step": 18795
    },
    {
      "entropy": 5.4122942924499515,
      "epoch": 2.1114168585387767,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004561135985052285,
      "loss": 5.1062,
      "mean_token_accuracy": 0.2018197387456894,
      "num_tokens": 41957929.0,
      "step": 18800
    },
    {
      "entropy": 5.340630054473877,
      "epoch": 2.1119784354467344,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000456089773638076,
      "loss": 5.019,
      "mean_token_accuracy": 0.2096169888973236,
      "num_tokens": 41968677.0,
      "step": 18805
    },
    {
      "entropy": 5.368287897109985,
      "epoch": 2.112540012354692,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004560659430049548,
      "loss": 5.1848,
      "mean_token_accuracy": 0.20199057161808015,
      "num_tokens": 41980330.0,
      "step": 18810
    },
    {
      "entropy": 5.384235477447509,
      "epoch": 2.1131015892626497,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004560421066066237,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20321005284786225,
      "num_tokens": 41991258.0,
      "step": 18815
    },
    {
      "entropy": 5.425364303588867,
      "epoch": 2.113663166170607,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045601826444384175,
      "loss": 5.1754,
      "mean_token_accuracy": 0.20385170727968216,
      "num_tokens": 42002679.0,
      "step": 18820
    },
    {
      "entropy": 5.3336811542510985,
      "epoch": 2.1142247430785646,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045599441651736805,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20498909950256347,
      "num_tokens": 42012845.0,
      "step": 18825
    },
    {
      "entropy": 5.416022539138794,
      "epoch": 2.1147863199865222,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045597056282796196,
      "loss": 5.1357,
      "mean_token_accuracy": 0.19906553626060486,
      "num_tokens": 42024461.0,
      "step": 18830
    },
    {
      "entropy": 5.394451904296875,
      "epoch": 2.11534789689448,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004559467033763829,
      "loss": 5.1325,
      "mean_token_accuracy": 0.1978890925645828,
      "num_tokens": 42036996.0,
      "step": 18835
    },
    {
      "entropy": 5.421212673187256,
      "epoch": 2.115909473802437,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004559228381633907,
      "loss": 5.1856,
      "mean_token_accuracy": 0.20177469849586488,
      "num_tokens": 42048731.0,
      "step": 18840
    },
    {
      "entropy": 5.436614751815796,
      "epoch": 2.1164710507103948,
      "grad_norm": 1.3125,
      "learning_rate": 0.00045589896718974517,
      "loss": 5.1418,
      "mean_token_accuracy": 0.20479384660720826,
      "num_tokens": 42060396.0,
      "step": 18845
    },
    {
      "entropy": 5.266221141815185,
      "epoch": 2.1170326276183524,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004558750904562063,
      "loss": 5.0276,
      "mean_token_accuracy": 0.20432794243097305,
      "num_tokens": 42072037.0,
      "step": 18850
    },
    {
      "entropy": 5.321339845657349,
      "epoch": 2.11759420452631,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004558512079635345,
      "loss": 4.9937,
      "mean_token_accuracy": 0.2105395168066025,
      "num_tokens": 42082577.0,
      "step": 18855
    },
    {
      "entropy": 5.2712665557861325,
      "epoch": 2.1181557814342673,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004558273197124899,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2145879715681076,
      "num_tokens": 42093978.0,
      "step": 18860
    },
    {
      "entropy": 5.368838691711426,
      "epoch": 2.118717358342225,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045580342570383347,
      "loss": 5.2118,
      "mean_token_accuracy": 0.19626391530036927,
      "num_tokens": 42105773.0,
      "step": 18865
    },
    {
      "entropy": 5.392467069625854,
      "epoch": 2.1192789352501826,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004557795259383258,
      "loss": 5.1076,
      "mean_token_accuracy": 0.2028869792819023,
      "num_tokens": 42116232.0,
      "step": 18870
    },
    {
      "entropy": 5.366229677200318,
      "epoch": 2.11984051215814,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045575562041672794,
      "loss": 5.1246,
      "mean_token_accuracy": 0.2042296499013901,
      "num_tokens": 42127813.0,
      "step": 18875
    },
    {
      "entropy": 5.278764152526856,
      "epoch": 2.1204020890660975,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045573170913980094,
      "loss": 5.0527,
      "mean_token_accuracy": 0.2064984291791916,
      "num_tokens": 42138054.0,
      "step": 18880
    },
    {
      "entropy": 5.4260131359100345,
      "epoch": 2.120963665974055,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045570779210830627,
      "loss": 5.1162,
      "mean_token_accuracy": 0.2021315097808838,
      "num_tokens": 42149575.0,
      "step": 18885
    },
    {
      "entropy": 5.341203308105468,
      "epoch": 2.121525242882013,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004556838693230054,
      "loss": 5.1403,
      "mean_token_accuracy": 0.20253828912973404,
      "num_tokens": 42160824.0,
      "step": 18890
    },
    {
      "entropy": 5.36500768661499,
      "epoch": 2.12208681978997,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045565994078466005,
      "loss": 5.1726,
      "mean_token_accuracy": 0.20073347687721252,
      "num_tokens": 42171088.0,
      "step": 18895
    },
    {
      "entropy": 5.359429359436035,
      "epoch": 2.1226483966979277,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045563600649403213,
      "loss": 5.1225,
      "mean_token_accuracy": 0.19776681810617447,
      "num_tokens": 42182112.0,
      "step": 18900
    },
    {
      "entropy": 5.396687746047974,
      "epoch": 2.1232099736058854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004556120664518838,
      "loss": 5.0126,
      "mean_token_accuracy": 0.21290193051099776,
      "num_tokens": 42193563.0,
      "step": 18905
    },
    {
      "entropy": 5.344514322280884,
      "epoch": 2.123771550513843,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004555881206589771,
      "loss": 5.0774,
      "mean_token_accuracy": 0.20512334555387496,
      "num_tokens": 42204891.0,
      "step": 18910
    },
    {
      "entropy": 5.34842848777771,
      "epoch": 2.1243331274218002,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004555641691160746,
      "loss": 5.1409,
      "mean_token_accuracy": 0.20819477289915084,
      "num_tokens": 42217226.0,
      "step": 18915
    },
    {
      "entropy": 5.438216209411621,
      "epoch": 2.124894704329758,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000455540211823939,
      "loss": 5.1245,
      "mean_token_accuracy": 0.20053599327802657,
      "num_tokens": 42228134.0,
      "step": 18920
    },
    {
      "entropy": 5.368012237548828,
      "epoch": 2.1254562812377156,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045551624878333296,
      "loss": 5.1375,
      "mean_token_accuracy": 0.19445285350084304,
      "num_tokens": 42239868.0,
      "step": 18925
    },
    {
      "entropy": 5.40369348526001,
      "epoch": 2.1260178581456732,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045549227999501957,
      "loss": 5.1827,
      "mean_token_accuracy": 0.2057632938027382,
      "num_tokens": 42250297.0,
      "step": 18930
    },
    {
      "entropy": 5.324860286712647,
      "epoch": 2.1265794350536305,
      "grad_norm": 1.125,
      "learning_rate": 0.00045546830545976203,
      "loss": 4.8736,
      "mean_token_accuracy": 0.21715236008167266,
      "num_tokens": 42261263.0,
      "step": 18935
    },
    {
      "entropy": 5.367534923553467,
      "epoch": 2.127141011961588,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045544432517832356,
      "loss": 5.1578,
      "mean_token_accuracy": 0.2044275164604187,
      "num_tokens": 42272357.0,
      "step": 18940
    },
    {
      "entropy": 5.397897100448608,
      "epoch": 2.1277025888695458,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045542033915146787,
      "loss": 5.1755,
      "mean_token_accuracy": 0.2024521067738533,
      "num_tokens": 42283630.0,
      "step": 18945
    },
    {
      "entropy": 5.397818183898925,
      "epoch": 2.1282641657775034,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045539634737995857,
      "loss": 5.1926,
      "mean_token_accuracy": 0.18823411017656327,
      "num_tokens": 42294981.0,
      "step": 18950
    },
    {
      "entropy": 5.3841241836547855,
      "epoch": 2.1288257426854607,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045537234986455964,
      "loss": 5.0626,
      "mean_token_accuracy": 0.2016092523932457,
      "num_tokens": 42304877.0,
      "step": 18955
    },
    {
      "entropy": 5.359314012527466,
      "epoch": 2.1293873195934183,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004553483466060351,
      "loss": 5.1419,
      "mean_token_accuracy": 0.2008747860789299,
      "num_tokens": 42316325.0,
      "step": 18960
    },
    {
      "entropy": 5.362393522262574,
      "epoch": 2.129948896501376,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045532433760514925,
      "loss": 5.1005,
      "mean_token_accuracy": 0.2047993540763855,
      "num_tokens": 42327437.0,
      "step": 18965
    },
    {
      "entropy": 5.285092449188232,
      "epoch": 2.130510473409333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004553003228626665,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21103420108556747,
      "num_tokens": 42337742.0,
      "step": 18970
    },
    {
      "entropy": 5.371410608291626,
      "epoch": 2.131072050317291,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004552763023793517,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20686433613300323,
      "num_tokens": 42348511.0,
      "step": 18975
    },
    {
      "entropy": 5.328984928131104,
      "epoch": 2.1316336272252485,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004552522761559694,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2069281354546547,
      "num_tokens": 42358977.0,
      "step": 18980
    },
    {
      "entropy": 5.377198505401611,
      "epoch": 2.132195204133206,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00045522824419328475,
      "loss": 5.1558,
      "mean_token_accuracy": 0.195808307826519,
      "num_tokens": 42371609.0,
      "step": 18985
    },
    {
      "entropy": 5.409285163879394,
      "epoch": 2.1327567810411634,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045520420649206283,
      "loss": 5.118,
      "mean_token_accuracy": 0.20816231667995452,
      "num_tokens": 42383150.0,
      "step": 18990
    },
    {
      "entropy": 5.336454439163208,
      "epoch": 2.133318357949121,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004551801630530692,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20229487717151642,
      "num_tokens": 42395150.0,
      "step": 18995
    },
    {
      "entropy": 5.3731800556182865,
      "epoch": 2.1338799348570787,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004551561138770693,
      "loss": 5.1323,
      "mean_token_accuracy": 0.20241184532642365,
      "num_tokens": 42406348.0,
      "step": 19000
    },
    {
      "entropy": 5.400448369979858,
      "epoch": 2.1344415117650364,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045513205896482875,
      "loss": 5.1327,
      "mean_token_accuracy": 0.1975383535027504,
      "num_tokens": 42417183.0,
      "step": 19005
    },
    {
      "entropy": 5.400464868545532,
      "epoch": 2.1350030886729936,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045510799831711366,
      "loss": 5.059,
      "mean_token_accuracy": 0.21160545349121093,
      "num_tokens": 42427805.0,
      "step": 19010
    },
    {
      "entropy": 5.338586044311524,
      "epoch": 2.1355646655809513,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00045508393193469,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20181637853384018,
      "num_tokens": 42439521.0,
      "step": 19015
    },
    {
      "entropy": 5.408268022537231,
      "epoch": 2.136126242488909,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004550598598183241,
      "loss": 5.1918,
      "mean_token_accuracy": 0.2013215333223343,
      "num_tokens": 42451201.0,
      "step": 19020
    },
    {
      "entropy": 5.379426908493042,
      "epoch": 2.1366878193968666,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004550357819687825,
      "loss": 5.0836,
      "mean_token_accuracy": 0.2061127632856369,
      "num_tokens": 42462315.0,
      "step": 19025
    },
    {
      "entropy": 5.5117699146270756,
      "epoch": 2.137249396304824,
      "grad_norm": 1.2265625,
      "learning_rate": 0.0004550116983868317,
      "loss": 5.285,
      "mean_token_accuracy": 0.18784965127706527,
      "num_tokens": 42473993.0,
      "step": 19030
    },
    {
      "entropy": 5.482213878631592,
      "epoch": 2.1378109732127815,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004549876090732386,
      "loss": 5.2536,
      "mean_token_accuracy": 0.19378787726163865,
      "num_tokens": 42484307.0,
      "step": 19035
    },
    {
      "entropy": 5.3942736148834225,
      "epoch": 2.138372550120739,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004549635140287703,
      "loss": 5.1679,
      "mean_token_accuracy": 0.20006123036146164,
      "num_tokens": 42497305.0,
      "step": 19040
    },
    {
      "entropy": 5.431675863265991,
      "epoch": 2.138934127028697,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045493941325419385,
      "loss": 5.14,
      "mean_token_accuracy": 0.20057855248451234,
      "num_tokens": 42508307.0,
      "step": 19045
    },
    {
      "entropy": 5.408052158355713,
      "epoch": 2.139495703936654,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004549153067502768,
      "loss": 5.2166,
      "mean_token_accuracy": 0.19544317275285722,
      "num_tokens": 42519946.0,
      "step": 19050
    },
    {
      "entropy": 5.302248191833496,
      "epoch": 2.1400572808446117,
      "grad_norm": 1.296875,
      "learning_rate": 0.00045489119451778645,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2079508736729622,
      "num_tokens": 42530737.0,
      "step": 19055
    },
    {
      "entropy": 5.256875944137573,
      "epoch": 2.1406188577525693,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004548670765574908,
      "loss": 4.9866,
      "mean_token_accuracy": 0.208221635222435,
      "num_tokens": 42542184.0,
      "step": 19060
    },
    {
      "entropy": 5.339967250823975,
      "epoch": 2.1411804346605265,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045484295287015757,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19821846038103103,
      "num_tokens": 42553816.0,
      "step": 19065
    },
    {
      "entropy": 5.455509948730469,
      "epoch": 2.141742011568484,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004548188234565551,
      "loss": 5.1374,
      "mean_token_accuracy": 0.20365287512540817,
      "num_tokens": 42564685.0,
      "step": 19070
    },
    {
      "entropy": 5.329450416564941,
      "epoch": 2.142303588476442,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045479468831745144,
      "loss": 5.0673,
      "mean_token_accuracy": 0.19669991582632065,
      "num_tokens": 42574864.0,
      "step": 19075
    },
    {
      "entropy": 5.373766469955444,
      "epoch": 2.1428651653843995,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004547705474536152,
      "loss": 5.0581,
      "mean_token_accuracy": 0.2014237716794014,
      "num_tokens": 42585214.0,
      "step": 19080
    },
    {
      "entropy": 5.436512613296509,
      "epoch": 2.1434267422923567,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045474640086581506,
      "loss": 5.1898,
      "mean_token_accuracy": 0.1994388997554779,
      "num_tokens": 42596017.0,
      "step": 19085
    },
    {
      "entropy": 5.375260305404663,
      "epoch": 2.1439883192003144,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045472224855481975,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20052024573087693,
      "num_tokens": 42607903.0,
      "step": 19090
    },
    {
      "entropy": 5.349408578872681,
      "epoch": 2.144549896108272,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045469809052139844,
      "loss": 5.09,
      "mean_token_accuracy": 0.20410326421260833,
      "num_tokens": 42618690.0,
      "step": 19095
    },
    {
      "entropy": 5.357032299041748,
      "epoch": 2.1451114730162297,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045467392676632016,
      "loss": 5.0643,
      "mean_token_accuracy": 0.20388213843107222,
      "num_tokens": 42629655.0,
      "step": 19100
    },
    {
      "entropy": 5.380858087539673,
      "epoch": 2.145673049924187,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004546497572903543,
      "loss": 5.175,
      "mean_token_accuracy": 0.2018699511885643,
      "num_tokens": 42641214.0,
      "step": 19105
    },
    {
      "entropy": 5.420915651321411,
      "epoch": 2.1462346268321446,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004546255820942706,
      "loss": 5.2072,
      "mean_token_accuracy": 0.19548473805189132,
      "num_tokens": 42653829.0,
      "step": 19110
    },
    {
      "entropy": 5.444293260574341,
      "epoch": 2.1467962037401023,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004546014011788386,
      "loss": 5.1443,
      "mean_token_accuracy": 0.19995381385087968,
      "num_tokens": 42665245.0,
      "step": 19115
    },
    {
      "entropy": 5.382050609588623,
      "epoch": 2.14735778064806,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004545772145448284,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20089350789785385,
      "num_tokens": 42676409.0,
      "step": 19120
    },
    {
      "entropy": 5.291347551345825,
      "epoch": 2.147919357556017,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045455302219301,
      "loss": 4.9341,
      "mean_token_accuracy": 0.21362725645303726,
      "num_tokens": 42686586.0,
      "step": 19125
    },
    {
      "entropy": 5.242350149154663,
      "epoch": 2.148480934463975,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004545288241241538,
      "loss": 5.01,
      "mean_token_accuracy": 0.20715139508247377,
      "num_tokens": 42696597.0,
      "step": 19130
    },
    {
      "entropy": 5.4127318382263185,
      "epoch": 2.1490425113719325,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004545046203390302,
      "loss": 5.2062,
      "mean_token_accuracy": 0.20166823267936707,
      "num_tokens": 42707447.0,
      "step": 19135
    },
    {
      "entropy": 5.532575607299805,
      "epoch": 2.14960408827989,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00045448041083840973,
      "loss": 5.2197,
      "mean_token_accuracy": 0.19463301599025726,
      "num_tokens": 42717533.0,
      "step": 19140
    },
    {
      "entropy": 5.41377100944519,
      "epoch": 2.1501656651878474,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045445619562306354,
      "loss": 5.2045,
      "mean_token_accuracy": 0.1980792388319969,
      "num_tokens": 42730563.0,
      "step": 19145
    },
    {
      "entropy": 5.351421356201172,
      "epoch": 2.150727242095805,
      "grad_norm": 1.125,
      "learning_rate": 0.00045443197469376243,
      "loss": 5.0902,
      "mean_token_accuracy": 0.20530505180358888,
      "num_tokens": 42740301.0,
      "step": 19150
    },
    {
      "entropy": 5.391275930404663,
      "epoch": 2.1512888190037627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004544077480512776,
      "loss": 5.1223,
      "mean_token_accuracy": 0.19607978612184523,
      "num_tokens": 42752296.0,
      "step": 19155
    },
    {
      "entropy": 5.316630172729492,
      "epoch": 2.15185039591172,
      "grad_norm": 1.84375,
      "learning_rate": 0.00045438351569638055,
      "loss": 5.0041,
      "mean_token_accuracy": 0.20520310550928117,
      "num_tokens": 42762530.0,
      "step": 19160
    },
    {
      "entropy": 5.329446506500244,
      "epoch": 2.1524119728196776,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045435927762984277,
      "loss": 5.0524,
      "mean_token_accuracy": 0.2042991414666176,
      "num_tokens": 42773059.0,
      "step": 19165
    },
    {
      "entropy": 5.351168632507324,
      "epoch": 2.152973549727635,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000454335033852436,
      "loss": 5.0734,
      "mean_token_accuracy": 0.2085629239678383,
      "num_tokens": 42783832.0,
      "step": 19170
    },
    {
      "entropy": 5.420539522171021,
      "epoch": 2.153535126635593,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004543107843649322,
      "loss": 5.1799,
      "mean_token_accuracy": 0.2021641343832016,
      "num_tokens": 42794565.0,
      "step": 19175
    },
    {
      "entropy": 5.446712017059326,
      "epoch": 2.15409670354355,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004542865291681035,
      "loss": 5.2172,
      "mean_token_accuracy": 0.19348224252462387,
      "num_tokens": 42805536.0,
      "step": 19180
    },
    {
      "entropy": 5.4203441619873045,
      "epoch": 2.1546582804515078,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045426226826272213,
      "loss": 5.1533,
      "mean_token_accuracy": 0.1961967185139656,
      "num_tokens": 42816718.0,
      "step": 19185
    },
    {
      "entropy": 5.381061983108521,
      "epoch": 2.1552198573594654,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004542380016495607,
      "loss": 5.1205,
      "mean_token_accuracy": 0.20412277281284333,
      "num_tokens": 42830183.0,
      "step": 19190
    },
    {
      "entropy": 5.337467813491822,
      "epoch": 2.155781434267423,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045421372932939177,
      "loss": 5.121,
      "mean_token_accuracy": 0.20671890676021576,
      "num_tokens": 42841241.0,
      "step": 19195
    },
    {
      "entropy": 5.545665168762207,
      "epoch": 2.1563430111753803,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004541894513029881,
      "loss": 5.1928,
      "mean_token_accuracy": 0.19893884807825088,
      "num_tokens": 42852775.0,
      "step": 19200
    },
    {
      "entropy": 5.53910346031189,
      "epoch": 2.156904588083338,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045416516757112286,
      "loss": 5.3023,
      "mean_token_accuracy": 0.19325364530086517,
      "num_tokens": 42863947.0,
      "step": 19205
    },
    {
      "entropy": 5.295586633682251,
      "epoch": 2.1574661649912956,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045414087813456925,
      "loss": 5.1491,
      "mean_token_accuracy": 0.2045189008116722,
      "num_tokens": 42874794.0,
      "step": 19210
    },
    {
      "entropy": 5.39025387763977,
      "epoch": 2.1580277418992533,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004541165829941006,
      "loss": 5.1138,
      "mean_token_accuracy": 0.20128049850463867,
      "num_tokens": 42886038.0,
      "step": 19215
    },
    {
      "entropy": 5.40678038597107,
      "epoch": 2.1585893188072105,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004540922821504904,
      "loss": 5.1361,
      "mean_token_accuracy": 0.19972163587808608,
      "num_tokens": 42897243.0,
      "step": 19220
    },
    {
      "entropy": 5.334144973754883,
      "epoch": 2.159150895715168,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004540679756045125,
      "loss": 5.0406,
      "mean_token_accuracy": 0.20674597918987275,
      "num_tokens": 42908802.0,
      "step": 19225
    },
    {
      "entropy": 5.4347789764404295,
      "epoch": 2.159712472623126,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004540436633569407,
      "loss": 5.2187,
      "mean_token_accuracy": 0.19259276539087294,
      "num_tokens": 42921156.0,
      "step": 19230
    },
    {
      "entropy": 5.292076110839844,
      "epoch": 2.1602740495310835,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045401934540854937,
      "loss": 5.0214,
      "mean_token_accuracy": 0.2119014024734497,
      "num_tokens": 42931336.0,
      "step": 19235
    },
    {
      "entropy": 5.338836240768432,
      "epoch": 2.1608356264390407,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004539950217601127,
      "loss": 5.1159,
      "mean_token_accuracy": 0.21029542833566667,
      "num_tokens": 42942312.0,
      "step": 19240
    },
    {
      "entropy": 5.3649670600891115,
      "epoch": 2.1613972033469984,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045397069241240497,
      "loss": 5.1585,
      "mean_token_accuracy": 0.20633459836244583,
      "num_tokens": 42953429.0,
      "step": 19245
    },
    {
      "entropy": 5.35712513923645,
      "epoch": 2.161958780254956,
      "grad_norm": 1.140625,
      "learning_rate": 0.000453946357366201,
      "loss": 5.0249,
      "mean_token_accuracy": 0.20839226990938187,
      "num_tokens": 42963354.0,
      "step": 19250
    },
    {
      "entropy": 5.443540811538696,
      "epoch": 2.1625203571629132,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045392201662227565,
      "loss": 5.1823,
      "mean_token_accuracy": 0.20375896096229554,
      "num_tokens": 42974643.0,
      "step": 19255
    },
    {
      "entropy": 5.38089509010315,
      "epoch": 2.163081934070871,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045389767018140395,
      "loss": 5.1531,
      "mean_token_accuracy": 0.20015639811754227,
      "num_tokens": 42985821.0,
      "step": 19260
    },
    {
      "entropy": 5.446358633041382,
      "epoch": 2.1636435109788286,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00045387331804436096,
      "loss": 5.2,
      "mean_token_accuracy": 0.1969418615102768,
      "num_tokens": 42998258.0,
      "step": 19265
    },
    {
      "entropy": 5.393283224105835,
      "epoch": 2.1642050878867862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004538489602119222,
      "loss": 5.0901,
      "mean_token_accuracy": 0.20232222378253936,
      "num_tokens": 43009117.0,
      "step": 19270
    },
    {
      "entropy": 5.311589145660401,
      "epoch": 2.1647666647947434,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004538245966848632,
      "loss": 5.093,
      "mean_token_accuracy": 0.20592110455036164,
      "num_tokens": 43020437.0,
      "step": 19275
    },
    {
      "entropy": 5.365633010864258,
      "epoch": 2.165328241702701,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045380022746395953,
      "loss": 5.1432,
      "mean_token_accuracy": 0.2059076115489006,
      "num_tokens": 43031722.0,
      "step": 19280
    },
    {
      "entropy": 5.354097557067871,
      "epoch": 2.1658898186106588,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045377585254998743,
      "loss": 5.0683,
      "mean_token_accuracy": 0.21458153575658798,
      "num_tokens": 43042957.0,
      "step": 19285
    },
    {
      "entropy": 5.334741640090942,
      "epoch": 2.1664513955186164,
      "grad_norm": 1.125,
      "learning_rate": 0.0004537514719437228,
      "loss": 5.1346,
      "mean_token_accuracy": 0.19941189587116243,
      "num_tokens": 43054086.0,
      "step": 19290
    },
    {
      "entropy": 5.476854515075684,
      "epoch": 2.1670129724265736,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045372708564594196,
      "loss": 5.2166,
      "mean_token_accuracy": 0.1965984582901001,
      "num_tokens": 43066259.0,
      "step": 19295
    },
    {
      "entropy": 5.29880313873291,
      "epoch": 2.1675745493345313,
      "grad_norm": 1.234375,
      "learning_rate": 0.00045370269365742143,
      "loss": 4.9672,
      "mean_token_accuracy": 0.20954586416482926,
      "num_tokens": 43076837.0,
      "step": 19300
    },
    {
      "entropy": 5.291219902038574,
      "epoch": 2.168136126242489,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045367829597893776,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20618225187063216,
      "num_tokens": 43087802.0,
      "step": 19305
    },
    {
      "entropy": 5.333664178848267,
      "epoch": 2.1686977031504466,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045365389261126793,
      "loss": 5.0559,
      "mean_token_accuracy": 0.2061526283621788,
      "num_tokens": 43097562.0,
      "step": 19310
    },
    {
      "entropy": 5.246506690979004,
      "epoch": 2.169259280058404,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045362948355518883,
      "loss": 5.0574,
      "mean_token_accuracy": 0.2059615060687065,
      "num_tokens": 43107973.0,
      "step": 19315
    },
    {
      "entropy": 5.28773832321167,
      "epoch": 2.1698208569663615,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004536050688114777,
      "loss": 5.0005,
      "mean_token_accuracy": 0.21197632253170012,
      "num_tokens": 43118923.0,
      "step": 19320
    },
    {
      "entropy": 5.357081127166748,
      "epoch": 2.170382433874319,
      "grad_norm": 0.984375,
      "learning_rate": 0.00045358064838091187,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20547584295272828,
      "num_tokens": 43130904.0,
      "step": 19325
    },
    {
      "entropy": 5.378326272964477,
      "epoch": 2.170944010782277,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045355622226426903,
      "loss": 5.0714,
      "mean_token_accuracy": 0.19961223602294922,
      "num_tokens": 43142293.0,
      "step": 19330
    },
    {
      "entropy": 5.366404390335083,
      "epoch": 2.171505587690234,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00045353179046232674,
      "loss": 5.1525,
      "mean_token_accuracy": 0.20788997709751128,
      "num_tokens": 43154340.0,
      "step": 19335
    },
    {
      "entropy": 5.35880036354065,
      "epoch": 2.1720671645981917,
      "grad_norm": 1.1640625,
      "learning_rate": 0.000453507352975863,
      "loss": 5.0839,
      "mean_token_accuracy": 0.20857978314161302,
      "num_tokens": 43165924.0,
      "step": 19340
    },
    {
      "entropy": 5.330966997146606,
      "epoch": 2.1726287415061494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045348290980565595,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20029636919498445,
      "num_tokens": 43177394.0,
      "step": 19345
    },
    {
      "entropy": 5.34565486907959,
      "epoch": 2.1731903184141066,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045345846095248376,
      "loss": 5.033,
      "mean_token_accuracy": 0.20690608024597168,
      "num_tokens": 43188043.0,
      "step": 19350
    },
    {
      "entropy": 5.399140167236328,
      "epoch": 2.1737518953220643,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045343400641712503,
      "loss": 5.223,
      "mean_token_accuracy": 0.19903456568717956,
      "num_tokens": 43199466.0,
      "step": 19355
    },
    {
      "entropy": 5.291401958465576,
      "epoch": 2.174313472230022,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045340954620035827,
      "loss": 5.041,
      "mean_token_accuracy": 0.20945289880037307,
      "num_tokens": 43210673.0,
      "step": 19360
    },
    {
      "entropy": 5.373338937759399,
      "epoch": 2.1748750491379796,
      "grad_norm": 1.265625,
      "learning_rate": 0.00045338508030296246,
      "loss": 5.1413,
      "mean_token_accuracy": 0.2028835415840149,
      "num_tokens": 43221263.0,
      "step": 19365
    },
    {
      "entropy": 5.445145750045777,
      "epoch": 2.175436626045937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045336060872571646,
      "loss": 5.1452,
      "mean_token_accuracy": 0.20750940591096878,
      "num_tokens": 43232590.0,
      "step": 19370
    },
    {
      "entropy": 5.432615900039673,
      "epoch": 2.1759982029538945,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004533361314693994,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20465612709522246,
      "num_tokens": 43243693.0,
      "step": 19375
    },
    {
      "entropy": 5.369443607330322,
      "epoch": 2.176559779861852,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045331164853479084,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20119786709547044,
      "num_tokens": 43254526.0,
      "step": 19380
    },
    {
      "entropy": 5.482165908813476,
      "epoch": 2.17712135676981,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00045328715992267024,
      "loss": 5.1699,
      "mean_token_accuracy": 0.20255340039730071,
      "num_tokens": 43264490.0,
      "step": 19385
    },
    {
      "entropy": 5.335905981063843,
      "epoch": 2.177682933677767,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045326266563381734,
      "loss": 5.0744,
      "mean_token_accuracy": 0.20453088730573654,
      "num_tokens": 43275012.0,
      "step": 19390
    },
    {
      "entropy": 5.373665809631348,
      "epoch": 2.1782445105857247,
      "grad_norm": 1.09375,
      "learning_rate": 0.00045323816566901205,
      "loss": 5.0717,
      "mean_token_accuracy": 0.20717393457889557,
      "num_tokens": 43285194.0,
      "step": 19395
    },
    {
      "entropy": 5.394247961044312,
      "epoch": 2.1788060874936823,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004532136600290343,
      "loss": 5.1573,
      "mean_token_accuracy": 0.20441700518131256,
      "num_tokens": 43298159.0,
      "step": 19400
    },
    {
      "entropy": 5.421025085449219,
      "epoch": 2.17936766440164,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004531891487146646,
      "loss": 5.1896,
      "mean_token_accuracy": 0.1968286395072937,
      "num_tokens": 43309664.0,
      "step": 19405
    },
    {
      "entropy": 5.401861763000488,
      "epoch": 2.179929241309597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004531646317266833,
      "loss": 5.148,
      "mean_token_accuracy": 0.20249048471450806,
      "num_tokens": 43320399.0,
      "step": 19410
    },
    {
      "entropy": 5.362013149261474,
      "epoch": 2.180490818217555,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045314010906587096,
      "loss": 4.9774,
      "mean_token_accuracy": 0.20451103746891022,
      "num_tokens": 43331510.0,
      "step": 19415
    },
    {
      "entropy": 5.417592191696167,
      "epoch": 2.1810523951255125,
      "grad_norm": 1.125,
      "learning_rate": 0.00045311558073300853,
      "loss": 5.1927,
      "mean_token_accuracy": 0.20606657415628432,
      "num_tokens": 43342844.0,
      "step": 19420
    },
    {
      "entropy": 5.326347398757934,
      "epoch": 2.18161397203347,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004530910467288769,
      "loss": 5.0493,
      "mean_token_accuracy": 0.2047366663813591,
      "num_tokens": 43352556.0,
      "step": 19425
    },
    {
      "entropy": 5.335011720657349,
      "epoch": 2.1821755489414274,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045306650705425727,
      "loss": 5.0249,
      "mean_token_accuracy": 0.21149512082338334,
      "num_tokens": 43363357.0,
      "step": 19430
    },
    {
      "entropy": 5.400990533828735,
      "epoch": 2.182737125849385,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045304196170993093,
      "loss": 5.1052,
      "mean_token_accuracy": 0.20090225636959075,
      "num_tokens": 43374697.0,
      "step": 19435
    },
    {
      "entropy": 5.439014291763305,
      "epoch": 2.1832987027573427,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004530174106966795,
      "loss": 5.1927,
      "mean_token_accuracy": 0.19984963089227675,
      "num_tokens": 43386826.0,
      "step": 19440
    },
    {
      "entropy": 5.378395462036133,
      "epoch": 2.1838602796653,
      "grad_norm": 1.125,
      "learning_rate": 0.00045299285401528475,
      "loss": 5.1514,
      "mean_token_accuracy": 0.20468613654375076,
      "num_tokens": 43398352.0,
      "step": 19445
    },
    {
      "entropy": 5.39046721458435,
      "epoch": 2.1844218565732576,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004529682916665285,
      "loss": 5.0938,
      "mean_token_accuracy": 0.20705513656139374,
      "num_tokens": 43409509.0,
      "step": 19450
    },
    {
      "entropy": 5.324642896652222,
      "epoch": 2.1849834334812153,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045294372365119276,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20113499611616134,
      "num_tokens": 43420195.0,
      "step": 19455
    },
    {
      "entropy": 5.392933559417725,
      "epoch": 2.185545010389173,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004529191499700598,
      "loss": 5.1545,
      "mean_token_accuracy": 0.20458621978759767,
      "num_tokens": 43430599.0,
      "step": 19460
    },
    {
      "entropy": 5.326816844940185,
      "epoch": 2.18610658729713,
      "grad_norm": 1.875,
      "learning_rate": 0.0004528945706239122,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2033589869737625,
      "num_tokens": 43442212.0,
      "step": 19465
    },
    {
      "entropy": 5.334196901321411,
      "epoch": 2.186668164205088,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045286998561353236,
      "loss": 5.0422,
      "mean_token_accuracy": 0.20268322974443437,
      "num_tokens": 43452631.0,
      "step": 19470
    },
    {
      "entropy": 5.398529100418091,
      "epoch": 2.1872297411130455,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004528453949397033,
      "loss": 5.1037,
      "mean_token_accuracy": 0.20496261417865752,
      "num_tokens": 43463323.0,
      "step": 19475
    },
    {
      "entropy": 5.3585466861724855,
      "epoch": 2.187791318021003,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045282079860320786,
      "loss": 5.146,
      "mean_token_accuracy": 0.1944469153881073,
      "num_tokens": 43474451.0,
      "step": 19480
    },
    {
      "entropy": 5.3097704410552975,
      "epoch": 2.1883528949289603,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045279619660482925,
      "loss": 5.0154,
      "mean_token_accuracy": 0.2084156483411789,
      "num_tokens": 43485948.0,
      "step": 19485
    },
    {
      "entropy": 5.416886806488037,
      "epoch": 2.188914471836918,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045277158894535075,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20817996859550475,
      "num_tokens": 43498955.0,
      "step": 19490
    },
    {
      "entropy": 5.394818067550659,
      "epoch": 2.1894760487448757,
      "grad_norm": 1.15625,
      "learning_rate": 0.00045274697562555584,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20979901105165483,
      "num_tokens": 43509416.0,
      "step": 19495
    },
    {
      "entropy": 5.336586952209473,
      "epoch": 2.1900376256528333,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004527223566462284,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20413439124822616,
      "num_tokens": 43520387.0,
      "step": 19500
    },
    {
      "entropy": 5.302551651000977,
      "epoch": 2.1905992025607905,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004526977320081522,
      "loss": 5.0925,
      "mean_token_accuracy": 0.20479379296302797,
      "num_tokens": 43532006.0,
      "step": 19505
    },
    {
      "entropy": 5.313959646224975,
      "epoch": 2.191160779468748,
      "grad_norm": 1.125,
      "learning_rate": 0.0004526731017121111,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2114452600479126,
      "num_tokens": 43542303.0,
      "step": 19510
    },
    {
      "entropy": 5.426939916610718,
      "epoch": 2.191722356376706,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00045264846575888973,
      "loss": 5.1966,
      "mean_token_accuracy": 0.1989064946770668,
      "num_tokens": 43553711.0,
      "step": 19515
    },
    {
      "entropy": 5.290425586700439,
      "epoch": 2.1922839332846635,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045262382414927235,
      "loss": 5.0395,
      "mean_token_accuracy": 0.20560196191072463,
      "num_tokens": 43564488.0,
      "step": 19520
    },
    {
      "entropy": 5.35864291191101,
      "epoch": 2.1928455101926208,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00045259917688404336,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20329554229974747,
      "num_tokens": 43575414.0,
      "step": 19525
    },
    {
      "entropy": 5.313883113861084,
      "epoch": 2.1934070871005784,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004525745239639877,
      "loss": 5.0453,
      "mean_token_accuracy": 0.2041175529360771,
      "num_tokens": 43586789.0,
      "step": 19530
    },
    {
      "entropy": 5.315013122558594,
      "epoch": 2.193968664008536,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045254986538989045,
      "loss": 5.0366,
      "mean_token_accuracy": 0.21320376843214034,
      "num_tokens": 43597423.0,
      "step": 19535
    },
    {
      "entropy": 5.354296493530273,
      "epoch": 2.1945302409164933,
      "grad_norm": 1.125,
      "learning_rate": 0.0004525252011625366,
      "loss": 5.048,
      "mean_token_accuracy": 0.20356505662202834,
      "num_tokens": 43606703.0,
      "step": 19540
    },
    {
      "entropy": 5.391881132125855,
      "epoch": 2.195091817824451,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004525005312827114,
      "loss": 5.1942,
      "mean_token_accuracy": 0.1980681985616684,
      "num_tokens": 43617869.0,
      "step": 19545
    },
    {
      "entropy": 5.49667911529541,
      "epoch": 2.1956533947324086,
      "grad_norm": 1.0,
      "learning_rate": 0.0004524758557512005,
      "loss": 5.2404,
      "mean_token_accuracy": 0.20164641737937927,
      "num_tokens": 43629102.0,
      "step": 19550
    },
    {
      "entropy": 5.426077079772949,
      "epoch": 2.1962149716403663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004524511745687895,
      "loss": 5.1447,
      "mean_token_accuracy": 0.20592319518327712,
      "num_tokens": 43640421.0,
      "step": 19555
    },
    {
      "entropy": 5.389101362228393,
      "epoch": 2.1967765485483235,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004524264877362643,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20764122605323793,
      "num_tokens": 43650516.0,
      "step": 19560
    },
    {
      "entropy": 5.352014350891113,
      "epoch": 2.197338125456281,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045240179525441094,
      "loss": 5.073,
      "mean_token_accuracy": 0.21349542289972306,
      "num_tokens": 43661439.0,
      "step": 19565
    },
    {
      "entropy": 5.329970359802246,
      "epoch": 2.197899702364239,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004523770971240156,
      "loss": 5.1507,
      "mean_token_accuracy": 0.20773282051086425,
      "num_tokens": 43672137.0,
      "step": 19570
    },
    {
      "entropy": 5.392208003997803,
      "epoch": 2.1984612792721965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004523523933458647,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20392244905233384,
      "num_tokens": 43683612.0,
      "step": 19575
    },
    {
      "entropy": 5.358365488052368,
      "epoch": 2.1990228561801537,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004523276839207448,
      "loss": 5.0673,
      "mean_token_accuracy": 0.2079744428396225,
      "num_tokens": 43694331.0,
      "step": 19580
    },
    {
      "entropy": 5.413495016098023,
      "epoch": 2.1995844330881114,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045230296884944264,
      "loss": 5.1776,
      "mean_token_accuracy": 0.19865314811468124,
      "num_tokens": 43706572.0,
      "step": 19585
    },
    {
      "entropy": 5.369610548019409,
      "epoch": 2.200146009996069,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045227824813274517,
      "loss": 5.0546,
      "mean_token_accuracy": 0.21129221469163895,
      "num_tokens": 43718472.0,
      "step": 19590
    },
    {
      "entropy": 5.359939241409302,
      "epoch": 2.2007075869040267,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004522535217714396,
      "loss": 5.1536,
      "mean_token_accuracy": 0.20403508394956588,
      "num_tokens": 43729642.0,
      "step": 19595
    },
    {
      "entropy": 5.370769739151001,
      "epoch": 2.201269163811984,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004522287897663131,
      "loss": 5.1598,
      "mean_token_accuracy": 0.19789824783802032,
      "num_tokens": 43739846.0,
      "step": 19600
    },
    {
      "entropy": 5.3660805225372314,
      "epoch": 2.2018307407199416,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045220405211815327,
      "loss": 5.0025,
      "mean_token_accuracy": 0.20831260979175567,
      "num_tokens": 43750265.0,
      "step": 19605
    },
    {
      "entropy": 5.420605897903442,
      "epoch": 2.202392317627899,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045217930882774754,
      "loss": 5.0997,
      "mean_token_accuracy": 0.20011068284511566,
      "num_tokens": 43761245.0,
      "step": 19610
    },
    {
      "entropy": 5.4113771438598635,
      "epoch": 2.202953894535857,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045215455989588396,
      "loss": 5.1101,
      "mean_token_accuracy": 0.2053752437233925,
      "num_tokens": 43772677.0,
      "step": 19615
    },
    {
      "entropy": 5.393328142166138,
      "epoch": 2.203515471443814,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045212980532335055,
      "loss": 5.1856,
      "mean_token_accuracy": 0.20019663870334625,
      "num_tokens": 43783128.0,
      "step": 19620
    },
    {
      "entropy": 5.335208940505981,
      "epoch": 2.2040770483517718,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045210504511093534,
      "loss": 5.1159,
      "mean_token_accuracy": 0.20663616955280303,
      "num_tokens": 43795323.0,
      "step": 19625
    },
    {
      "entropy": 5.456618356704712,
      "epoch": 2.2046386252597294,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004520802792594269,
      "loss": 5.1846,
      "mean_token_accuracy": 0.199416783452034,
      "num_tokens": 43806917.0,
      "step": 19630
    },
    {
      "entropy": 5.483849906921387,
      "epoch": 2.2052002021676866,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045205550776961366,
      "loss": 5.1837,
      "mean_token_accuracy": 0.19885862320661546,
      "num_tokens": 43818963.0,
      "step": 19635
    },
    {
      "entropy": 5.2997434616088865,
      "epoch": 2.2057617790756443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004520307306422844,
      "loss": 5.0396,
      "mean_token_accuracy": 0.20195687115192412,
      "num_tokens": 43830163.0,
      "step": 19640
    },
    {
      "entropy": 5.36118974685669,
      "epoch": 2.206323355983602,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00045200594787822795,
      "loss": 5.0472,
      "mean_token_accuracy": 0.20602291822433472,
      "num_tokens": 43840434.0,
      "step": 19645
    },
    {
      "entropy": 5.263969039916992,
      "epoch": 2.2068849328915596,
      "grad_norm": 1.0,
      "learning_rate": 0.0004519811594782335,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21548182815313338,
      "num_tokens": 43851086.0,
      "step": 19650
    },
    {
      "entropy": 5.290554857254028,
      "epoch": 2.207446509799517,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045195636544309033,
      "loss": 5.1428,
      "mean_token_accuracy": 0.19613641500473022,
      "num_tokens": 43862834.0,
      "step": 19655
    },
    {
      "entropy": 5.323722648620605,
      "epoch": 2.2080080867074745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045193156577358784,
      "loss": 5.1284,
      "mean_token_accuracy": 0.1946035608649254,
      "num_tokens": 43874868.0,
      "step": 19660
    },
    {
      "entropy": 5.416198492050171,
      "epoch": 2.208569663615432,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004519067604705157,
      "loss": 5.1028,
      "mean_token_accuracy": 0.21198291033506395,
      "num_tokens": 43886175.0,
      "step": 19665
    },
    {
      "entropy": 5.344945955276489,
      "epoch": 2.20913124052339,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045188194953466363,
      "loss": 5.0767,
      "mean_token_accuracy": 0.20829330682754515,
      "num_tokens": 43896489.0,
      "step": 19670
    },
    {
      "entropy": 5.324801588058472,
      "epoch": 2.209692817431347,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045185713296682175,
      "loss": 5.0667,
      "mean_token_accuracy": 0.20812975168228148,
      "num_tokens": 43907759.0,
      "step": 19675
    },
    {
      "entropy": 5.364258432388306,
      "epoch": 2.2102543943393047,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045183231076778005,
      "loss": 5.0739,
      "mean_token_accuracy": 0.20600459724664688,
      "num_tokens": 43918136.0,
      "step": 19680
    },
    {
      "entropy": 5.387272834777832,
      "epoch": 2.2108159712472624,
      "grad_norm": 1.0625,
      "learning_rate": 0.00045180748293832906,
      "loss": 5.1362,
      "mean_token_accuracy": 0.19692451357841492,
      "num_tokens": 43929662.0,
      "step": 19685
    },
    {
      "entropy": 5.3353719234466555,
      "epoch": 2.21137754815522,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00045178264947925924,
      "loss": 5.0981,
      "mean_token_accuracy": 0.21062570661306382,
      "num_tokens": 43939599.0,
      "step": 19690
    },
    {
      "entropy": 5.35045371055603,
      "epoch": 2.2119391250631772,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045175781039136133,
      "loss": 5.0469,
      "mean_token_accuracy": 0.20328034162521363,
      "num_tokens": 43949728.0,
      "step": 19695
    },
    {
      "entropy": 5.447231912612915,
      "epoch": 2.212500701971135,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004517329656754261,
      "loss": 5.1643,
      "mean_token_accuracy": 0.2044721633195877,
      "num_tokens": 43961147.0,
      "step": 19700
    },
    {
      "entropy": 5.426426267623901,
      "epoch": 2.2130622788790926,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004517081153322448,
      "loss": 5.145,
      "mean_token_accuracy": 0.20442102551460267,
      "num_tokens": 43972541.0,
      "step": 19705
    },
    {
      "entropy": 5.330085802078247,
      "epoch": 2.2136238557870502,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004516832593626085,
      "loss": 5.1502,
      "mean_token_accuracy": 0.2003696918487549,
      "num_tokens": 43983540.0,
      "step": 19710
    },
    {
      "entropy": 5.3000773906707765,
      "epoch": 2.2141854326950074,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004516583977673087,
      "loss": 5.0217,
      "mean_token_accuracy": 0.21495030373334884,
      "num_tokens": 43994255.0,
      "step": 19715
    },
    {
      "entropy": 5.448230648040772,
      "epoch": 2.214747009602965,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045163353054713704,
      "loss": 5.1942,
      "mean_token_accuracy": 0.2009427309036255,
      "num_tokens": 44004364.0,
      "step": 19720
    },
    {
      "entropy": 5.360700273513794,
      "epoch": 2.2153085865109228,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004516086577028852,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20249146223068237,
      "num_tokens": 44015487.0,
      "step": 19725
    },
    {
      "entropy": 5.445820569992065,
      "epoch": 2.21587016341888,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004515837792353453,
      "loss": 5.1451,
      "mean_token_accuracy": 0.20734512954950332,
      "num_tokens": 44026974.0,
      "step": 19730
    },
    {
      "entropy": 5.398342323303223,
      "epoch": 2.2164317403268377,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004515588951453093,
      "loss": 5.0968,
      "mean_token_accuracy": 0.2005889296531677,
      "num_tokens": 44038718.0,
      "step": 19735
    },
    {
      "entropy": 5.309185647964478,
      "epoch": 2.2169933172347953,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004515340054335696,
      "loss": 5.0344,
      "mean_token_accuracy": 0.20823158621788024,
      "num_tokens": 44048821.0,
      "step": 19740
    },
    {
      "entropy": 5.382176685333252,
      "epoch": 2.217554894142753,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004515091101009188,
      "loss": 5.1941,
      "mean_token_accuracy": 0.19722867459058763,
      "num_tokens": 44060973.0,
      "step": 19745
    },
    {
      "entropy": 5.506192588806153,
      "epoch": 2.21811647105071,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004514842091481493,
      "loss": 5.1572,
      "mean_token_accuracy": 0.20578921735286712,
      "num_tokens": 44073595.0,
      "step": 19750
    },
    {
      "entropy": 5.273137712478638,
      "epoch": 2.218678047958668,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004514593025760543,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20357028245925904,
      "num_tokens": 44084860.0,
      "step": 19755
    },
    {
      "entropy": 5.377425861358643,
      "epoch": 2.2192396248666255,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045143439038542653,
      "loss": 5.1767,
      "mean_token_accuracy": 0.2013503357768059,
      "num_tokens": 44096026.0,
      "step": 19760
    },
    {
      "entropy": 5.456210136413574,
      "epoch": 2.219801201774583,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004514094725770594,
      "loss": 5.0861,
      "mean_token_accuracy": 0.200937819480896,
      "num_tokens": 44107981.0,
      "step": 19765
    },
    {
      "entropy": 5.338053417205811,
      "epoch": 2.2203627786825404,
      "grad_norm": 1.140625,
      "learning_rate": 0.00045138454915174625,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20015855878591537,
      "num_tokens": 44119589.0,
      "step": 19770
    },
    {
      "entropy": 5.332802581787109,
      "epoch": 2.220924355590498,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045135962011028064,
      "loss": 5.1509,
      "mean_token_accuracy": 0.1982612982392311,
      "num_tokens": 44131094.0,
      "step": 19775
    },
    {
      "entropy": 5.404224967956543,
      "epoch": 2.2214859324984557,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004513346854534564,
      "loss": 5.1277,
      "mean_token_accuracy": 0.19807063937187194,
      "num_tokens": 44141348.0,
      "step": 19780
    },
    {
      "entropy": 5.436205863952637,
      "epoch": 2.2220475094064134,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004513097451820673,
      "loss": 5.2522,
      "mean_token_accuracy": 0.19823910742998124,
      "num_tokens": 44152085.0,
      "step": 19785
    },
    {
      "entropy": 5.328860902786255,
      "epoch": 2.2226090863143706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045128479929690756,
      "loss": 5.0329,
      "mean_token_accuracy": 0.21555811166763306,
      "num_tokens": 44163518.0,
      "step": 19790
    },
    {
      "entropy": 5.359676456451416,
      "epoch": 2.2231706632223283,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004512598477987714,
      "loss": 5.0936,
      "mean_token_accuracy": 0.20120801776647568,
      "num_tokens": 44175513.0,
      "step": 19795
    },
    {
      "entropy": 5.435778951644897,
      "epoch": 2.223732240130286,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045123489068845334,
      "loss": 5.0984,
      "mean_token_accuracy": 0.20364794433116912,
      "num_tokens": 44186660.0,
      "step": 19800
    },
    {
      "entropy": 5.384709167480469,
      "epoch": 2.2242938170382436,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045120992796674804,
      "loss": 5.1591,
      "mean_token_accuracy": 0.20311979800462723,
      "num_tokens": 44197613.0,
      "step": 19805
    },
    {
      "entropy": 5.327611017227173,
      "epoch": 2.224855393946201,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004511849596344503,
      "loss": 5.0929,
      "mean_token_accuracy": 0.1949809193611145,
      "num_tokens": 44209469.0,
      "step": 19810
    },
    {
      "entropy": 5.3979620933532715,
      "epoch": 2.2254169708541585,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045115998569235507,
      "loss": 5.1106,
      "mean_token_accuracy": 0.20224408209323883,
      "num_tokens": 44220672.0,
      "step": 19815
    },
    {
      "entropy": 5.404706764221191,
      "epoch": 2.225978547762116,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00045113500614125756,
      "loss": 5.1353,
      "mean_token_accuracy": 0.1975605458021164,
      "num_tokens": 44232867.0,
      "step": 19820
    },
    {
      "entropy": 5.3518811702728275,
      "epoch": 2.2265401246700733,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004511100209819532,
      "loss": 5.1349,
      "mean_token_accuracy": 0.20274464040994644,
      "num_tokens": 44245567.0,
      "step": 19825
    },
    {
      "entropy": 5.368127870559692,
      "epoch": 2.227101701578031,
      "grad_norm": 1.28125,
      "learning_rate": 0.00045108503021523737,
      "loss": 5.079,
      "mean_token_accuracy": 0.21095803529024124,
      "num_tokens": 44257449.0,
      "step": 19830
    },
    {
      "entropy": 5.392900133132935,
      "epoch": 2.2276632784859887,
      "grad_norm": 1.109375,
      "learning_rate": 0.00045106003384190604,
      "loss": 5.158,
      "mean_token_accuracy": 0.20273813903331755,
      "num_tokens": 44268890.0,
      "step": 19835
    },
    {
      "entropy": 5.3991600513458256,
      "epoch": 2.2282248553939463,
      "grad_norm": 1.2734375,
      "learning_rate": 0.00045103503186275486,
      "loss": 5.1235,
      "mean_token_accuracy": 0.19733493626117707,
      "num_tokens": 44280719.0,
      "step": 19840
    },
    {
      "entropy": 5.406705570220947,
      "epoch": 2.2287864323019035,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045101002427857997,
      "loss": 5.1985,
      "mean_token_accuracy": 0.19442322105169296,
      "num_tokens": 44292567.0,
      "step": 19845
    },
    {
      "entropy": 5.426033639907837,
      "epoch": 2.229348009209861,
      "grad_norm": 1.171875,
      "learning_rate": 0.00045098501109017765,
      "loss": 5.1626,
      "mean_token_accuracy": 0.19541232883930207,
      "num_tokens": 44304177.0,
      "step": 19850
    },
    {
      "entropy": 5.3671142578125,
      "epoch": 2.229909586117819,
      "grad_norm": 1.1875,
      "learning_rate": 0.00045095999229834437,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2114976167678833,
      "num_tokens": 44315142.0,
      "step": 19855
    },
    {
      "entropy": 5.344931602478027,
      "epoch": 2.2304711630257765,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004509349679038767,
      "loss": 5.2031,
      "mean_token_accuracy": 0.20557105392217637,
      "num_tokens": 44327457.0,
      "step": 19860
    },
    {
      "entropy": 5.4604042053222654,
      "epoch": 2.2310327399337337,
      "grad_norm": 1.203125,
      "learning_rate": 0.00045090993790757137,
      "loss": 5.1912,
      "mean_token_accuracy": 0.1956818789243698,
      "num_tokens": 44338978.0,
      "step": 19865
    },
    {
      "entropy": 5.432954740524292,
      "epoch": 2.2315943168416914,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004508849023102253,
      "loss": 5.1877,
      "mean_token_accuracy": 0.201189324259758,
      "num_tokens": 44349121.0,
      "step": 19870
    },
    {
      "entropy": 5.328174829483032,
      "epoch": 2.232155893749649,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004508598611126359,
      "loss": 5.0698,
      "mean_token_accuracy": 0.20756586492061616,
      "num_tokens": 44360753.0,
      "step": 19875
    },
    {
      "entropy": 5.439252185821533,
      "epoch": 2.2327174706576067,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004508348143156002,
      "loss": 5.1405,
      "mean_token_accuracy": 0.20391618758440017,
      "num_tokens": 44371977.0,
      "step": 19880
    },
    {
      "entropy": 5.358951044082642,
      "epoch": 2.233279047565564,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00045080976191991587,
      "loss": 5.056,
      "mean_token_accuracy": 0.2025028347969055,
      "num_tokens": 44383077.0,
      "step": 19885
    },
    {
      "entropy": 5.354318141937256,
      "epoch": 2.2338406244735216,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004507847039263805,
      "loss": 5.0823,
      "mean_token_accuracy": 0.20178060680627824,
      "num_tokens": 44394827.0,
      "step": 19890
    },
    {
      "entropy": 5.31257586479187,
      "epoch": 2.2344022013814793,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000450759640335792,
      "loss": 5.0339,
      "mean_token_accuracy": 0.2095889165997505,
      "num_tokens": 44406594.0,
      "step": 19895
    },
    {
      "entropy": 5.313387823104859,
      "epoch": 2.234963778289437,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004507345711489484,
      "loss": 5.0308,
      "mean_token_accuracy": 0.20813892632722855,
      "num_tokens": 44417842.0,
      "step": 19900
    },
    {
      "entropy": 5.397504663467407,
      "epoch": 2.235525355197394,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045070949636664785,
      "loss": 5.0849,
      "mean_token_accuracy": 0.20354894399642945,
      "num_tokens": 44428234.0,
      "step": 19905
    },
    {
      "entropy": 5.300754404067993,
      "epoch": 2.236086932105352,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004506844159896888,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20825165659189224,
      "num_tokens": 44437504.0,
      "step": 19910
    },
    {
      "entropy": 5.341244459152222,
      "epoch": 2.2366485090133095,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004506593300188698,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20614471286535263,
      "num_tokens": 44447873.0,
      "step": 19915
    },
    {
      "entropy": 5.3794753074646,
      "epoch": 2.2372100859212667,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004506342384549897,
      "loss": 5.1599,
      "mean_token_accuracy": 0.20417623221874237,
      "num_tokens": 44458990.0,
      "step": 19920
    },
    {
      "entropy": 5.341534519195557,
      "epoch": 2.2377716628292244,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004506091412988472,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20220936834812164,
      "num_tokens": 44469374.0,
      "step": 19925
    },
    {
      "entropy": 5.381109189987183,
      "epoch": 2.238333239737182,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045058403855124165,
      "loss": 5.0862,
      "mean_token_accuracy": 0.20963959246873856,
      "num_tokens": 44479629.0,
      "step": 19930
    },
    {
      "entropy": 5.453382539749145,
      "epoch": 2.2388948166451397,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004505589302129721,
      "loss": 5.1692,
      "mean_token_accuracy": 0.202591173350811,
      "num_tokens": 44490603.0,
      "step": 19935
    },
    {
      "entropy": 5.270648097991943,
      "epoch": 2.239456393553097,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004505338162848382,
      "loss": 5.0221,
      "mean_token_accuracy": 0.2111395612359047,
      "num_tokens": 44501527.0,
      "step": 19940
    },
    {
      "entropy": 5.382872438430786,
      "epoch": 2.2400179704610546,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004505086967676395,
      "loss": 5.2318,
      "mean_token_accuracy": 0.20090459436178207,
      "num_tokens": 44512619.0,
      "step": 19945
    },
    {
      "entropy": 5.519139194488526,
      "epoch": 2.240579547369012,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004504835716621757,
      "loss": 5.2381,
      "mean_token_accuracy": 0.19484602212905883,
      "num_tokens": 44524639.0,
      "step": 19950
    },
    {
      "entropy": 5.323319053649902,
      "epoch": 2.24114112427697,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004504584409692471,
      "loss": 5.0797,
      "mean_token_accuracy": 0.2111170247197151,
      "num_tokens": 44535013.0,
      "step": 19955
    },
    {
      "entropy": 5.48799500465393,
      "epoch": 2.241702701184927,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00045043330468965364,
      "loss": 5.1801,
      "mean_token_accuracy": 0.20255401879549026,
      "num_tokens": 44546838.0,
      "step": 19960
    },
    {
      "entropy": 5.371854686737061,
      "epoch": 2.2422642780928848,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045040816282419566,
      "loss": 5.0517,
      "mean_token_accuracy": 0.2095311999320984,
      "num_tokens": 44557984.0,
      "step": 19965
    },
    {
      "entropy": 5.301491069793701,
      "epoch": 2.2428258550008424,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00045038301537367374,
      "loss": 5.038,
      "mean_token_accuracy": 0.2070087045431137,
      "num_tokens": 44569029.0,
      "step": 19970
    },
    {
      "entropy": 5.374565362930298,
      "epoch": 2.2433874319088,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00045035786233888864,
      "loss": 5.1263,
      "mean_token_accuracy": 0.2037488654255867,
      "num_tokens": 44578804.0,
      "step": 19975
    },
    {
      "entropy": 5.384415769577027,
      "epoch": 2.2439490088167573,
      "grad_norm": 1.078125,
      "learning_rate": 0.00045033270372064123,
      "loss": 5.0535,
      "mean_token_accuracy": 0.20555941760540009,
      "num_tokens": 44589362.0,
      "step": 19980
    },
    {
      "entropy": 5.350485324859619,
      "epoch": 2.244510585724715,
      "grad_norm": 1.046875,
      "learning_rate": 0.00045030753951973243,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19952967315912246,
      "num_tokens": 44600137.0,
      "step": 19985
    },
    {
      "entropy": 5.395564842224121,
      "epoch": 2.2450721626326726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045028236973696363,
      "loss": 5.1014,
      "mean_token_accuracy": 0.21357431411743164,
      "num_tokens": 44611707.0,
      "step": 19990
    },
    {
      "entropy": 5.346336603164673,
      "epoch": 2.2456337395406303,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004502571943731362,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20618945360183716,
      "num_tokens": 44621141.0,
      "step": 19995
    },
    {
      "entropy": 5.354413080215454,
      "epoch": 2.2461953164485875,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004502320134290517,
      "loss": 5.1663,
      "mean_token_accuracy": 0.2022166520357132,
      "num_tokens": 44632043.0,
      "step": 20000
    },
    {
      "entropy": 5.433373689651489,
      "epoch": 2.246756893356545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00045020682690551203,
      "loss": 5.1719,
      "mean_token_accuracy": 0.19867394715547562,
      "num_tokens": 44643905.0,
      "step": 20005
    },
    {
      "entropy": 5.354638767242432,
      "epoch": 2.247318470264503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00045018163480331887,
      "loss": 5.092,
      "mean_token_accuracy": 0.19866699278354644,
      "num_tokens": 44655105.0,
      "step": 20010
    },
    {
      "entropy": 5.321398401260376,
      "epoch": 2.24788004717246,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00045015643712327465,
      "loss": 5.0607,
      "mean_token_accuracy": 0.20430415272712707,
      "num_tokens": 44667096.0,
      "step": 20015
    },
    {
      "entropy": 5.421066999435425,
      "epoch": 2.2484416240804177,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004501312338661815,
      "loss": 5.1125,
      "mean_token_accuracy": 0.20775338858366013,
      "num_tokens": 44678206.0,
      "step": 20020
    },
    {
      "entropy": 5.301611232757568,
      "epoch": 2.2490032009883754,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004501060250328419,
      "loss": 5.0415,
      "mean_token_accuracy": 0.21100276857614517,
      "num_tokens": 44688619.0,
      "step": 20025
    },
    {
      "entropy": 5.336159944534302,
      "epoch": 2.249564777896333,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004500808106240586,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20205643624067307,
      "num_tokens": 44700757.0,
      "step": 20030
    },
    {
      "entropy": 5.25946569442749,
      "epoch": 2.2501263548042907,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004500555906406343,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2110808774828911,
      "num_tokens": 44710822.0,
      "step": 20035
    },
    {
      "entropy": 5.249242115020752,
      "epoch": 2.250687931712248,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004500303650833722,
      "loss": 4.9749,
      "mean_token_accuracy": 0.21068553030490875,
      "num_tokens": 44722163.0,
      "step": 20040
    },
    {
      "entropy": 5.465045642852783,
      "epoch": 2.2512495086202056,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004500051339530754,
      "loss": 5.1433,
      "mean_token_accuracy": 0.20191844403743744,
      "num_tokens": 44734271.0,
      "step": 20045
    },
    {
      "entropy": 5.450015640258789,
      "epoch": 2.2518110855281632,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004499798972505472,
      "loss": 5.1054,
      "mean_token_accuracy": 0.2066001459956169,
      "num_tokens": 44745345.0,
      "step": 20050
    },
    {
      "entropy": 5.291317558288574,
      "epoch": 2.2523726624361204,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044995465497659127,
      "loss": 5.0829,
      "mean_token_accuracy": 0.2002398267388344,
      "num_tokens": 44756986.0,
      "step": 20055
    },
    {
      "entropy": 5.267534399032593,
      "epoch": 2.252934239344078,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004499294071320112,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21111230552196503,
      "num_tokens": 44767259.0,
      "step": 20060
    },
    {
      "entropy": 5.306210947036743,
      "epoch": 2.2534958162520358,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000449904153717611,
      "loss": 5.1415,
      "mean_token_accuracy": 0.2020959287881851,
      "num_tokens": 44778884.0,
      "step": 20065
    },
    {
      "entropy": 5.383560037612915,
      "epoch": 2.2540573931599934,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004498788947341947,
      "loss": 5.0274,
      "mean_token_accuracy": 0.21112231761217118,
      "num_tokens": 44789216.0,
      "step": 20070
    },
    {
      "entropy": 5.259118604660034,
      "epoch": 2.2546189700679506,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004498536301825665,
      "loss": 4.9797,
      "mean_token_accuracy": 0.2160826653242111,
      "num_tokens": 44799357.0,
      "step": 20075
    },
    {
      "entropy": 5.292740345001221,
      "epoch": 2.2551805469759083,
      "grad_norm": 1.046875,
      "learning_rate": 0.000449828360063531,
      "loss": 5.1188,
      "mean_token_accuracy": 0.2020603522658348,
      "num_tokens": 44811040.0,
      "step": 20080
    },
    {
      "entropy": 5.424952650070191,
      "epoch": 2.255742123883866,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044980308437789267,
      "loss": 5.1848,
      "mean_token_accuracy": 0.19354504048824311,
      "num_tokens": 44822103.0,
      "step": 20085
    },
    {
      "entropy": 5.370583724975586,
      "epoch": 2.2563037007918236,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004497778031264564,
      "loss": 5.099,
      "mean_token_accuracy": 0.20245604664087297,
      "num_tokens": 44834712.0,
      "step": 20090
    },
    {
      "entropy": 5.352909231185913,
      "epoch": 2.256865277699781,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044975251631002694,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20266429781913758,
      "num_tokens": 44845412.0,
      "step": 20095
    },
    {
      "entropy": 5.315562152862549,
      "epoch": 2.2574268546077385,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004497272239294097,
      "loss": 5.0962,
      "mean_token_accuracy": 0.20537223666906357,
      "num_tokens": 44857720.0,
      "step": 20100
    },
    {
      "entropy": 5.363692808151245,
      "epoch": 2.257988431515696,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004497019259854099,
      "loss": 5.0783,
      "mean_token_accuracy": 0.20686988085508345,
      "num_tokens": 44868253.0,
      "step": 20105
    },
    {
      "entropy": 5.362688875198364,
      "epoch": 2.2585500084236534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044967662247883295,
      "loss": 5.07,
      "mean_token_accuracy": 0.20320221036672592,
      "num_tokens": 44878525.0,
      "step": 20110
    },
    {
      "entropy": 5.258663511276245,
      "epoch": 2.259111585331611,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044965131341048454,
      "loss": 5.0383,
      "mean_token_accuracy": 0.2083526760339737,
      "num_tokens": 44889565.0,
      "step": 20115
    },
    {
      "entropy": 5.301581478118896,
      "epoch": 2.2596731622395687,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004496259987811707,
      "loss": 5.0188,
      "mean_token_accuracy": 0.20787357091903685,
      "num_tokens": 44901191.0,
      "step": 20120
    },
    {
      "entropy": 5.308367681503296,
      "epoch": 2.2602347391475264,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004496006785916973,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21561262756586075,
      "num_tokens": 44911483.0,
      "step": 20125
    },
    {
      "entropy": 5.395339632034302,
      "epoch": 2.260796316055484,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044957535284287055,
      "loss": 5.1769,
      "mean_token_accuracy": 0.20435143560171126,
      "num_tokens": 44923537.0,
      "step": 20130
    },
    {
      "entropy": 5.434361934661865,
      "epoch": 2.2613578929634413,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044955002153549686,
      "loss": 5.1269,
      "mean_token_accuracy": 0.19831878393888475,
      "num_tokens": 44935780.0,
      "step": 20135
    },
    {
      "entropy": 5.375442409515381,
      "epoch": 2.261919469871399,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004495246846703828,
      "loss": 5.1175,
      "mean_token_accuracy": 0.19969422072172166,
      "num_tokens": 44946879.0,
      "step": 20140
    },
    {
      "entropy": 5.389518928527832,
      "epoch": 2.2624810467793566,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044949934224833505,
      "loss": 5.1085,
      "mean_token_accuracy": 0.20671920329332352,
      "num_tokens": 44958385.0,
      "step": 20145
    },
    {
      "entropy": 5.379944324493408,
      "epoch": 2.263042623687314,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004494739942701606,
      "loss": 5.1993,
      "mean_token_accuracy": 0.19508476853370665,
      "num_tokens": 44970597.0,
      "step": 20150
    },
    {
      "entropy": 5.386096668243408,
      "epoch": 2.2636042005952715,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044944864073666655,
      "loss": 5.1281,
      "mean_token_accuracy": 0.20448593348264693,
      "num_tokens": 44983123.0,
      "step": 20155
    },
    {
      "entropy": 5.358403873443604,
      "epoch": 2.264165777503229,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044942328164866,
      "loss": 5.057,
      "mean_token_accuracy": 0.1983027771115303,
      "num_tokens": 44993378.0,
      "step": 20160
    },
    {
      "entropy": 5.292265844345093,
      "epoch": 2.2647273544111868,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004493979170069486,
      "loss": 5.0345,
      "mean_token_accuracy": 0.20365310162305833,
      "num_tokens": 45004905.0,
      "step": 20165
    },
    {
      "entropy": 5.2693897724151615,
      "epoch": 2.265288931319144,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044937254681233984,
      "loss": 5.0825,
      "mean_token_accuracy": 0.20016768723726272,
      "num_tokens": 45016640.0,
      "step": 20170
    },
    {
      "entropy": 5.436551666259765,
      "epoch": 2.2658505082271017,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044934717106564164,
      "loss": 5.2111,
      "mean_token_accuracy": 0.1984803095459938,
      "num_tokens": 45028285.0,
      "step": 20175
    },
    {
      "entropy": 5.449406576156616,
      "epoch": 2.2664120851350593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044932178976766183,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2022205501794815,
      "num_tokens": 45038924.0,
      "step": 20180
    },
    {
      "entropy": 5.353724431991577,
      "epoch": 2.266973662043017,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044929640291920865,
      "loss": 5.056,
      "mean_token_accuracy": 0.20464457124471663,
      "num_tokens": 45050483.0,
      "step": 20185
    },
    {
      "entropy": 5.4004720687866214,
      "epoch": 2.267535238950974,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004492710105210904,
      "loss": 5.245,
      "mean_token_accuracy": 0.19622887969017028,
      "num_tokens": 45063242.0,
      "step": 20190
    },
    {
      "entropy": 5.343885803222657,
      "epoch": 2.268096815858932,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044924561257411554,
      "loss": 5.1141,
      "mean_token_accuracy": 0.19760920554399491,
      "num_tokens": 45074057.0,
      "step": 20195
    },
    {
      "entropy": 5.344647550582886,
      "epoch": 2.2686583927668895,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004492202090790928,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19460835456848144,
      "num_tokens": 45085488.0,
      "step": 20200
    },
    {
      "entropy": 5.482232999801636,
      "epoch": 2.2692199696748467,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004491948000368311,
      "loss": 5.2132,
      "mean_token_accuracy": 0.20241600573062896,
      "num_tokens": 45096303.0,
      "step": 20205
    },
    {
      "entropy": 5.376873874664307,
      "epoch": 2.2697815465828044,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044916938544813936,
      "loss": 4.9879,
      "mean_token_accuracy": 0.20667999237775803,
      "num_tokens": 45106411.0,
      "step": 20210
    },
    {
      "entropy": 5.339910507202148,
      "epoch": 2.270343123490762,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044914396531382686,
      "loss": 5.1678,
      "mean_token_accuracy": 0.20378995239734649,
      "num_tokens": 45117355.0,
      "step": 20215
    },
    {
      "entropy": 5.338074731826782,
      "epoch": 2.2709047003987197,
      "grad_norm": 1.203125,
      "learning_rate": 0.00044911853963470295,
      "loss": 5.0337,
      "mean_token_accuracy": 0.20728358030319213,
      "num_tokens": 45127612.0,
      "step": 20220
    },
    {
      "entropy": 5.4230427742004395,
      "epoch": 2.2714662773066774,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004490931084115772,
      "loss": 5.2073,
      "mean_token_accuracy": 0.20217876881361008,
      "num_tokens": 45138495.0,
      "step": 20225
    },
    {
      "entropy": 5.357241296768189,
      "epoch": 2.2720278542146346,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044906767164525936,
      "loss": 5.0394,
      "mean_token_accuracy": 0.2094038411974907,
      "num_tokens": 45149727.0,
      "step": 20230
    },
    {
      "entropy": 5.350690031051636,
      "epoch": 2.2725894311225923,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004490422293365594,
      "loss": 5.1197,
      "mean_token_accuracy": 0.20096703469753266,
      "num_tokens": 45160921.0,
      "step": 20235
    },
    {
      "entropy": 5.323499155044556,
      "epoch": 2.27315100803055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004490167814862872,
      "loss": 5.0442,
      "mean_token_accuracy": 0.20805609822273255,
      "num_tokens": 45172607.0,
      "step": 20240
    },
    {
      "entropy": 5.415520191192627,
      "epoch": 2.273712584938507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004489913280952533,
      "loss": 5.2208,
      "mean_token_accuracy": 0.20074901282787322,
      "num_tokens": 45186330.0,
      "step": 20245
    },
    {
      "entropy": 5.22049913406372,
      "epoch": 2.274274161846465,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004489658691642681,
      "loss": 4.9735,
      "mean_token_accuracy": 0.21477728039026261,
      "num_tokens": 45198239.0,
      "step": 20250
    },
    {
      "entropy": 5.241580486297607,
      "epoch": 2.2748357387544225,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044894040469414207,
      "loss": 4.9788,
      "mean_token_accuracy": 0.20981811881065368,
      "num_tokens": 45209295.0,
      "step": 20255
    },
    {
      "entropy": 5.399084043502808,
      "epoch": 2.27539731566238,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044891493468568607,
      "loss": 5.1023,
      "mean_token_accuracy": 0.20668914020061493,
      "num_tokens": 45220210.0,
      "step": 20260
    },
    {
      "entropy": 5.36901364326477,
      "epoch": 2.2759588925703373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044888945913971106,
      "loss": 5.1334,
      "mean_token_accuracy": 0.2017901986837387,
      "num_tokens": 45231809.0,
      "step": 20265
    },
    {
      "entropy": 5.363797187805176,
      "epoch": 2.276520469478295,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044886397805702835,
      "loss": 5.0948,
      "mean_token_accuracy": 0.21255529522895814,
      "num_tokens": 45242847.0,
      "step": 20270
    },
    {
      "entropy": 5.308997535705567,
      "epoch": 2.2770820463862527,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044883849143844914,
      "loss": 5.04,
      "mean_token_accuracy": 0.2152576506137848,
      "num_tokens": 45253357.0,
      "step": 20275
    },
    {
      "entropy": 5.272324180603027,
      "epoch": 2.2776436232942103,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004488129992847848,
      "loss": 4.9577,
      "mean_token_accuracy": 0.21194023191928862,
      "num_tokens": 45264607.0,
      "step": 20280
    },
    {
      "entropy": 5.447886610031128,
      "epoch": 2.2782052002021675,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004487875015968473,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20291971266269684,
      "num_tokens": 45275211.0,
      "step": 20285
    },
    {
      "entropy": 5.32122802734375,
      "epoch": 2.278766777110125,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004487619983754483,
      "loss": 5.0104,
      "mean_token_accuracy": 0.21178892701864244,
      "num_tokens": 45286665.0,
      "step": 20290
    },
    {
      "entropy": 5.324554443359375,
      "epoch": 2.279328354018083,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044873648962139983,
      "loss": 5.1315,
      "mean_token_accuracy": 0.2066076084971428,
      "num_tokens": 45298066.0,
      "step": 20295
    },
    {
      "entropy": 5.382396221160889,
      "epoch": 2.27988993092604,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004487109753355142,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20128075182437896,
      "num_tokens": 45308799.0,
      "step": 20300
    },
    {
      "entropy": 5.349091339111328,
      "epoch": 2.2804515078339977,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004486854555186037,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21205297410488128,
      "num_tokens": 45319932.0,
      "step": 20305
    },
    {
      "entropy": 5.3762444972991945,
      "epoch": 2.2810130847419554,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004486599301714809,
      "loss": 5.1163,
      "mean_token_accuracy": 0.20812519937753676,
      "num_tokens": 45330008.0,
      "step": 20310
    },
    {
      "entropy": 5.361810398101807,
      "epoch": 2.281574661649913,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004486343992949587,
      "loss": 5.1225,
      "mean_token_accuracy": 0.20168101638555527,
      "num_tokens": 45340877.0,
      "step": 20315
    },
    {
      "entropy": 5.341693639755249,
      "epoch": 2.2821362385578707,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044860886288984977,
      "loss": 5.091,
      "mean_token_accuracy": 0.21070897877216338,
      "num_tokens": 45351956.0,
      "step": 20320
    },
    {
      "entropy": 5.439002132415771,
      "epoch": 2.282697815465828,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485833209569674,
      "loss": 5.1671,
      "mean_token_accuracy": 0.20155526101589202,
      "num_tokens": 45363133.0,
      "step": 20325
    },
    {
      "entropy": 5.365840101242066,
      "epoch": 2.2832593923737856,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004485577734971247,
      "loss": 5.031,
      "mean_token_accuracy": 0.2058436632156372,
      "num_tokens": 45373917.0,
      "step": 20330
    },
    {
      "entropy": 5.391622018814087,
      "epoch": 2.2838209692817433,
      "grad_norm": 1.109375,
      "learning_rate": 0.00044853222051113525,
      "loss": 5.1477,
      "mean_token_accuracy": 0.20523791909217834,
      "num_tokens": 45385575.0,
      "step": 20335
    },
    {
      "entropy": 5.352527093887329,
      "epoch": 2.2843825461897005,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044850666199981244,
      "loss": 5.044,
      "mean_token_accuracy": 0.2044575184583664,
      "num_tokens": 45396615.0,
      "step": 20340
    },
    {
      "entropy": 5.337377595901489,
      "epoch": 2.284944123097658,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044848109796397035,
      "loss": 5.0634,
      "mean_token_accuracy": 0.20646312534809114,
      "num_tokens": 45408931.0,
      "step": 20345
    },
    {
      "entropy": 5.393699312210083,
      "epoch": 2.285505700005616,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044845552840442276,
      "loss": 5.1313,
      "mean_token_accuracy": 0.20015745013952255,
      "num_tokens": 45419993.0,
      "step": 20350
    },
    {
      "entropy": 5.39482045173645,
      "epoch": 2.2860672769135735,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044842995332198393,
      "loss": 5.1716,
      "mean_token_accuracy": 0.19468427151441575,
      "num_tokens": 45431889.0,
      "step": 20355
    },
    {
      "entropy": 5.298393058776855,
      "epoch": 2.2866288538215307,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044840437271746807,
      "loss": 5.1031,
      "mean_token_accuracy": 0.20591969937086105,
      "num_tokens": 45444233.0,
      "step": 20360
    },
    {
      "entropy": 5.3450030326843265,
      "epoch": 2.2871904307294884,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004483787865916898,
      "loss": 5.0613,
      "mean_token_accuracy": 0.204450361430645,
      "num_tokens": 45455743.0,
      "step": 20365
    },
    {
      "entropy": 5.307443475723266,
      "epoch": 2.287752007637446,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044835319494546366,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20327189117670058,
      "num_tokens": 45466272.0,
      "step": 20370
    },
    {
      "entropy": 5.390641450881958,
      "epoch": 2.2883135845454037,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044832759777960456,
      "loss": 5.0597,
      "mean_token_accuracy": 0.20406009256839752,
      "num_tokens": 45477828.0,
      "step": 20375
    },
    {
      "entropy": 5.346872234344483,
      "epoch": 2.288875161453361,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004483019950949276,
      "loss": 5.0767,
      "mean_token_accuracy": 0.19802024215459824,
      "num_tokens": 45489315.0,
      "step": 20380
    },
    {
      "entropy": 5.257917881011963,
      "epoch": 2.2894367383613186,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004482763868922478,
      "loss": 5.0341,
      "mean_token_accuracy": 0.2123546466231346,
      "num_tokens": 45499836.0,
      "step": 20385
    },
    {
      "entropy": 5.398345994949341,
      "epoch": 2.289998315269276,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004482507731723807,
      "loss": 5.1131,
      "mean_token_accuracy": 0.20347088426351548,
      "num_tokens": 45511106.0,
      "step": 20390
    },
    {
      "entropy": 5.366377496719361,
      "epoch": 2.2905598921772334,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004482251539361418,
      "loss": 5.0664,
      "mean_token_accuracy": 0.2021627590060234,
      "num_tokens": 45521314.0,
      "step": 20395
    },
    {
      "entropy": 5.408805418014526,
      "epoch": 2.291121469085191,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481995291843467,
      "loss": 5.1054,
      "mean_token_accuracy": 0.20929235517978667,
      "num_tokens": 45531675.0,
      "step": 20400
    },
    {
      "entropy": 5.403948450088501,
      "epoch": 2.2916830459931488,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004481738989178115,
      "loss": 5.1312,
      "mean_token_accuracy": 0.2043342873454094,
      "num_tokens": 45542064.0,
      "step": 20405
    },
    {
      "entropy": 5.359946775436401,
      "epoch": 2.2922446229011064,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004481482631373522,
      "loss": 5.1511,
      "mean_token_accuracy": 0.20445766001939775,
      "num_tokens": 45553764.0,
      "step": 20410
    },
    {
      "entropy": 5.469867324829101,
      "epoch": 2.292806199809064,
      "grad_norm": 1.0,
      "learning_rate": 0.00044812262184378503,
      "loss": 5.1338,
      "mean_token_accuracy": 0.2060307428240776,
      "num_tokens": 45565302.0,
      "step": 20415
    },
    {
      "entropy": 5.241159534454345,
      "epoch": 2.2933677767170213,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00044809697503792647,
      "loss": 4.9976,
      "mean_token_accuracy": 0.2138414904475212,
      "num_tokens": 45577165.0,
      "step": 20420
    },
    {
      "entropy": 5.329216718673706,
      "epoch": 2.293929353624979,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044807132272059306,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21068395674228668,
      "num_tokens": 45587771.0,
      "step": 20425
    },
    {
      "entropy": 5.359926366806031,
      "epoch": 2.2944909305329366,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004480456648926016,
      "loss": 5.0721,
      "mean_token_accuracy": 0.2165664792060852,
      "num_tokens": 45598815.0,
      "step": 20430
    },
    {
      "entropy": 5.3172706127166744,
      "epoch": 2.295052507440894,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004480200015547691,
      "loss": 5.0404,
      "mean_token_accuracy": 0.20626538395881652,
      "num_tokens": 45609007.0,
      "step": 20435
    },
    {
      "entropy": 5.3091613292694095,
      "epoch": 2.2956140843488515,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00044799433270791255,
      "loss": 5.1451,
      "mean_token_accuracy": 0.1992732360959053,
      "num_tokens": 45621323.0,
      "step": 20440
    },
    {
      "entropy": 5.302775812149048,
      "epoch": 2.296175661256809,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004479686583528494,
      "loss": 4.919,
      "mean_token_accuracy": 0.21014518290758133,
      "num_tokens": 45631895.0,
      "step": 20445
    },
    {
      "entropy": 5.397558259963989,
      "epoch": 2.296737238164767,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004479429784903971,
      "loss": 5.079,
      "mean_token_accuracy": 0.20835694819688796,
      "num_tokens": 45643204.0,
      "step": 20450
    },
    {
      "entropy": 5.351912784576416,
      "epoch": 2.297298815072724,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004479172931213733,
      "loss": 5.1216,
      "mean_token_accuracy": 0.20591072738170624,
      "num_tokens": 45654607.0,
      "step": 20455
    },
    {
      "entropy": 5.2486944675445555,
      "epoch": 2.2978603919806817,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004478916022465958,
      "loss": 5.0222,
      "mean_token_accuracy": 0.2072629675269127,
      "num_tokens": 45666746.0,
      "step": 20460
    },
    {
      "entropy": 5.288469028472901,
      "epoch": 2.2984219688886394,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004478659058668827,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2085390046238899,
      "num_tokens": 45676663.0,
      "step": 20465
    },
    {
      "entropy": 5.297271299362182,
      "epoch": 2.298983545796597,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044784020398305204,
      "loss": 5.0139,
      "mean_token_accuracy": 0.20700441151857377,
      "num_tokens": 45688746.0,
      "step": 20470
    },
    {
      "entropy": 5.43873553276062,
      "epoch": 2.2995451227045542,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004478144965959222,
      "loss": 5.2207,
      "mean_token_accuracy": 0.19751724898815154,
      "num_tokens": 45700199.0,
      "step": 20475
    },
    {
      "entropy": 5.3294299125671385,
      "epoch": 2.300106699612512,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044778878370631183,
      "loss": 5.0706,
      "mean_token_accuracy": 0.19786709547042847,
      "num_tokens": 45710771.0,
      "step": 20480
    },
    {
      "entropy": 5.385776519775391,
      "epoch": 2.3006682765204696,
      "grad_norm": 1.125,
      "learning_rate": 0.00044776306531503957,
      "loss": 5.2334,
      "mean_token_accuracy": 0.2028210461139679,
      "num_tokens": 45721957.0,
      "step": 20485
    },
    {
      "entropy": 5.3663506507873535,
      "epoch": 2.301229853428427,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044773734142292425,
      "loss": 5.0814,
      "mean_token_accuracy": 0.2008628711104393,
      "num_tokens": 45733904.0,
      "step": 20490
    },
    {
      "entropy": 5.4276511669158936,
      "epoch": 2.3017914303363844,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004477116120307851,
      "loss": 5.1055,
      "mean_token_accuracy": 0.20731922388076782,
      "num_tokens": 45744004.0,
      "step": 20495
    },
    {
      "entropy": 5.334802865982056,
      "epoch": 2.302353007244342,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004476858771394412,
      "loss": 5.0056,
      "mean_token_accuracy": 0.21643689423799514,
      "num_tokens": 45755130.0,
      "step": 20500
    },
    {
      "entropy": 5.297990846633911,
      "epoch": 2.3029145841522998,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004476601367497119,
      "loss": 5.1938,
      "mean_token_accuracy": 0.20154662430286407,
      "num_tokens": 45767353.0,
      "step": 20505
    },
    {
      "entropy": 5.414880561828613,
      "epoch": 2.3034761610602574,
      "grad_norm": 1.09375,
      "learning_rate": 0.000447634390862417,
      "loss": 5.2633,
      "mean_token_accuracy": 0.18926682323217392,
      "num_tokens": 45779266.0,
      "step": 20510
    },
    {
      "entropy": 5.455040025711059,
      "epoch": 2.3040377379682146,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044760863947837607,
      "loss": 5.2087,
      "mean_token_accuracy": 0.19775144755840302,
      "num_tokens": 45790601.0,
      "step": 20515
    },
    {
      "entropy": 5.3755511283874515,
      "epoch": 2.3045993148761723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044758288259840904,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20379408001899718,
      "num_tokens": 45801449.0,
      "step": 20520
    },
    {
      "entropy": 5.248305606842041,
      "epoch": 2.30516089178413,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004475571202233363,
      "loss": 5.0087,
      "mean_token_accuracy": 0.20786976516246797,
      "num_tokens": 45811737.0,
      "step": 20525
    },
    {
      "entropy": 5.30620698928833,
      "epoch": 2.305722468692087,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004475313523539778,
      "loss": 5.0525,
      "mean_token_accuracy": 0.21306243985891343,
      "num_tokens": 45822955.0,
      "step": 20530
    },
    {
      "entropy": 5.344698476791382,
      "epoch": 2.306284045600045,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044750557899115407,
      "loss": 5.1584,
      "mean_token_accuracy": 0.20897065103054047,
      "num_tokens": 45834697.0,
      "step": 20535
    },
    {
      "entropy": 5.3242103099823,
      "epoch": 2.3068456225080025,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474798001356859,
      "loss": 5.0818,
      "mean_token_accuracy": 0.2017083317041397,
      "num_tokens": 45846869.0,
      "step": 20540
    },
    {
      "entropy": 5.3313860416412355,
      "epoch": 2.30740719941596,
      "grad_norm": 1.09375,
      "learning_rate": 0.000447454015788394,
      "loss": 5.0561,
      "mean_token_accuracy": 0.20865903198719024,
      "num_tokens": 45858171.0,
      "step": 20545
    },
    {
      "entropy": 5.339020252227783,
      "epoch": 2.3079687763239174,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004474282259500992,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20026569962501525,
      "num_tokens": 45868715.0,
      "step": 20550
    },
    {
      "entropy": 5.328518295288086,
      "epoch": 2.308530353231875,
      "grad_norm": 1.0625,
      "learning_rate": 0.000447402430621623,
      "loss": 5.047,
      "mean_token_accuracy": 0.21342927515506743,
      "num_tokens": 45880569.0,
      "step": 20555
    },
    {
      "entropy": 5.35097484588623,
      "epoch": 2.3090919301398327,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004473766298037865,
      "loss": 5.0503,
      "mean_token_accuracy": 0.20786226093769072,
      "num_tokens": 45890452.0,
      "step": 20560
    },
    {
      "entropy": 5.453346920013428,
      "epoch": 2.3096535070477904,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044735082349741116,
      "loss": 5.1509,
      "mean_token_accuracy": 0.20034567266702652,
      "num_tokens": 45903125.0,
      "step": 20565
    },
    {
      "entropy": 5.327281141281128,
      "epoch": 2.3102150839557476,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044732501170331886,
      "loss": 5.0327,
      "mean_token_accuracy": 0.20685889571905136,
      "num_tokens": 45914961.0,
      "step": 20570
    },
    {
      "entropy": 5.394677209854126,
      "epoch": 2.3107766608637053,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004472991944223313,
      "loss": 5.1915,
      "mean_token_accuracy": 0.2007885307073593,
      "num_tokens": 45927198.0,
      "step": 20575
    },
    {
      "entropy": 5.379442930221558,
      "epoch": 2.311338237771663,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044727337165527056,
      "loss": 5.1514,
      "mean_token_accuracy": 0.20365616977214812,
      "num_tokens": 45937937.0,
      "step": 20580
    },
    {
      "entropy": 5.469306182861328,
      "epoch": 2.31189981467962,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044724754340295887,
      "loss": 5.1184,
      "mean_token_accuracy": 0.19479708522558212,
      "num_tokens": 45948983.0,
      "step": 20585
    },
    {
      "entropy": 5.330069351196289,
      "epoch": 2.312461391587578,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004472217096662186,
      "loss": 5.1593,
      "mean_token_accuracy": 0.19740949422121049,
      "num_tokens": 45960674.0,
      "step": 20590
    },
    {
      "entropy": 5.340665864944458,
      "epoch": 2.3130229684955355,
      "grad_norm": 1.125,
      "learning_rate": 0.0004471958704458723,
      "loss": 5.0642,
      "mean_token_accuracy": 0.2026139795780182,
      "num_tokens": 45971260.0,
      "step": 20595
    },
    {
      "entropy": 5.454844999313354,
      "epoch": 2.313584545403493,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044717002574274267,
      "loss": 5.1076,
      "mean_token_accuracy": 0.2028655782341957,
      "num_tokens": 45983942.0,
      "step": 20600
    },
    {
      "entropy": 5.366620349884033,
      "epoch": 2.314146122311451,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004471441755576527,
      "loss": 5.1374,
      "mean_token_accuracy": 0.19894261360168458,
      "num_tokens": 45995983.0,
      "step": 20605
    },
    {
      "entropy": 5.352132892608642,
      "epoch": 2.314707699219408,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044711831989142534,
      "loss": 5.051,
      "mean_token_accuracy": 0.20581570118665696,
      "num_tokens": 46007074.0,
      "step": 20610
    },
    {
      "entropy": 5.409793329238892,
      "epoch": 2.3152692761273657,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044709245874488405,
      "loss": 5.1218,
      "mean_token_accuracy": 0.20638673156499862,
      "num_tokens": 46018376.0,
      "step": 20615
    },
    {
      "entropy": 5.322412204742432,
      "epoch": 2.3158308530353233,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004470665921188519,
      "loss": 5.0611,
      "mean_token_accuracy": 0.2064543291926384,
      "num_tokens": 46029192.0,
      "step": 20620
    },
    {
      "entropy": 5.291636943817139,
      "epoch": 2.3163924299432805,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044704072001415285,
      "loss": 5.1161,
      "mean_token_accuracy": 0.20536373406648636,
      "num_tokens": 46040735.0,
      "step": 20625
    },
    {
      "entropy": 5.317258739471436,
      "epoch": 2.316954006851238,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044701484243161055,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21010254919528962,
      "num_tokens": 46050859.0,
      "step": 20630
    },
    {
      "entropy": 5.281309604644775,
      "epoch": 2.317515583759196,
      "grad_norm": 1.0625,
      "learning_rate": 0.000446988959372049,
      "loss": 5.064,
      "mean_token_accuracy": 0.20660900175571442,
      "num_tokens": 46060704.0,
      "step": 20635
    },
    {
      "entropy": 5.290012550354004,
      "epoch": 2.3180771606671535,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004469630708362922,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20604062527418138,
      "num_tokens": 46072317.0,
      "step": 20640
    },
    {
      "entropy": 5.441315746307373,
      "epoch": 2.3186387375751107,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004469371768251646,
      "loss": 5.1971,
      "mean_token_accuracy": 0.197893688082695,
      "num_tokens": 46083823.0,
      "step": 20645
    },
    {
      "entropy": 5.458332681655884,
      "epoch": 2.3192003144830684,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004469112773394906,
      "loss": 5.1344,
      "mean_token_accuracy": 0.20321834683418274,
      "num_tokens": 46093920.0,
      "step": 20650
    },
    {
      "entropy": 5.314505338668823,
      "epoch": 2.319761891391026,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004468853723800948,
      "loss": 5.0522,
      "mean_token_accuracy": 0.20753435641527176,
      "num_tokens": 46104618.0,
      "step": 20655
    },
    {
      "entropy": 5.310702323913574,
      "epoch": 2.3203234682989837,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004468594619478021,
      "loss": 4.9439,
      "mean_token_accuracy": 0.20728110522031784,
      "num_tokens": 46114748.0,
      "step": 20660
    },
    {
      "entropy": 5.315951251983643,
      "epoch": 2.320885045206941,
      "grad_norm": 1.3203125,
      "learning_rate": 0.0004468335460434375,
      "loss": 5.13,
      "mean_token_accuracy": 0.2100660338997841,
      "num_tokens": 46125125.0,
      "step": 20665
    },
    {
      "entropy": 5.197822427749633,
      "epoch": 2.3214466221148986,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004468076246678261,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20035115629434586,
      "num_tokens": 46135877.0,
      "step": 20670
    },
    {
      "entropy": 5.449482107162476,
      "epoch": 2.3220081990228563,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044678169782179334,
      "loss": 5.2214,
      "mean_token_accuracy": 0.19868141710758208,
      "num_tokens": 46146769.0,
      "step": 20675
    },
    {
      "entropy": 5.369632196426392,
      "epoch": 2.3225697759308135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044675576550616466,
      "loss": 5.0781,
      "mean_token_accuracy": 0.2070179268717766,
      "num_tokens": 46158408.0,
      "step": 20680
    },
    {
      "entropy": 5.40155987739563,
      "epoch": 2.323131352838771,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00044672982772176584,
      "loss": 5.1948,
      "mean_token_accuracy": 0.19960886240005493,
      "num_tokens": 46169961.0,
      "step": 20685
    },
    {
      "entropy": 5.39210786819458,
      "epoch": 2.323692929746729,
      "grad_norm": 1.125,
      "learning_rate": 0.0004467038844694227,
      "loss": 5.146,
      "mean_token_accuracy": 0.20854398906230925,
      "num_tokens": 46181750.0,
      "step": 20690
    },
    {
      "entropy": 5.325721645355225,
      "epoch": 2.3242545066546865,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004466779357499612,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2084916576743126,
      "num_tokens": 46192745.0,
      "step": 20695
    },
    {
      "entropy": 5.3451276302337645,
      "epoch": 2.324816083562644,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044665198156420774,
      "loss": 5.1159,
      "mean_token_accuracy": 0.205216620862484,
      "num_tokens": 46203267.0,
      "step": 20700
    },
    {
      "entropy": 5.349807405471802,
      "epoch": 2.3253776604706013,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044662602191298856,
      "loss": 5.0778,
      "mean_token_accuracy": 0.20753962248563768,
      "num_tokens": 46214966.0,
      "step": 20705
    },
    {
      "entropy": 5.289715766906738,
      "epoch": 2.325939237378559,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044660005679713027,
      "loss": 5.0207,
      "mean_token_accuracy": 0.20870077461004258,
      "num_tokens": 46225755.0,
      "step": 20710
    },
    {
      "entropy": 5.3680132865905765,
      "epoch": 2.3265008142865167,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044657408621745964,
      "loss": 5.1263,
      "mean_token_accuracy": 0.20525988042354584,
      "num_tokens": 46237091.0,
      "step": 20715
    },
    {
      "entropy": 5.442149209976196,
      "epoch": 2.327062391194474,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004465481101748036,
      "loss": 5.1223,
      "mean_token_accuracy": 0.20220893621444702,
      "num_tokens": 46249250.0,
      "step": 20720
    },
    {
      "entropy": 5.31956672668457,
      "epoch": 2.3276239681024316,
      "grad_norm": 1.15625,
      "learning_rate": 0.00044652212866998914,
      "loss": 5.0585,
      "mean_token_accuracy": 0.205973057448864,
      "num_tokens": 46259866.0,
      "step": 20725
    },
    {
      "entropy": 5.262479543685913,
      "epoch": 2.328185545010389,
      "grad_norm": 1.3125,
      "learning_rate": 0.00044649614170384354,
      "loss": 5.0145,
      "mean_token_accuracy": 0.21774810999631883,
      "num_tokens": 46270888.0,
      "step": 20730
    },
    {
      "entropy": 5.322278022766113,
      "epoch": 2.328747121918347,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044647014927719436,
      "loss": 5.0805,
      "mean_token_accuracy": 0.20544798970222472,
      "num_tokens": 46281109.0,
      "step": 20735
    },
    {
      "entropy": 5.329541921615601,
      "epoch": 2.329308698826304,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044644415139086913,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21546088755130768,
      "num_tokens": 46292171.0,
      "step": 20740
    },
    {
      "entropy": 5.331709527969361,
      "epoch": 2.3298702757342618,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004464181480456955,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21046677976846695,
      "num_tokens": 46302919.0,
      "step": 20745
    },
    {
      "entropy": 5.228147411346436,
      "epoch": 2.3304318526422194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004463921392425017,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21352233737707138,
      "num_tokens": 46313199.0,
      "step": 20750
    },
    {
      "entropy": 5.387055540084839,
      "epoch": 2.330993429550177,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044636612498211566,
      "loss": 5.102,
      "mean_token_accuracy": 0.20706415325403213,
      "num_tokens": 46324316.0,
      "step": 20755
    },
    {
      "entropy": 5.324500703811646,
      "epoch": 2.3315550064581343,
      "grad_norm": 1.1875,
      "learning_rate": 0.00044634010526536575,
      "loss": 5.0875,
      "mean_token_accuracy": 0.19993732422590255,
      "num_tokens": 46335767.0,
      "step": 20760
    },
    {
      "entropy": 5.290089464187622,
      "epoch": 2.332116583366092,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004463140800930804,
      "loss": 5.0173,
      "mean_token_accuracy": 0.20296950042247772,
      "num_tokens": 46347840.0,
      "step": 20765
    },
    {
      "entropy": 5.341397285461426,
      "epoch": 2.3326781602740496,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004462880494660883,
      "loss": 5.1658,
      "mean_token_accuracy": 0.20478006899356843,
      "num_tokens": 46359766.0,
      "step": 20770
    },
    {
      "entropy": 5.403687334060669,
      "epoch": 2.333239737182007,
      "grad_norm": 1.28125,
      "learning_rate": 0.00044626201338521834,
      "loss": 5.072,
      "mean_token_accuracy": 0.2126522809267044,
      "num_tokens": 46371426.0,
      "step": 20775
    },
    {
      "entropy": 5.345255136489868,
      "epoch": 2.3338013140899645,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004462359718512993,
      "loss": 5.1325,
      "mean_token_accuracy": 0.2010212928056717,
      "num_tokens": 46382257.0,
      "step": 20780
    },
    {
      "entropy": 5.387854814529419,
      "epoch": 2.334362890997922,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044620992486516073,
      "loss": 5.0963,
      "mean_token_accuracy": 0.201078100502491,
      "num_tokens": 46393689.0,
      "step": 20785
    },
    {
      "entropy": 5.355895090103149,
      "epoch": 2.33492446790588,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004461838724276316,
      "loss": 5.1381,
      "mean_token_accuracy": 0.2025910422205925,
      "num_tokens": 46404856.0,
      "step": 20790
    },
    {
      "entropy": 5.495230627059937,
      "epoch": 2.3354860448138375,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044615781453954164,
      "loss": 5.1356,
      "mean_token_accuracy": 0.2005625531077385,
      "num_tokens": 46415395.0,
      "step": 20795
    },
    {
      "entropy": 5.444233465194702,
      "epoch": 2.3360476217217947,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044613175120172035,
      "loss": 5.159,
      "mean_token_accuracy": 0.20015651881694793,
      "num_tokens": 46426773.0,
      "step": 20800
    },
    {
      "entropy": 5.280200576782226,
      "epoch": 2.3366091986297524,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004461056824149979,
      "loss": 4.9938,
      "mean_token_accuracy": 0.21180358231067659,
      "num_tokens": 46438206.0,
      "step": 20805
    },
    {
      "entropy": 5.3591855525970455,
      "epoch": 2.33717077553771,
      "grad_norm": 1.09375,
      "learning_rate": 0.000446079608180204,
      "loss": 5.1014,
      "mean_token_accuracy": 0.20576141625642777,
      "num_tokens": 46449922.0,
      "step": 20810
    },
    {
      "entropy": 5.35233588218689,
      "epoch": 2.3377323524456672,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004460535284981692,
      "loss": 5.0399,
      "mean_token_accuracy": 0.20728129148483276,
      "num_tokens": 46461715.0,
      "step": 20815
    },
    {
      "entropy": 5.305027294158935,
      "epoch": 2.338293929353625,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004460274433697235,
      "loss": 5.062,
      "mean_token_accuracy": 0.20290228426456453,
      "num_tokens": 46473136.0,
      "step": 20820
    },
    {
      "entropy": 5.2961078643798825,
      "epoch": 2.3388555062615826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044600135279569784,
      "loss": 4.9688,
      "mean_token_accuracy": 0.21417010873556136,
      "num_tokens": 46483644.0,
      "step": 20825
    },
    {
      "entropy": 5.393818044662476,
      "epoch": 2.3394170831695402,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044597525677692273,
      "loss": 5.1099,
      "mean_token_accuracy": 0.2055003121495247,
      "num_tokens": 46494896.0,
      "step": 20830
    },
    {
      "entropy": 5.369697618484497,
      "epoch": 2.3399786600774974,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004459491553142291,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20360004305839538,
      "num_tokens": 46505415.0,
      "step": 20835
    },
    {
      "entropy": 5.343874931335449,
      "epoch": 2.340540236985455,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044592304840844813,
      "loss": 5.0795,
      "mean_token_accuracy": 0.21130837500095367,
      "num_tokens": 46516280.0,
      "step": 20840
    },
    {
      "entropy": 5.247337532043457,
      "epoch": 2.3411018138934128,
      "grad_norm": 1.25,
      "learning_rate": 0.000445896936060411,
      "loss": 4.9661,
      "mean_token_accuracy": 0.212342968583107,
      "num_tokens": 46526456.0,
      "step": 20845
    },
    {
      "entropy": 5.396631860733033,
      "epoch": 2.3416633908013704,
      "grad_norm": 1.0,
      "learning_rate": 0.00044587081827094906,
      "loss": 5.1044,
      "mean_token_accuracy": 0.2046203389763832,
      "num_tokens": 46537547.0,
      "step": 20850
    },
    {
      "entropy": 5.297669506072998,
      "epoch": 2.3422249677093276,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044584469504089407,
      "loss": 5.0109,
      "mean_token_accuracy": 0.20883893966674805,
      "num_tokens": 46549424.0,
      "step": 20855
    },
    {
      "entropy": 5.323129177093506,
      "epoch": 2.3427865446172853,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004458185663710777,
      "loss": 5.0401,
      "mean_token_accuracy": 0.20588389039039612,
      "num_tokens": 46560234.0,
      "step": 20860
    },
    {
      "entropy": 5.3471111297607425,
      "epoch": 2.343348121525243,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004457924322623319,
      "loss": 5.0706,
      "mean_token_accuracy": 0.205361407995224,
      "num_tokens": 46570847.0,
      "step": 20865
    },
    {
      "entropy": 5.3007354736328125,
      "epoch": 2.3439096984332,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044576629271548884,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20300960689783096,
      "num_tokens": 46582379.0,
      "step": 20870
    },
    {
      "entropy": 5.332913112640381,
      "epoch": 2.344471275341158,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004457401477313808,
      "loss": 5.0643,
      "mean_token_accuracy": 0.1977205455303192,
      "num_tokens": 46594669.0,
      "step": 20875
    },
    {
      "entropy": 5.227885961532593,
      "epoch": 2.3450328522491155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004457139973108403,
      "loss": 4.9473,
      "mean_token_accuracy": 0.21855945587158204,
      "num_tokens": 46605834.0,
      "step": 20880
    },
    {
      "entropy": 5.2900628566741945,
      "epoch": 2.345594429157073,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004456878414546998,
      "loss": 5.0671,
      "mean_token_accuracy": 0.20854760855436325,
      "num_tokens": 46616684.0,
      "step": 20885
    },
    {
      "entropy": 5.372613620758057,
      "epoch": 2.346156006065031,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044566168016379234,
      "loss": 5.1221,
      "mean_token_accuracy": 0.2065764605998993,
      "num_tokens": 46628045.0,
      "step": 20890
    },
    {
      "entropy": 5.2980249404907225,
      "epoch": 2.346717582972988,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004456355134389507,
      "loss": 5.0685,
      "mean_token_accuracy": 0.20762595236301423,
      "num_tokens": 46639455.0,
      "step": 20895
    },
    {
      "entropy": 5.456133842468262,
      "epoch": 2.3472791598809457,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044560934128100814,
      "loss": 5.284,
      "mean_token_accuracy": 0.19438371509313584,
      "num_tokens": 46651025.0,
      "step": 20900
    },
    {
      "entropy": 5.400032234191895,
      "epoch": 2.3478407367889034,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044558316369079805,
      "loss": 5.0819,
      "mean_token_accuracy": 0.20437851101160048,
      "num_tokens": 46661859.0,
      "step": 20905
    },
    {
      "entropy": 5.361554145812988,
      "epoch": 2.3484023136968606,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044555698066915385,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20833214223384858,
      "num_tokens": 46673441.0,
      "step": 20910
    },
    {
      "entropy": 5.3677009582519535,
      "epoch": 2.3489638906048182,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004455307922169092,
      "loss": 5.135,
      "mean_token_accuracy": 0.19757814109325408,
      "num_tokens": 46684959.0,
      "step": 20915
    },
    {
      "entropy": 5.382802724838257,
      "epoch": 2.349525467512776,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004455045983348979,
      "loss": 5.1583,
      "mean_token_accuracy": 0.2111970826983452,
      "num_tokens": 46694940.0,
      "step": 20920
    },
    {
      "entropy": 5.344420480728149,
      "epoch": 2.3500870444207336,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004454783990239542,
      "loss": 5.0241,
      "mean_token_accuracy": 0.20831120908260345,
      "num_tokens": 46706184.0,
      "step": 20925
    },
    {
      "entropy": 5.3587822914123535,
      "epoch": 2.350648621328691,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044545219428491215,
      "loss": 5.0885,
      "mean_token_accuracy": 0.2099097564816475,
      "num_tokens": 46718267.0,
      "step": 20930
    },
    {
      "entropy": 5.357988739013672,
      "epoch": 2.3512101982366485,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044542598411860605,
      "loss": 5.1076,
      "mean_token_accuracy": 0.20367925763130187,
      "num_tokens": 46729293.0,
      "step": 20935
    },
    {
      "entropy": 5.324302291870117,
      "epoch": 2.351771775144606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004453997685258706,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21211587488651276,
      "num_tokens": 46740214.0,
      "step": 20940
    },
    {
      "entropy": 5.3821464538574215,
      "epoch": 2.3523333520525638,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044537354750754033,
      "loss": 5.1297,
      "mean_token_accuracy": 0.2028684973716736,
      "num_tokens": 46751269.0,
      "step": 20945
    },
    {
      "entropy": 5.2749505043029785,
      "epoch": 2.352894928960521,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004453473210644503,
      "loss": 4.9143,
      "mean_token_accuracy": 0.2171613872051239,
      "num_tokens": 46761851.0,
      "step": 20950
    },
    {
      "entropy": 5.264743423461914,
      "epoch": 2.3534565058684787,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004453210891974355,
      "loss": 5.0511,
      "mean_token_accuracy": 0.20909276604652405,
      "num_tokens": 46773441.0,
      "step": 20955
    },
    {
      "entropy": 5.231562900543213,
      "epoch": 2.3540180827764363,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044529485190733113,
      "loss": 5.0377,
      "mean_token_accuracy": 0.20571806877851487,
      "num_tokens": 46785016.0,
      "step": 20960
    },
    {
      "entropy": 5.360168600082398,
      "epoch": 2.3545796596843935,
      "grad_norm": 1.125,
      "learning_rate": 0.0004452686091949727,
      "loss": 5.0852,
      "mean_token_accuracy": 0.20057010650634766,
      "num_tokens": 46796059.0,
      "step": 20965
    },
    {
      "entropy": 5.282875728607178,
      "epoch": 2.355141236592351,
      "grad_norm": 1.09375,
      "learning_rate": 0.00044524236106119563,
      "loss": 5.0102,
      "mean_token_accuracy": 0.20759591907262803,
      "num_tokens": 46805725.0,
      "step": 20970
    },
    {
      "entropy": 5.342052459716797,
      "epoch": 2.355702813500309,
      "grad_norm": 1.125,
      "learning_rate": 0.00044521610750683586,
      "loss": 5.0277,
      "mean_token_accuracy": 0.21056323796510695,
      "num_tokens": 46816300.0,
      "step": 20975
    },
    {
      "entropy": 5.333779621124267,
      "epoch": 2.3562643904082665,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004451898485327291,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2018620103597641,
      "num_tokens": 46827498.0,
      "step": 20980
    },
    {
      "entropy": 5.328069257736206,
      "epoch": 2.356825967316224,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004451635841397116,
      "loss": 5.0641,
      "mean_token_accuracy": 0.19992548525333403,
      "num_tokens": 46838416.0,
      "step": 20985
    },
    {
      "entropy": 5.406435823440551,
      "epoch": 2.3573875442241814,
      "grad_norm": 1.25,
      "learning_rate": 0.0004451373143286195,
      "loss": 5.1626,
      "mean_token_accuracy": 0.20134071558713912,
      "num_tokens": 46849371.0,
      "step": 20990
    },
    {
      "entropy": 5.3002152919769285,
      "epoch": 2.357949121132139,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044511103910028945,
      "loss": 5.0934,
      "mean_token_accuracy": 0.20992300510406495,
      "num_tokens": 46860095.0,
      "step": 20995
    },
    {
      "entropy": 5.382495164871216,
      "epoch": 2.3585106980400967,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004450847584555579,
      "loss": 5.1664,
      "mean_token_accuracy": 0.19720554053783418,
      "num_tokens": 46871217.0,
      "step": 21000
    },
    {
      "epoch": 2.3585106980400967,
      "eval_entropy": 5.249399135615182,
      "eval_loss": 5.22935152053833,
      "eval_mean_token_accuracy": 0.20687413622084924,
      "eval_num_tokens": 46871217.0,
      "eval_runtime": 23.9477,
      "eval_samples_per_second": 1295.03,
      "eval_steps_per_second": 161.894,
      "step": 21000
    },
    {
      "entropy": 5.341647577285767,
      "epoch": 2.359072274948054,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004450584723952617,
      "loss": 5.0975,
      "mean_token_accuracy": 0.2063847690820694,
      "num_tokens": 46881224.0,
      "step": 21005
    },
    {
      "entropy": 5.353741645812988,
      "epoch": 2.3596338518560116,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004450321809202377,
      "loss": 5.0831,
      "mean_token_accuracy": 0.20559806525707244,
      "num_tokens": 46892673.0,
      "step": 21010
    },
    {
      "entropy": 5.483761119842529,
      "epoch": 2.3601954287639693,
      "grad_norm": 1.125,
      "learning_rate": 0.0004450058840313232,
      "loss": 5.1751,
      "mean_token_accuracy": 0.2048836573958397,
      "num_tokens": 46902753.0,
      "step": 21015
    },
    {
      "entropy": 5.257661056518555,
      "epoch": 2.360757005671927,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044497958172935536,
      "loss": 4.9944,
      "mean_token_accuracy": 0.2143484428524971,
      "num_tokens": 46914093.0,
      "step": 21020
    },
    {
      "entropy": 5.296432018280029,
      "epoch": 2.361318582579884,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004449532740151717,
      "loss": 5.0762,
      "mean_token_accuracy": 0.21013642549514772,
      "num_tokens": 46924946.0,
      "step": 21025
    },
    {
      "entropy": 5.36100959777832,
      "epoch": 2.361880159487842,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044492696088960985,
      "loss": 5.099,
      "mean_token_accuracy": 0.20259494632482528,
      "num_tokens": 46936756.0,
      "step": 21030
    },
    {
      "entropy": 5.450031900405884,
      "epoch": 2.3624417363957995,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004449006423535078,
      "loss": 5.1476,
      "mean_token_accuracy": 0.20377867966890334,
      "num_tokens": 46948552.0,
      "step": 21035
    },
    {
      "entropy": 5.433702087402343,
      "epoch": 2.363003313303757,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00044487431840770325,
      "loss": 5.1097,
      "mean_token_accuracy": 0.2046427994966507,
      "num_tokens": 46960926.0,
      "step": 21040
    },
    {
      "entropy": 5.392832708358765,
      "epoch": 2.3635648902117143,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004448479890530347,
      "loss": 5.0825,
      "mean_token_accuracy": 0.1995529741048813,
      "num_tokens": 46972657.0,
      "step": 21045
    },
    {
      "entropy": 5.3493246078491214,
      "epoch": 2.364126467119672,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00044482165429034014,
      "loss": 5.1138,
      "mean_token_accuracy": 0.20816900283098222,
      "num_tokens": 46984585.0,
      "step": 21050
    },
    {
      "entropy": 5.272209167480469,
      "epoch": 2.3646880440276297,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004447953141204583,
      "loss": 5.1273,
      "mean_token_accuracy": 0.20490482449531555,
      "num_tokens": 46996052.0,
      "step": 21055
    },
    {
      "entropy": 5.410894584655762,
      "epoch": 2.365249620935587,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044476896854422785,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21246029883623124,
      "num_tokens": 47006185.0,
      "step": 21060
    },
    {
      "entropy": 5.303319454193115,
      "epoch": 2.3658111978435445,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044474261756248745,
      "loss": 5.0944,
      "mean_token_accuracy": 0.20836876332759857,
      "num_tokens": 47016647.0,
      "step": 21065
    },
    {
      "entropy": 5.347690677642822,
      "epoch": 2.366372774751502,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004447162611760764,
      "loss": 5.1185,
      "mean_token_accuracy": 0.2054924950003624,
      "num_tokens": 47026847.0,
      "step": 21070
    },
    {
      "entropy": 5.368099546432495,
      "epoch": 2.36693435165946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004446898993858338,
      "loss": 5.1529,
      "mean_token_accuracy": 0.20075261145830153,
      "num_tokens": 47037948.0,
      "step": 21075
    },
    {
      "entropy": 5.375207567214966,
      "epoch": 2.3674959285674175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044466353219259884,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2074829012155533,
      "num_tokens": 47047751.0,
      "step": 21080
    },
    {
      "entropy": 5.3143524646759035,
      "epoch": 2.3680575054753747,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004446371595972114,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20662340819835662,
      "num_tokens": 47059589.0,
      "step": 21085
    },
    {
      "entropy": 5.294400787353515,
      "epoch": 2.3686190823833324,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004446107816005109,
      "loss": 5.088,
      "mean_token_accuracy": 0.2050166144967079,
      "num_tokens": 47070060.0,
      "step": 21090
    },
    {
      "entropy": 5.4353595733642575,
      "epoch": 2.36918065929129,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004445843982033373,
      "loss": 5.1229,
      "mean_token_accuracy": 0.20905786603689194,
      "num_tokens": 47081419.0,
      "step": 21095
    },
    {
      "entropy": 5.319814777374267,
      "epoch": 2.3697422361992473,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004445580094065307,
      "loss": 5.1015,
      "mean_token_accuracy": 0.19646659642457961,
      "num_tokens": 47093701.0,
      "step": 21100
    },
    {
      "entropy": 5.339945650100708,
      "epoch": 2.370303813107205,
      "grad_norm": 1.203125,
      "learning_rate": 0.0004445316152109313,
      "loss": 5.077,
      "mean_token_accuracy": 0.20609214752912522,
      "num_tokens": 47104930.0,
      "step": 21105
    },
    {
      "entropy": 5.346409940719605,
      "epoch": 2.3708653900151626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004445052156173796,
      "loss": 5.0407,
      "mean_token_accuracy": 0.20314418971538545,
      "num_tokens": 47115109.0,
      "step": 21110
    },
    {
      "entropy": 5.289762735366821,
      "epoch": 2.3714269669231203,
      "grad_norm": 1.21875,
      "learning_rate": 0.0004444788106267161,
      "loss": 5.0077,
      "mean_token_accuracy": 0.20990897417068483,
      "num_tokens": 47125386.0,
      "step": 21115
    },
    {
      "entropy": 5.387632083892822,
      "epoch": 2.3719885438310775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004444524002397815,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20521261543035507,
      "num_tokens": 47136066.0,
      "step": 21120
    },
    {
      "entropy": 5.3451180934906,
      "epoch": 2.372550120739035,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004444259844574167,
      "loss": 5.1478,
      "mean_token_accuracy": 0.2050887942314148,
      "num_tokens": 47147551.0,
      "step": 21125
    },
    {
      "entropy": 5.331465435028076,
      "epoch": 2.373111697646993,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044439956328046296,
      "loss": 5.0531,
      "mean_token_accuracy": 0.20601211935281755,
      "num_tokens": 47158798.0,
      "step": 21130
    },
    {
      "entropy": 5.45786485671997,
      "epoch": 2.3736732745549505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004443731367097613,
      "loss": 5.1698,
      "mean_token_accuracy": 0.2002708211541176,
      "num_tokens": 47169519.0,
      "step": 21135
    },
    {
      "entropy": 5.320242595672608,
      "epoch": 2.3742348514629077,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044434670474615334,
      "loss": 5.1034,
      "mean_token_accuracy": 0.20686566978693008,
      "num_tokens": 47180147.0,
      "step": 21140
    },
    {
      "entropy": 5.383879566192627,
      "epoch": 2.3747964283708654,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044432026739048064,
      "loss": 5.1478,
      "mean_token_accuracy": 0.20741900354623793,
      "num_tokens": 47191832.0,
      "step": 21145
    },
    {
      "entropy": 5.302240037918091,
      "epoch": 2.375358005278823,
      "grad_norm": 1.0625,
      "learning_rate": 0.000444293824643585,
      "loss": 5.0131,
      "mean_token_accuracy": 0.21372061967849731,
      "num_tokens": 47201979.0,
      "step": 21150
    },
    {
      "entropy": 5.3776772022247314,
      "epoch": 2.3759195821867802,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004442673765063082,
      "loss": 5.159,
      "mean_token_accuracy": 0.2037729248404503,
      "num_tokens": 47212271.0,
      "step": 21155
    },
    {
      "entropy": 5.327552843093872,
      "epoch": 2.376481159094738,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004442409229794925,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20425171107053758,
      "num_tokens": 47223412.0,
      "step": 21160
    },
    {
      "entropy": 5.4290323734283445,
      "epoch": 2.3770427360026956,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004442144640639803,
      "loss": 5.2056,
      "mean_token_accuracy": 0.19357030242681503,
      "num_tokens": 47235919.0,
      "step": 21165
    },
    {
      "entropy": 5.371686887741089,
      "epoch": 2.377604312910653,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004441879997606138,
      "loss": 5.0746,
      "mean_token_accuracy": 0.19923140853643417,
      "num_tokens": 47246568.0,
      "step": 21170
    },
    {
      "entropy": 5.426797580718994,
      "epoch": 2.378165889818611,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004441615300702358,
      "loss": 5.1609,
      "mean_token_accuracy": 0.2015797659754753,
      "num_tokens": 47259474.0,
      "step": 21175
    },
    {
      "entropy": 5.378620767593384,
      "epoch": 2.378727466726568,
      "grad_norm": 1.0,
      "learning_rate": 0.0004441350549936891,
      "loss": 5.0658,
      "mean_token_accuracy": 0.2095206707715988,
      "num_tokens": 47270906.0,
      "step": 21180
    },
    {
      "entropy": 5.319374465942383,
      "epoch": 2.3792890436345258,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004441085745318167,
      "loss": 5.1567,
      "mean_token_accuracy": 0.1941419407725334,
      "num_tokens": 47283341.0,
      "step": 21185
    },
    {
      "entropy": 5.370115613937378,
      "epoch": 2.3798506205424834,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044408208868546164,
      "loss": 5.1033,
      "mean_token_accuracy": 0.20189952552318574,
      "num_tokens": 47294590.0,
      "step": 21190
    },
    {
      "entropy": 5.259399795532227,
      "epoch": 2.3804121974504406,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004440555974554673,
      "loss": 4.9233,
      "mean_token_accuracy": 0.20971108078956605,
      "num_tokens": 47305111.0,
      "step": 21195
    },
    {
      "entropy": 5.30864782333374,
      "epoch": 2.3809737743583983,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00044402910084267717,
      "loss": 5.0873,
      "mean_token_accuracy": 0.20617498755455016,
      "num_tokens": 47316145.0,
      "step": 21200
    },
    {
      "entropy": 5.3256970882415775,
      "epoch": 2.381535351266356,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004440025988479349,
      "loss": 5.0199,
      "mean_token_accuracy": 0.2093463808298111,
      "num_tokens": 47327085.0,
      "step": 21205
    },
    {
      "entropy": 5.42002387046814,
      "epoch": 2.3820969281743136,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044397609147208436,
      "loss": 5.2145,
      "mean_token_accuracy": 0.2037497028708458,
      "num_tokens": 47338308.0,
      "step": 21210
    },
    {
      "entropy": 5.419722270965576,
      "epoch": 2.382658505082271,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004439495787159695,
      "loss": 5.1686,
      "mean_token_accuracy": 0.2023970901966095,
      "num_tokens": 47348717.0,
      "step": 21215
    },
    {
      "entropy": 5.389338254928589,
      "epoch": 2.3832200819902285,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044392306058043453,
      "loss": 5.025,
      "mean_token_accuracy": 0.21648618280887605,
      "num_tokens": 47359804.0,
      "step": 21220
    },
    {
      "entropy": 5.313509607315064,
      "epoch": 2.383781658898186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004438965370663238,
      "loss": 5.0781,
      "mean_token_accuracy": 0.1954860046505928,
      "num_tokens": 47370589.0,
      "step": 21225
    },
    {
      "entropy": 5.3334324836730955,
      "epoch": 2.384343235806144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004438700081744819,
      "loss": 5.0282,
      "mean_token_accuracy": 0.20652858018875123,
      "num_tokens": 47381586.0,
      "step": 21230
    },
    {
      "entropy": 5.298813772201538,
      "epoch": 2.384904812714101,
      "grad_norm": 0.921875,
      "learning_rate": 0.00044384347390575337,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21492155343294145,
      "num_tokens": 47394064.0,
      "step": 21235
    },
    {
      "entropy": 5.279388666152954,
      "epoch": 2.3854663896220587,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004438169342609831,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20651203840970994,
      "num_tokens": 47405371.0,
      "step": 21240
    },
    {
      "entropy": 5.2457249641418455,
      "epoch": 2.3860279665300164,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044379038924101627,
      "loss": 5.0718,
      "mean_token_accuracy": 0.20939249396324158,
      "num_tokens": 47416446.0,
      "step": 21245
    },
    {
      "entropy": 5.297568845748901,
      "epoch": 2.3865895434379736,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004437638388466979,
      "loss": 4.9317,
      "mean_token_accuracy": 0.21583558470010758,
      "num_tokens": 47426588.0,
      "step": 21250
    },
    {
      "entropy": 5.355429553985596,
      "epoch": 2.3871511203459312,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044373728307887353,
      "loss": 5.1919,
      "mean_token_accuracy": 0.1953586995601654,
      "num_tokens": 47438015.0,
      "step": 21255
    },
    {
      "entropy": 5.470928478240967,
      "epoch": 2.387712697253889,
      "grad_norm": 1.0,
      "learning_rate": 0.00044371072193838857,
      "loss": 5.2405,
      "mean_token_accuracy": 0.19062355458736419,
      "num_tokens": 47448953.0,
      "step": 21260
    },
    {
      "entropy": 5.383955001831055,
      "epoch": 2.3882742741618466,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044368415542608874,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20967313796281814,
      "num_tokens": 47459669.0,
      "step": 21265
    },
    {
      "entropy": 5.453853607177734,
      "epoch": 2.3888358510698042,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044365758354282005,
      "loss": 5.1347,
      "mean_token_accuracy": 0.20502011030912398,
      "num_tokens": 47471018.0,
      "step": 21270
    },
    {
      "entropy": 5.4318805694580075,
      "epoch": 2.3893974279777614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004436310062894285,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20526975244283677,
      "num_tokens": 47481613.0,
      "step": 21275
    },
    {
      "entropy": 5.291565847396851,
      "epoch": 2.389959004885719,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004436044236667603,
      "loss": 4.9675,
      "mean_token_accuracy": 0.2067270278930664,
      "num_tokens": 47491245.0,
      "step": 21280
    },
    {
      "entropy": 5.269408512115478,
      "epoch": 2.3905205817936768,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004435778356756618,
      "loss": 5.0356,
      "mean_token_accuracy": 0.2126204028725624,
      "num_tokens": 47502308.0,
      "step": 21285
    },
    {
      "entropy": 5.272086763381958,
      "epoch": 2.391082158701634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044355124231697967,
      "loss": 5.0343,
      "mean_token_accuracy": 0.2127940446138382,
      "num_tokens": 47512799.0,
      "step": 21290
    },
    {
      "entropy": 5.422924423217774,
      "epoch": 2.3916437356095916,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004435246435915607,
      "loss": 5.1265,
      "mean_token_accuracy": 0.2077869400382042,
      "num_tokens": 47524740.0,
      "step": 21295
    },
    {
      "entropy": 5.430826139450073,
      "epoch": 2.3922053125175493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004434980395002516,
      "loss": 5.1587,
      "mean_token_accuracy": 0.20434848964214325,
      "num_tokens": 47535785.0,
      "step": 21300
    },
    {
      "entropy": 5.286251258850098,
      "epoch": 2.392766889425507,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044347143004389955,
      "loss": 4.9454,
      "mean_token_accuracy": 0.22040904611349105,
      "num_tokens": 47545698.0,
      "step": 21305
    },
    {
      "entropy": 5.268867540359497,
      "epoch": 2.393328466333464,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004434448152233519,
      "loss": 5.1219,
      "mean_token_accuracy": 0.2128116801381111,
      "num_tokens": 47559786.0,
      "step": 21310
    },
    {
      "entropy": 5.355034732818604,
      "epoch": 2.393890043241422,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044341819503945594,
      "loss": 5.0632,
      "mean_token_accuracy": 0.20909431129693984,
      "num_tokens": 47570267.0,
      "step": 21315
    },
    {
      "entropy": 5.4178314208984375,
      "epoch": 2.3944516201493795,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004433915694930594,
      "loss": 5.1876,
      "mean_token_accuracy": 0.19788068532943726,
      "num_tokens": 47580462.0,
      "step": 21320
    },
    {
      "entropy": 5.319781064987183,
      "epoch": 2.395013197057337,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004433649385850099,
      "loss": 5.022,
      "mean_token_accuracy": 0.20876433849334716,
      "num_tokens": 47590651.0,
      "step": 21325
    },
    {
      "entropy": 5.4964902877807615,
      "epoch": 2.3955747739652944,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004433383023161555,
      "loss": 5.2295,
      "mean_token_accuracy": 0.20375731587409973,
      "num_tokens": 47602263.0,
      "step": 21330
    },
    {
      "entropy": 5.38461332321167,
      "epoch": 2.396136350873252,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004433116606873443,
      "loss": 5.0448,
      "mean_token_accuracy": 0.20963948369026184,
      "num_tokens": 47613570.0,
      "step": 21335
    },
    {
      "entropy": 5.373274469375611,
      "epoch": 2.3966979277812097,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044328501369942454,
      "loss": 5.1389,
      "mean_token_accuracy": 0.21095341295003892,
      "num_tokens": 47624915.0,
      "step": 21340
    },
    {
      "entropy": 5.378404521942139,
      "epoch": 2.3972595046891674,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004432583613532446,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20307647436857224,
      "num_tokens": 47635974.0,
      "step": 21345
    },
    {
      "entropy": 5.308145666122437,
      "epoch": 2.3978210815971246,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004432317036496533,
      "loss": 5.0506,
      "mean_token_accuracy": 0.21103573888540267,
      "num_tokens": 47647312.0,
      "step": 21350
    },
    {
      "entropy": 5.252079200744629,
      "epoch": 2.3983826585050823,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00044320504058949916,
      "loss": 4.9454,
      "mean_token_accuracy": 0.21121157258749007,
      "num_tokens": 47659224.0,
      "step": 21355
    },
    {
      "entropy": 5.324654579162598,
      "epoch": 2.39894423541304,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004431783721736314,
      "loss": 5.1043,
      "mean_token_accuracy": 0.20820595026016236,
      "num_tokens": 47670720.0,
      "step": 21360
    },
    {
      "entropy": 5.367992401123047,
      "epoch": 2.3995058123209976,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044315169840289896,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20719297975301743,
      "num_tokens": 47680646.0,
      "step": 21365
    },
    {
      "entropy": 5.420795822143555,
      "epoch": 2.400067389228955,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004431250192781513,
      "loss": 5.075,
      "mean_token_accuracy": 0.2064395233988762,
      "num_tokens": 47692155.0,
      "step": 21370
    },
    {
      "entropy": 5.460869264602661,
      "epoch": 2.4006289661369125,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004430983348002378,
      "loss": 5.2218,
      "mean_token_accuracy": 0.19611628502607345,
      "num_tokens": 47703662.0,
      "step": 21375
    },
    {
      "entropy": 5.339164876937867,
      "epoch": 2.40119054304487,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044307164497000813,
      "loss": 5.0228,
      "mean_token_accuracy": 0.21200441122055053,
      "num_tokens": 47713617.0,
      "step": 21380
    },
    {
      "entropy": 5.314976453781128,
      "epoch": 2.4017521199528273,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044304494978831204,
      "loss": 5.0364,
      "mean_token_accuracy": 0.21192807853221893,
      "num_tokens": 47725051.0,
      "step": 21385
    },
    {
      "entropy": 5.2491583824157715,
      "epoch": 2.402313696860785,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004430182492559997,
      "loss": 4.9823,
      "mean_token_accuracy": 0.2157183676958084,
      "num_tokens": 47735023.0,
      "step": 21390
    },
    {
      "entropy": 5.3216947555542,
      "epoch": 2.4028752737687427,
      "grad_norm": 1.015625,
      "learning_rate": 0.000442991543373921,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2116788446903229,
      "num_tokens": 47746252.0,
      "step": 21395
    },
    {
      "entropy": 5.3307723045349125,
      "epoch": 2.4034368506767003,
      "grad_norm": 1.0,
      "learning_rate": 0.0004429648321429264,
      "loss": 5.148,
      "mean_token_accuracy": 0.20355014652013778,
      "num_tokens": 47756980.0,
      "step": 21400
    },
    {
      "entropy": 5.3625846862792965,
      "epoch": 2.4039984275846575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044293811556386656,
      "loss": 5.0946,
      "mean_token_accuracy": 0.20974625796079635,
      "num_tokens": 47767633.0,
      "step": 21405
    },
    {
      "entropy": 5.234905815124511,
      "epoch": 2.404560004492615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004429113936375918,
      "loss": 5.0235,
      "mean_token_accuracy": 0.20763096660375596,
      "num_tokens": 47779346.0,
      "step": 21410
    },
    {
      "entropy": 5.463612461090088,
      "epoch": 2.405121581400573,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044288466636495323,
      "loss": 5.1621,
      "mean_token_accuracy": 0.2036971554160118,
      "num_tokens": 47791039.0,
      "step": 21415
    },
    {
      "entropy": 5.3825828552246096,
      "epoch": 2.4056831583085305,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00044285793374680184,
      "loss": 5.1711,
      "mean_token_accuracy": 0.20380015969276427,
      "num_tokens": 47802096.0,
      "step": 21420
    },
    {
      "entropy": 5.3517224311828615,
      "epoch": 2.4062447352164877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044283119578398866,
      "loss": 5.0943,
      "mean_token_accuracy": 0.19694539159536362,
      "num_tokens": 47814087.0,
      "step": 21425
    },
    {
      "entropy": 5.311621570587159,
      "epoch": 2.4068063121244454,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004428044524773651,
      "loss": 5.0407,
      "mean_token_accuracy": 0.2067922458052635,
      "num_tokens": 47824656.0,
      "step": 21430
    },
    {
      "entropy": 5.370716524124146,
      "epoch": 2.407367889032403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004427777038277827,
      "loss": 5.0549,
      "mean_token_accuracy": 0.21415073275566102,
      "num_tokens": 47836045.0,
      "step": 21435
    },
    {
      "entropy": 5.37744026184082,
      "epoch": 2.4079294659403607,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044275094983609313,
      "loss": 5.0818,
      "mean_token_accuracy": 0.20760655850172044,
      "num_tokens": 47846494.0,
      "step": 21440
    },
    {
      "entropy": 5.3970489501953125,
      "epoch": 2.408491042848318,
      "grad_norm": 1.125,
      "learning_rate": 0.00044272419050314833,
      "loss": 5.1567,
      "mean_token_accuracy": 0.20415105074644088,
      "num_tokens": 47858689.0,
      "step": 21445
    },
    {
      "entropy": 5.262633895874023,
      "epoch": 2.4090526197562756,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044269742582980017,
      "loss": 5.0246,
      "mean_token_accuracy": 0.20916135609149933,
      "num_tokens": 47870534.0,
      "step": 21450
    },
    {
      "entropy": 5.3451025009155275,
      "epoch": 2.4096141966642333,
      "grad_norm": 1.125,
      "learning_rate": 0.000442670655816901,
      "loss": 5.0504,
      "mean_token_accuracy": 0.20984792411327363,
      "num_tokens": 47880354.0,
      "step": 21455
    },
    {
      "entropy": 5.412451076507568,
      "epoch": 2.410175773572191,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00044264388046530304,
      "loss": 5.1087,
      "mean_token_accuracy": 0.2025397926568985,
      "num_tokens": 47892528.0,
      "step": 21460
    },
    {
      "entropy": 5.4135151386260985,
      "epoch": 2.410737350480148,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044261709977585893,
      "loss": 5.137,
      "mean_token_accuracy": 0.21014703214168548,
      "num_tokens": 47904653.0,
      "step": 21465
    },
    {
      "entropy": 5.3896568775177,
      "epoch": 2.411298927388106,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004425903137494214,
      "loss": 5.1905,
      "mean_token_accuracy": 0.20199224650859832,
      "num_tokens": 47916452.0,
      "step": 21470
    },
    {
      "entropy": 5.387000322341919,
      "epoch": 2.4118605042960635,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004425635223868433,
      "loss": 5.0576,
      "mean_token_accuracy": 0.20262714028358458,
      "num_tokens": 47927746.0,
      "step": 21475
    },
    {
      "entropy": 5.331913232803345,
      "epoch": 2.4124220812040207,
      "grad_norm": 0.953125,
      "learning_rate": 0.00044253672568897754,
      "loss": 5.1075,
      "mean_token_accuracy": 0.20034135580062867,
      "num_tokens": 47941057.0,
      "step": 21480
    },
    {
      "entropy": 5.368620824813843,
      "epoch": 2.4129836581119783,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004425099236566775,
      "loss": 5.1578,
      "mean_token_accuracy": 0.2046702727675438,
      "num_tokens": 47951912.0,
      "step": 21485
    },
    {
      "entropy": 5.368574905395508,
      "epoch": 2.413545235019936,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044248311629079657,
      "loss": 5.0729,
      "mean_token_accuracy": 0.20773589462041855,
      "num_tokens": 47963483.0,
      "step": 21490
    },
    {
      "entropy": 5.4525984764099125,
      "epoch": 2.4141068119278937,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004424563035921882,
      "loss": 5.1631,
      "mean_token_accuracy": 0.20306519120931626,
      "num_tokens": 47975047.0,
      "step": 21495
    },
    {
      "entropy": 5.3824256420135494,
      "epoch": 2.414668388835851,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00044242948556170634,
      "loss": 5.1423,
      "mean_token_accuracy": 0.1951976850628853,
      "num_tokens": 47986432.0,
      "step": 21500
    },
    {
      "entropy": 5.3406003475189205,
      "epoch": 2.4152299657438085,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044240266220020455,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2087495967745781,
      "num_tokens": 47997283.0,
      "step": 21505
    },
    {
      "entropy": 5.303210973739624,
      "epoch": 2.415791542651766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004423758335085371,
      "loss": 5.0567,
      "mean_token_accuracy": 0.2031462609767914,
      "num_tokens": 48008353.0,
      "step": 21510
    },
    {
      "entropy": 5.379553174972534,
      "epoch": 2.416353119559724,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00044234899948755827,
      "loss": 5.1178,
      "mean_token_accuracy": 0.2045558586716652,
      "num_tokens": 48020065.0,
      "step": 21515
    },
    {
      "entropy": 5.394641256332397,
      "epoch": 2.416914696467681,
      "grad_norm": 1.125,
      "learning_rate": 0.00044232216013812226,
      "loss": 5.1239,
      "mean_token_accuracy": 0.20197992473840715,
      "num_tokens": 48030062.0,
      "step": 21520
    },
    {
      "entropy": 5.284262943267822,
      "epoch": 2.4174762733756388,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004422953154610838,
      "loss": 5.0591,
      "mean_token_accuracy": 0.20199424773454666,
      "num_tokens": 48039639.0,
      "step": 21525
    },
    {
      "entropy": 5.302639150619507,
      "epoch": 2.4180378502835964,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004422684654572977,
      "loss": 5.1198,
      "mean_token_accuracy": 0.19989813417196273,
      "num_tokens": 48050428.0,
      "step": 21530
    },
    {
      "entropy": 5.410748481750488,
      "epoch": 2.418599427191554,
      "grad_norm": 1.046875,
      "learning_rate": 0.00044224161012761874,
      "loss": 5.0523,
      "mean_token_accuracy": 0.20056888908147813,
      "num_tokens": 48061214.0,
      "step": 21535
    },
    {
      "entropy": 5.28414511680603,
      "epoch": 2.4191610040995113,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044221474947290205,
      "loss": 4.9666,
      "mean_token_accuracy": 0.20817440003156662,
      "num_tokens": 48071983.0,
      "step": 21540
    },
    {
      "entropy": 5.294326639175415,
      "epoch": 2.419722581007469,
      "grad_norm": 1.125,
      "learning_rate": 0.0004421878834940029,
      "loss": 4.989,
      "mean_token_accuracy": 0.21179089695215225,
      "num_tokens": 48083411.0,
      "step": 21545
    },
    {
      "entropy": 5.40851149559021,
      "epoch": 2.4202841579154266,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044216101219177667,
      "loss": 5.0913,
      "mean_token_accuracy": 0.198056860268116,
      "num_tokens": 48094718.0,
      "step": 21550
    },
    {
      "entropy": 5.367299032211304,
      "epoch": 2.4208457348233843,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00044213413556707905,
      "loss": 5.0943,
      "mean_token_accuracy": 0.2020697236061096,
      "num_tokens": 48106663.0,
      "step": 21555
    },
    {
      "entropy": 5.372581863403321,
      "epoch": 2.4214073117313415,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004421072536207656,
      "loss": 5.0988,
      "mean_token_accuracy": 0.20899490267038345,
      "num_tokens": 48117232.0,
      "step": 21560
    },
    {
      "entropy": 5.386395597457886,
      "epoch": 2.421968888639299,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044208036635369247,
      "loss": 5.0981,
      "mean_token_accuracy": 0.20874228179454804,
      "num_tokens": 48128817.0,
      "step": 21565
    },
    {
      "entropy": 5.361799240112305,
      "epoch": 2.422530465547257,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004420534737667157,
      "loss": 5.0583,
      "mean_token_accuracy": 0.21485474556684495,
      "num_tokens": 48140893.0,
      "step": 21570
    },
    {
      "entropy": 5.273307228088379,
      "epoch": 2.423092042455214,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044202657586069153,
      "loss": 4.9592,
      "mean_token_accuracy": 0.20872742384672166,
      "num_tokens": 48152661.0,
      "step": 21575
    },
    {
      "entropy": 5.322945261001587,
      "epoch": 2.4236536193631717,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044199967263647635,
      "loss": 5.083,
      "mean_token_accuracy": 0.2078230768442154,
      "num_tokens": 48163342.0,
      "step": 21580
    },
    {
      "entropy": 5.361215877532959,
      "epoch": 2.4242151962711294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004419727640949268,
      "loss": 5.1025,
      "mean_token_accuracy": 0.20635554194450378,
      "num_tokens": 48175321.0,
      "step": 21585
    },
    {
      "entropy": 5.399764728546143,
      "epoch": 2.424776773179087,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004419458502368998,
      "loss": 5.1265,
      "mean_token_accuracy": 0.19914610534906388,
      "num_tokens": 48186847.0,
      "step": 21590
    },
    {
      "entropy": 5.387173461914062,
      "epoch": 2.4253383500870442,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00044191893106325207,
      "loss": 5.2069,
      "mean_token_accuracy": 0.20496724545955658,
      "num_tokens": 48198840.0,
      "step": 21595
    },
    {
      "entropy": 5.4551849365234375,
      "epoch": 2.425899926995002,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004418920065748409,
      "loss": 5.1699,
      "mean_token_accuracy": 0.19796381741762162,
      "num_tokens": 48209948.0,
      "step": 21600
    },
    {
      "entropy": 5.355071926116944,
      "epoch": 2.4264615039029596,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044186507677252345,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20967598557472228,
      "num_tokens": 48220125.0,
      "step": 21605
    },
    {
      "entropy": 5.31331844329834,
      "epoch": 2.427023080810917,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004418381416571573,
      "loss": 5.0607,
      "mean_token_accuracy": 0.21032873541116714,
      "num_tokens": 48230736.0,
      "step": 21610
    },
    {
      "entropy": 5.267429399490356,
      "epoch": 2.4275846577188744,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004418112012296,
      "loss": 4.9857,
      "mean_token_accuracy": 0.21232457607984542,
      "num_tokens": 48241387.0,
      "step": 21615
    },
    {
      "entropy": 5.3571868419647215,
      "epoch": 2.428146234626832,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044178425549070934,
      "loss": 5.0749,
      "mean_token_accuracy": 0.20747715681791307,
      "num_tokens": 48252279.0,
      "step": 21620
    },
    {
      "entropy": 5.391210126876831,
      "epoch": 2.4287078115347898,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00044175730444134326,
      "loss": 5.1025,
      "mean_token_accuracy": 0.2100407987833023,
      "num_tokens": 48264235.0,
      "step": 21625
    },
    {
      "entropy": 5.298579549789428,
      "epoch": 2.4292693884427474,
      "grad_norm": 1.015625,
      "learning_rate": 0.00044173034808236,
      "loss": 4.9645,
      "mean_token_accuracy": 0.21341065764427186,
      "num_tokens": 48276194.0,
      "step": 21630
    },
    {
      "entropy": 5.3699544906616214,
      "epoch": 2.4298309653507046,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004417033864146177,
      "loss": 5.1268,
      "mean_token_accuracy": 0.2062091514468193,
      "num_tokens": 48287129.0,
      "step": 21635
    },
    {
      "entropy": 5.303134155273438,
      "epoch": 2.4303925422586623,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004416764194389751,
      "loss": 5.0722,
      "mean_token_accuracy": 0.2022764876484871,
      "num_tokens": 48298850.0,
      "step": 21640
    },
    {
      "entropy": 5.399888324737549,
      "epoch": 2.43095411916662,
      "grad_norm": 1.25,
      "learning_rate": 0.0004416494471562904,
      "loss": 5.1048,
      "mean_token_accuracy": 0.20562578737735748,
      "num_tokens": 48309968.0,
      "step": 21645
    },
    {
      "entropy": 5.416429615020752,
      "epoch": 2.4315156960745776,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004416224695674229,
      "loss": 5.0688,
      "mean_token_accuracy": 0.20902974009513856,
      "num_tokens": 48320419.0,
      "step": 21650
    },
    {
      "entropy": 5.378803968429565,
      "epoch": 2.432077272982535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044159548667323126,
      "loss": 5.1201,
      "mean_token_accuracy": 0.20475222468376159,
      "num_tokens": 48332611.0,
      "step": 21655
    },
    {
      "entropy": 5.2934410572052,
      "epoch": 2.4326388498904925,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004415684984745747,
      "loss": 5.103,
      "mean_token_accuracy": 0.2100343033671379,
      "num_tokens": 48344551.0,
      "step": 21660
    },
    {
      "entropy": 5.45091700553894,
      "epoch": 2.43320042679845,
      "grad_norm": 1.0,
      "learning_rate": 0.0004415415049723126,
      "loss": 5.1017,
      "mean_token_accuracy": 0.20578287690877914,
      "num_tokens": 48355142.0,
      "step": 21665
    },
    {
      "entropy": 5.3230249881744385,
      "epoch": 2.4337620037064074,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004415145061673043,
      "loss": 5.0905,
      "mean_token_accuracy": 0.20696451067924498,
      "num_tokens": 48367342.0,
      "step": 21670
    },
    {
      "entropy": 5.347601747512817,
      "epoch": 2.434323580614365,
      "grad_norm": 1.125,
      "learning_rate": 0.0004414875020604096,
      "loss": 5.1166,
      "mean_token_accuracy": 0.20455300360918044,
      "num_tokens": 48378608.0,
      "step": 21675
    },
    {
      "entropy": 5.405285406112671,
      "epoch": 2.4348851575223227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004414604926524882,
      "loss": 5.1026,
      "mean_token_accuracy": 0.20413862764835358,
      "num_tokens": 48390455.0,
      "step": 21680
    },
    {
      "entropy": 5.422958421707153,
      "epoch": 2.4354467344302804,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004414334779444002,
      "loss": 5.0891,
      "mean_token_accuracy": 0.20364355593919753,
      "num_tokens": 48401723.0,
      "step": 21685
    },
    {
      "entropy": 5.3235921382904055,
      "epoch": 2.4360083113382376,
      "grad_norm": 0.984375,
      "learning_rate": 0.00044140645793700573,
      "loss": 5.1004,
      "mean_token_accuracy": 0.20195840895175934,
      "num_tokens": 48414208.0,
      "step": 21690
    },
    {
      "entropy": 5.3246602535247805,
      "epoch": 2.4365698882461952,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00044137943263116515,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20879900008440017,
      "num_tokens": 48425978.0,
      "step": 21695
    },
    {
      "entropy": 5.315085554122925,
      "epoch": 2.437131465154153,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004413524020277388,
      "loss": 4.9818,
      "mean_token_accuracy": 0.21727731674909592,
      "num_tokens": 48435988.0,
      "step": 21700
    },
    {
      "entropy": 5.354332971572876,
      "epoch": 2.4376930420621106,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004413253661275875,
      "loss": 5.0591,
      "mean_token_accuracy": 0.20790113955736161,
      "num_tokens": 48447217.0,
      "step": 21705
    },
    {
      "entropy": 5.310920715332031,
      "epoch": 2.438254618970068,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00044129832493157195,
      "loss": 5.1167,
      "mean_token_accuracy": 0.20421452969312667,
      "num_tokens": 48458597.0,
      "step": 21710
    },
    {
      "entropy": 5.281848382949829,
      "epoch": 2.4388161958780255,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044127127844055335,
      "loss": 5.0606,
      "mean_token_accuracy": 0.21008180379867553,
      "num_tokens": 48470601.0,
      "step": 21715
    },
    {
      "entropy": 5.37006106376648,
      "epoch": 2.439377772785983,
      "grad_norm": 1.140625,
      "learning_rate": 0.00044124422665539273,
      "loss": 5.0585,
      "mean_token_accuracy": 0.205963134765625,
      "num_tokens": 48481570.0,
      "step": 21720
    },
    {
      "entropy": 5.4434418201446535,
      "epoch": 2.4399393496939408,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004412171695769515,
      "loss": 5.2086,
      "mean_token_accuracy": 0.1992409661412239,
      "num_tokens": 48492146.0,
      "step": 21725
    },
    {
      "entropy": 5.340870189666748,
      "epoch": 2.440500926601898,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044119010720609106,
      "loss": 5.08,
      "mean_token_accuracy": 0.20240859389305116,
      "num_tokens": 48504644.0,
      "step": 21730
    },
    {
      "entropy": 5.309919643402099,
      "epoch": 2.4410625035098557,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044116303954367314,
      "loss": 5.0353,
      "mean_token_accuracy": 0.20731003135442733,
      "num_tokens": 48516405.0,
      "step": 21735
    },
    {
      "entropy": 5.315701293945312,
      "epoch": 2.4416240804178133,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004411359665905596,
      "loss": 5.1063,
      "mean_token_accuracy": 0.20202191323041915,
      "num_tokens": 48528759.0,
      "step": 21740
    },
    {
      "entropy": 5.483945512771607,
      "epoch": 2.442185657325771,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004411088883476125,
      "loss": 5.2121,
      "mean_token_accuracy": 0.1946186199784279,
      "num_tokens": 48540731.0,
      "step": 21745
    },
    {
      "entropy": 5.352399063110352,
      "epoch": 2.442747234233728,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00044108180481569387,
      "loss": 5.1385,
      "mean_token_accuracy": 0.2011985018849373,
      "num_tokens": 48553642.0,
      "step": 21750
    },
    {
      "entropy": 5.451523685455323,
      "epoch": 2.443308811141686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00044105471599566626,
      "loss": 5.1387,
      "mean_token_accuracy": 0.20625367313623427,
      "num_tokens": 48564596.0,
      "step": 21755
    },
    {
      "entropy": 5.390207433700562,
      "epoch": 2.4438703880496435,
      "grad_norm": 1.03125,
      "learning_rate": 0.000441027621888392,
      "loss": 5.1256,
      "mean_token_accuracy": 0.20331577956676483,
      "num_tokens": 48574521.0,
      "step": 21760
    },
    {
      "entropy": 5.302506351470948,
      "epoch": 2.4444319649576007,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044100052249473393,
      "loss": 5.0005,
      "mean_token_accuracy": 0.20864036828279495,
      "num_tokens": 48586299.0,
      "step": 21765
    },
    {
      "entropy": 5.383817577362061,
      "epoch": 2.4449935418655584,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004409734178155548,
      "loss": 5.1473,
      "mean_token_accuracy": 0.2089768499135971,
      "num_tokens": 48598366.0,
      "step": 21770
    },
    {
      "entropy": 5.3598494052886965,
      "epoch": 2.445555118773516,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0004409463078517178,
      "loss": 5.0579,
      "mean_token_accuracy": 0.2060109168291092,
      "num_tokens": 48609659.0,
      "step": 21775
    },
    {
      "entropy": 5.31979718208313,
      "epoch": 2.4461166956814737,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00044091919260408585,
      "loss": 5.0291,
      "mean_token_accuracy": 0.2140071287751198,
      "num_tokens": 48621623.0,
      "step": 21780
    },
    {
      "entropy": 5.3187220096588135,
      "epoch": 2.446678272589431,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004408920720735225,
      "loss": 4.9871,
      "mean_token_accuracy": 0.20146605968475342,
      "num_tokens": 48632711.0,
      "step": 21785
    },
    {
      "entropy": 5.326915884017945,
      "epoch": 2.4472398494973886,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044086494626089127,
      "loss": 5.102,
      "mean_token_accuracy": 0.20842339992523193,
      "num_tokens": 48644518.0,
      "step": 21790
    },
    {
      "entropy": 5.342644882202149,
      "epoch": 2.4478014264053463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004408378151670558,
      "loss": 5.0958,
      "mean_token_accuracy": 0.20284921675920486,
      "num_tokens": 48655261.0,
      "step": 21795
    },
    {
      "entropy": 5.301572036743164,
      "epoch": 2.448363003313304,
      "grad_norm": 0.90625,
      "learning_rate": 0.00044081067879288,
      "loss": 5.0318,
      "mean_token_accuracy": 0.21258021742105485,
      "num_tokens": 48666701.0,
      "step": 21800
    },
    {
      "entropy": 5.273501682281494,
      "epoch": 2.448924580221261,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004407835371392279,
      "loss": 4.9741,
      "mean_token_accuracy": 0.21670688092708587,
      "num_tokens": 48677919.0,
      "step": 21805
    },
    {
      "entropy": 5.307366943359375,
      "epoch": 2.449486157129219,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004407563902069636,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2145888015627861,
      "num_tokens": 48688680.0,
      "step": 21810
    },
    {
      "entropy": 5.414508056640625,
      "epoch": 2.4500477340371765,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004407292379969515,
      "loss": 5.0714,
      "mean_token_accuracy": 0.20862776190042495,
      "num_tokens": 48701149.0,
      "step": 21815
    },
    {
      "entropy": 5.307842588424682,
      "epoch": 2.450609310945134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004407020805100563,
      "loss": 5.0468,
      "mean_token_accuracy": 0.2060738831758499,
      "num_tokens": 48712236.0,
      "step": 21820
    },
    {
      "entropy": 5.292262125015259,
      "epoch": 2.4511708878530913,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004406749177471425,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21422230303287507,
      "num_tokens": 48723923.0,
      "step": 21825
    },
    {
      "entropy": 5.276024055480957,
      "epoch": 2.451732464761049,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044064774970907516,
      "loss": 5.0169,
      "mean_token_accuracy": 0.21020504534244538,
      "num_tokens": 48734925.0,
      "step": 21830
    },
    {
      "entropy": 5.335070419311523,
      "epoch": 2.4522940416690067,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004406205763967193,
      "loss": 4.9734,
      "mean_token_accuracy": 0.20905134379863738,
      "num_tokens": 48745758.0,
      "step": 21835
    },
    {
      "entropy": 5.257679128646851,
      "epoch": 2.4528556185769643,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0004405933978109398,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20426300913095474,
      "num_tokens": 48756899.0,
      "step": 21840
    },
    {
      "entropy": 5.3833784580230715,
      "epoch": 2.4534171954849215,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044056621395260247,
      "loss": 5.0573,
      "mean_token_accuracy": 0.209712353348732,
      "num_tokens": 48766805.0,
      "step": 21845
    },
    {
      "entropy": 5.374932909011841,
      "epoch": 2.453978772392879,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00044053902482257265,
      "loss": 5.1073,
      "mean_token_accuracy": 0.20937247574329376,
      "num_tokens": 48779360.0,
      "step": 21850
    },
    {
      "entropy": 5.319709777832031,
      "epoch": 2.454540349300837,
      "grad_norm": 1.015625,
      "learning_rate": 0.000440511830421716,
      "loss": 5.0509,
      "mean_token_accuracy": 0.20643526315689087,
      "num_tokens": 48789836.0,
      "step": 21855
    },
    {
      "entropy": 5.281533241271973,
      "epoch": 2.455101926208794,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004404846307508985,
      "loss": 5.046,
      "mean_token_accuracy": 0.20688605755567552,
      "num_tokens": 48801327.0,
      "step": 21860
    },
    {
      "entropy": 5.278894233703613,
      "epoch": 2.4556635031167517,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044045742581098624,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21253372430801393,
      "num_tokens": 48812521.0,
      "step": 21865
    },
    {
      "entropy": 5.293280839920044,
      "epoch": 2.4562250800247094,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004404302156028453,
      "loss": 5.0275,
      "mean_token_accuracy": 0.2045602411031723,
      "num_tokens": 48824068.0,
      "step": 21870
    },
    {
      "entropy": 5.324436473846435,
      "epoch": 2.456786656932667,
      "grad_norm": 1.125,
      "learning_rate": 0.00044040300012734217,
      "loss": 5.0726,
      "mean_token_accuracy": 0.20400148630142212,
      "num_tokens": 48835183.0,
      "step": 21875
    },
    {
      "entropy": 5.36867094039917,
      "epoch": 2.4573482338406243,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004403757793853434,
      "loss": 5.1095,
      "mean_token_accuracy": 0.20888590067625046,
      "num_tokens": 48846616.0,
      "step": 21880
    },
    {
      "entropy": 5.3105193138122555,
      "epoch": 2.457909810748582,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004403485533777157,
      "loss": 5.0804,
      "mean_token_accuracy": 0.20703742802143096,
      "num_tokens": 48858159.0,
      "step": 21885
    },
    {
      "entropy": 5.343253517150879,
      "epoch": 2.4584713876565396,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004403213221053258,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21365494132041932,
      "num_tokens": 48870363.0,
      "step": 21890
    },
    {
      "entropy": 5.347671175003052,
      "epoch": 2.4590329645644973,
      "grad_norm": 1.078125,
      "learning_rate": 0.00044029408556904093,
      "loss": 5.1359,
      "mean_token_accuracy": 0.20451768934726716,
      "num_tokens": 48882036.0,
      "step": 21895
    },
    {
      "entropy": 5.366845655441284,
      "epoch": 2.4595945414724545,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004402668437697283,
      "loss": 5.09,
      "mean_token_accuracy": 0.20429982095956803,
      "num_tokens": 48893741.0,
      "step": 21900
    },
    {
      "entropy": 5.451215171813965,
      "epoch": 2.460156118380412,
      "grad_norm": 1.0625,
      "learning_rate": 0.00044023959670825533,
      "loss": 5.1748,
      "mean_token_accuracy": 0.20167700350284576,
      "num_tokens": 48906260.0,
      "step": 21905
    },
    {
      "entropy": 5.396780872344971,
      "epoch": 2.46071769528837,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004402123443854894,
      "loss": 5.1396,
      "mean_token_accuracy": 0.197510227560997,
      "num_tokens": 48918078.0,
      "step": 21910
    },
    {
      "entropy": 5.365047121047974,
      "epoch": 2.4612792721963275,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004401850868022984,
      "loss": 5.0776,
      "mean_token_accuracy": 0.20537584275007248,
      "num_tokens": 48928999.0,
      "step": 21915
    },
    {
      "entropy": 5.269875240325928,
      "epoch": 2.4618408491042847,
      "grad_norm": 1.0,
      "learning_rate": 0.0004401578239595501,
      "loss": 5.0339,
      "mean_token_accuracy": 0.20497289150953293,
      "num_tokens": 48940587.0,
      "step": 21920
    },
    {
      "entropy": 5.342274236679077,
      "epoch": 2.4624024260122424,
      "grad_norm": 1.03125,
      "learning_rate": 0.00044013055585811274,
      "loss": 5.1734,
      "mean_token_accuracy": 0.21042147427797317,
      "num_tokens": 48951632.0,
      "step": 21925
    },
    {
      "entropy": 5.445606517791748,
      "epoch": 2.4629640029202,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00044010328249885434,
      "loss": 5.237,
      "mean_token_accuracy": 0.20087085217237471,
      "num_tokens": 48963578.0,
      "step": 21930
    },
    {
      "entropy": 5.4253215312957765,
      "epoch": 2.4635255798281577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004400760038826434,
      "loss": 5.0802,
      "mean_token_accuracy": 0.20590412020683288,
      "num_tokens": 48974690.0,
      "step": 21935
    },
    {
      "entropy": 5.387612533569336,
      "epoch": 2.464087156736115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00044004872001034856,
      "loss": 5.1526,
      "mean_token_accuracy": 0.20595055669546128,
      "num_tokens": 48985871.0,
      "step": 21940
    },
    {
      "entropy": 5.263420867919922,
      "epoch": 2.4646487336440726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004400214308828384,
      "loss": 4.999,
      "mean_token_accuracy": 0.2128686174750328,
      "num_tokens": 48996475.0,
      "step": 21945
    },
    {
      "entropy": 5.425986289978027,
      "epoch": 2.46521031055203,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00043999413650098193,
      "loss": 5.1695,
      "mean_token_accuracy": 0.20305923968553544,
      "num_tokens": 49008900.0,
      "step": 21950
    },
    {
      "entropy": 5.28187575340271,
      "epoch": 2.4657718874599874,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004399668368656481,
      "loss": 5.0233,
      "mean_token_accuracy": 0.2087813511490822,
      "num_tokens": 49020583.0,
      "step": 21955
    },
    {
      "entropy": 5.404441881179809,
      "epoch": 2.466333464367945,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004399395319777062,
      "loss": 5.1437,
      "mean_token_accuracy": 0.19677996933460234,
      "num_tokens": 49031363.0,
      "step": 21960
    },
    {
      "entropy": 5.352469253540039,
      "epoch": 2.4668950412759028,
      "grad_norm": 1.265625,
      "learning_rate": 0.00043991222183802564,
      "loss": 5.1315,
      "mean_token_accuracy": 0.20676130503416063,
      "num_tokens": 49043404.0,
      "step": 21965
    },
    {
      "entropy": 5.424435806274414,
      "epoch": 2.4674566181838604,
      "grad_norm": 1.078125,
      "learning_rate": 0.000439884906447476,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20447663515806197,
      "num_tokens": 49054240.0,
      "step": 21970
    },
    {
      "entropy": 5.326626443862915,
      "epoch": 2.4680181950918176,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000439857585806927,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21729060560464858,
      "num_tokens": 49064730.0,
      "step": 21975
    },
    {
      "entropy": 5.38746829032898,
      "epoch": 2.4685797719997753,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043983025991724847,
      "loss": 5.1004,
      "mean_token_accuracy": 0.2054046183824539,
      "num_tokens": 49075218.0,
      "step": 21980
    },
    {
      "entropy": 5.3066481590271,
      "epoch": 2.469141348907733,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004398029287793106,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21284084469079972,
      "num_tokens": 49086182.0,
      "step": 21985
    },
    {
      "entropy": 5.374128580093384,
      "epoch": 2.4697029258156906,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00043977559239398353,
      "loss": 5.1611,
      "mean_token_accuracy": 0.19420846104621886,
      "num_tokens": 49099054.0,
      "step": 21990
    },
    {
      "entropy": 5.332332801818848,
      "epoch": 2.470264502723648,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004397482507621376,
      "loss": 5.0911,
      "mean_token_accuracy": 0.21246299743652344,
      "num_tokens": 49110234.0,
      "step": 21995
    },
    {
      "entropy": 5.326526260375976,
      "epoch": 2.4708260796316055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004397209038846436,
      "loss": 5.0466,
      "mean_token_accuracy": 0.2082221120595932,
      "num_tokens": 49121884.0,
      "step": 22000
    },
    {
      "entropy": 5.308582448959351,
      "epoch": 2.471387656539563,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043969355176237207,
      "loss": 5.025,
      "mean_token_accuracy": 0.20396019965410234,
      "num_tokens": 49133491.0,
      "step": 22005
    },
    {
      "entropy": 5.268691158294677,
      "epoch": 2.471949233447521,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004396661943961941,
      "loss": 5.0111,
      "mean_token_accuracy": 0.20542235672473907,
      "num_tokens": 49145895.0,
      "step": 22010
    },
    {
      "entropy": 5.428597402572632,
      "epoch": 2.472510810355478,
      "grad_norm": 1.125,
      "learning_rate": 0.0004396388317869805,
      "loss": 5.1267,
      "mean_token_accuracy": 0.20497975498437881,
      "num_tokens": 49157840.0,
      "step": 22015
    },
    {
      "entropy": 5.3687926769256595,
      "epoch": 2.4730723872634357,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004396114639356027,
      "loss": 5.0873,
      "mean_token_accuracy": 0.21366675049066544,
      "num_tokens": 49170391.0,
      "step": 22020
    },
    {
      "entropy": 5.347096347808838,
      "epoch": 2.4736339641713934,
      "grad_norm": 1.078125,
      "learning_rate": 0.000439584090842932,
      "loss": 5.0651,
      "mean_token_accuracy": 0.2035873830318451,
      "num_tokens": 49181721.0,
      "step": 22025
    },
    {
      "entropy": 5.331093502044678,
      "epoch": 2.474195541079351,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043955671250984005,
      "loss": 5.0872,
      "mean_token_accuracy": 0.20745128840208055,
      "num_tokens": 49192137.0,
      "step": 22030
    },
    {
      "entropy": 5.4110668182373045,
      "epoch": 2.4747571179873082,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043952932893719847,
      "loss": 5.1278,
      "mean_token_accuracy": 0.20717046409845352,
      "num_tokens": 49203260.0,
      "step": 22035
    },
    {
      "entropy": 5.344753456115723,
      "epoch": 2.475318694895266,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004395019401258794,
      "loss": 5.0043,
      "mean_token_accuracy": 0.20788085460662842,
      "num_tokens": 49214799.0,
      "step": 22040
    },
    {
      "entropy": 5.3041667461395265,
      "epoch": 2.4758802718032236,
      "grad_norm": 1.125,
      "learning_rate": 0.00043947454607675464,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2132069170475006,
      "num_tokens": 49224595.0,
      "step": 22045
    },
    {
      "entropy": 5.3213118553161625,
      "epoch": 2.476441848711181,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004394471467906966,
      "loss": 5.0609,
      "mean_token_accuracy": 0.21107788532972335,
      "num_tokens": 49235190.0,
      "step": 22050
    },
    {
      "entropy": 5.289295625686646,
      "epoch": 2.4770034256191384,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043941974226857763,
      "loss": 5.0346,
      "mean_token_accuracy": 0.2081354707479477,
      "num_tokens": 49247103.0,
      "step": 22055
    },
    {
      "entropy": 5.383853816986084,
      "epoch": 2.477565002527096,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004393923325112702,
      "loss": 5.1315,
      "mean_token_accuracy": 0.19898751974105836,
      "num_tokens": 49259874.0,
      "step": 22060
    },
    {
      "entropy": 5.240247774124145,
      "epoch": 2.4781265794350538,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004393649175196472,
      "loss": 4.9652,
      "mean_token_accuracy": 0.21667101234197617,
      "num_tokens": 49271059.0,
      "step": 22065
    },
    {
      "entropy": 5.282599878311157,
      "epoch": 2.478688156343011,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004393374972945815,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21667844653129578,
      "num_tokens": 49283142.0,
      "step": 22070
    },
    {
      "entropy": 5.356895399093628,
      "epoch": 2.4792497332509686,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043931007183694616,
      "loss": 5.0521,
      "mean_token_accuracy": 0.21300241947174073,
      "num_tokens": 49294022.0,
      "step": 22075
    },
    {
      "entropy": 5.345976972579956,
      "epoch": 2.4798113101589263,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043928264114761437,
      "loss": 5.0829,
      "mean_token_accuracy": 0.20595527142286302,
      "num_tokens": 49305849.0,
      "step": 22080
    },
    {
      "entropy": 5.332183599472046,
      "epoch": 2.480372887066884,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043925520522745953,
      "loss": 5.0618,
      "mean_token_accuracy": 0.2072546273469925,
      "num_tokens": 49316957.0,
      "step": 22085
    },
    {
      "entropy": 5.330669355392456,
      "epoch": 2.480934463974841,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004392277640773553,
      "loss": 5.0359,
      "mean_token_accuracy": 0.21170599460601808,
      "num_tokens": 49329040.0,
      "step": 22090
    },
    {
      "entropy": 5.311238145828247,
      "epoch": 2.481496040882799,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043920031769817543,
      "loss": 5.0817,
      "mean_token_accuracy": 0.20988859236240387,
      "num_tokens": 49340037.0,
      "step": 22095
    },
    {
      "entropy": 5.272458791732788,
      "epoch": 2.4820576177907565,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043917286609079364,
      "loss": 5.0647,
      "mean_token_accuracy": 0.20450074821710587,
      "num_tokens": 49350059.0,
      "step": 22100
    },
    {
      "entropy": 5.300284147262573,
      "epoch": 2.482619194698714,
      "grad_norm": 1.0,
      "learning_rate": 0.00043914540925608426,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21294322311878205,
      "num_tokens": 49360880.0,
      "step": 22105
    },
    {
      "entropy": 5.417384147644043,
      "epoch": 2.4831807716066714,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004391179471949213,
      "loss": 5.0279,
      "mean_token_accuracy": 0.20861021280288697,
      "num_tokens": 49370725.0,
      "step": 22110
    },
    {
      "entropy": 5.337517213821411,
      "epoch": 2.483742348514629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043909047990817917,
      "loss": 5.042,
      "mean_token_accuracy": 0.20742518454790115,
      "num_tokens": 49380991.0,
      "step": 22115
    },
    {
      "entropy": 5.242034387588501,
      "epoch": 2.4843039254225867,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043906300739673265,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20722330063581468,
      "num_tokens": 49390802.0,
      "step": 22120
    },
    {
      "entropy": 5.361000633239746,
      "epoch": 2.4848655023305444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043903552966145624,
      "loss": 5.0612,
      "mean_token_accuracy": 0.2088610291481018,
      "num_tokens": 49401507.0,
      "step": 22125
    },
    {
      "entropy": 5.38690299987793,
      "epoch": 2.4854270792385016,
      "grad_norm": 1.0625,
      "learning_rate": 0.000439008046703225,
      "loss": 5.1416,
      "mean_token_accuracy": 0.20007878690958023,
      "num_tokens": 49412022.0,
      "step": 22130
    },
    {
      "entropy": 5.334241580963135,
      "epoch": 2.4859886561464593,
      "grad_norm": 1.078125,
      "learning_rate": 0.000438980558522914,
      "loss": 5.0179,
      "mean_token_accuracy": 0.21491591036319732,
      "num_tokens": 49423294.0,
      "step": 22135
    },
    {
      "entropy": 5.294369840621949,
      "epoch": 2.486550233054417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004389530651213983,
      "loss": 4.98,
      "mean_token_accuracy": 0.20720698684453964,
      "num_tokens": 49433915.0,
      "step": 22140
    },
    {
      "entropy": 5.270742654800415,
      "epoch": 2.487111809962374,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043892556649955354,
      "loss": 5.099,
      "mean_token_accuracy": 0.20606036633253097,
      "num_tokens": 49445122.0,
      "step": 22145
    },
    {
      "entropy": 5.508612060546875,
      "epoch": 2.487673386870332,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004388980626582551,
      "loss": 5.2356,
      "mean_token_accuracy": 0.20265230685472488,
      "num_tokens": 49456276.0,
      "step": 22150
    },
    {
      "entropy": 5.373865556716919,
      "epoch": 2.4882349637782895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004388705535983788,
      "loss": 5.1161,
      "mean_token_accuracy": 0.2051602527499199,
      "num_tokens": 49467346.0,
      "step": 22155
    },
    {
      "entropy": 5.3075262069702145,
      "epoch": 2.488796540686247,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004388430393208005,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21123238205909728,
      "num_tokens": 49477635.0,
      "step": 22160
    },
    {
      "entropy": 5.335418272018432,
      "epoch": 2.4893581175942043,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004388155198263963,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20442467480897902,
      "num_tokens": 49491060.0,
      "step": 22165
    },
    {
      "entropy": 5.3818121433258055,
      "epoch": 2.489919694502162,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004387879951160424,
      "loss": 5.0455,
      "mean_token_accuracy": 0.21197994649410248,
      "num_tokens": 49502551.0,
      "step": 22170
    },
    {
      "entropy": 5.330615520477295,
      "epoch": 2.4904812714101197,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004387604651906152,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20503442287445067,
      "num_tokens": 49514505.0,
      "step": 22175
    },
    {
      "entropy": 5.256308507919312,
      "epoch": 2.4910428483180773,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004387329300509913,
      "loss": 4.9839,
      "mean_token_accuracy": 0.2080680325627327,
      "num_tokens": 49526301.0,
      "step": 22180
    },
    {
      "entropy": 5.249794816970825,
      "epoch": 2.4916044252260345,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043870538969804733,
      "loss": 5.0281,
      "mean_token_accuracy": 0.20875300765037536,
      "num_tokens": 49536546.0,
      "step": 22185
    },
    {
      "entropy": 5.409606504440307,
      "epoch": 2.492166002133992,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004386778441326603,
      "loss": 5.1506,
      "mean_token_accuracy": 0.20125253945589067,
      "num_tokens": 49548046.0,
      "step": 22190
    },
    {
      "entropy": 5.307662296295166,
      "epoch": 2.49272757904195,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00043865029335570725,
      "loss": 4.9987,
      "mean_token_accuracy": 0.21150004863739014,
      "num_tokens": 49560001.0,
      "step": 22195
    },
    {
      "entropy": 5.257001876831055,
      "epoch": 2.4932891559499075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004386227373680655,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21470770984888077,
      "num_tokens": 49571111.0,
      "step": 22200
    },
    {
      "entropy": 5.283322048187256,
      "epoch": 2.4938507328578647,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0004385951761706122,
      "loss": 5.005,
      "mean_token_accuracy": 0.2104405477643013,
      "num_tokens": 49582888.0,
      "step": 22205
    },
    {
      "entropy": 5.275071287155152,
      "epoch": 2.4944123097658224,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043856760976422506,
      "loss": 5.033,
      "mean_token_accuracy": 0.21279434859752655,
      "num_tokens": 49593710.0,
      "step": 22210
    },
    {
      "entropy": 5.3664508819580075,
      "epoch": 2.49497388667378,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043854003814978174,
      "loss": 5.09,
      "mean_token_accuracy": 0.20884661227464676,
      "num_tokens": 49604263.0,
      "step": 22215
    },
    {
      "entropy": 5.424262142181396,
      "epoch": 2.4955354635817377,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043851246132816025,
      "loss": 5.153,
      "mean_token_accuracy": 0.20305848121643066,
      "num_tokens": 49615152.0,
      "step": 22220
    },
    {
      "entropy": 5.420793151855468,
      "epoch": 2.496097040489695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004384848793002385,
      "loss": 5.1516,
      "mean_token_accuracy": 0.19985629469156266,
      "num_tokens": 49627778.0,
      "step": 22225
    },
    {
      "entropy": 5.36041226387024,
      "epoch": 2.4966586173976526,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004384572920668949,
      "loss": 5.0504,
      "mean_token_accuracy": 0.20892155766487122,
      "num_tokens": 49637701.0,
      "step": 22230
    },
    {
      "entropy": 5.283799266815185,
      "epoch": 2.4972201943056103,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004384296996290076,
      "loss": 5.0204,
      "mean_token_accuracy": 0.2194039762020111,
      "num_tokens": 49647680.0,
      "step": 22235
    },
    {
      "entropy": 5.390938949584961,
      "epoch": 2.4977817712135675,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00043840210198745533,
      "loss": 5.184,
      "mean_token_accuracy": 0.20452556014060974,
      "num_tokens": 49659419.0,
      "step": 22240
    },
    {
      "entropy": 5.322469711303711,
      "epoch": 2.498343348121525,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043837449914311676,
      "loss": 5.0772,
      "mean_token_accuracy": 0.21402415186166762,
      "num_tokens": 49671131.0,
      "step": 22245
    },
    {
      "entropy": 5.304863834381104,
      "epoch": 2.498904925029483,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043834689109687077,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2105807274580002,
      "num_tokens": 49681748.0,
      "step": 22250
    },
    {
      "entropy": 5.428316307067871,
      "epoch": 2.4994665019374405,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004383192778495964,
      "loss": 5.0864,
      "mean_token_accuracy": 0.2132682755589485,
      "num_tokens": 49693321.0,
      "step": 22255
    },
    {
      "entropy": 5.327401351928711,
      "epoch": 2.500028078845398,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000438291659402173,
      "loss": 5.0825,
      "mean_token_accuracy": 0.21267523467540742,
      "num_tokens": 49704419.0,
      "step": 22260
    },
    {
      "entropy": 5.358323097229004,
      "epoch": 2.5005896557533553,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043826403575547974,
      "loss": 5.0932,
      "mean_token_accuracy": 0.2085675746202469,
      "num_tokens": 49715692.0,
      "step": 22265
    },
    {
      "entropy": 5.422159147262573,
      "epoch": 2.501151232661313,
      "grad_norm": 1.1875,
      "learning_rate": 0.0004382364069103962,
      "loss": 5.1133,
      "mean_token_accuracy": 0.2023362249135971,
      "num_tokens": 49726324.0,
      "step": 22270
    },
    {
      "entropy": 5.275172376632691,
      "epoch": 2.5017128095692707,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043820877286780224,
      "loss": 5.0232,
      "mean_token_accuracy": 0.21031608581542968,
      "num_tokens": 49737069.0,
      "step": 22275
    },
    {
      "entropy": 5.318316316604614,
      "epoch": 2.502274386477228,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004381811336285776,
      "loss": 5.028,
      "mean_token_accuracy": 0.20997966825962067,
      "num_tokens": 49748507.0,
      "step": 22280
    },
    {
      "entropy": 5.350791120529175,
      "epoch": 2.5028359633851855,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004381534891936024,
      "loss": 5.018,
      "mean_token_accuracy": 0.20484119802713394,
      "num_tokens": 49759017.0,
      "step": 22285
    },
    {
      "entropy": 5.4028074741363525,
      "epoch": 2.503397540293143,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004381258395637568,
      "loss": 5.0868,
      "mean_token_accuracy": 0.20757929384708404,
      "num_tokens": 49771353.0,
      "step": 22290
    },
    {
      "entropy": 5.42965350151062,
      "epoch": 2.5039591172011004,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004380981847399212,
      "loss": 5.2158,
      "mean_token_accuracy": 0.19617853611707686,
      "num_tokens": 49782337.0,
      "step": 22295
    },
    {
      "entropy": 5.338736057281494,
      "epoch": 2.504520694109058,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043807052472297626,
      "loss": 4.9953,
      "mean_token_accuracy": 0.20998046696186065,
      "num_tokens": 49794212.0,
      "step": 22300
    },
    {
      "entropy": 5.293231058120727,
      "epoch": 2.5050822710170157,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004380428595138024,
      "loss": 5.0323,
      "mean_token_accuracy": 0.21372945755720138,
      "num_tokens": 49804602.0,
      "step": 22305
    },
    {
      "entropy": 5.397088479995728,
      "epoch": 2.5056438479249734,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043801518911328074,
      "loss": 5.1926,
      "mean_token_accuracy": 0.196171735227108,
      "num_tokens": 49816277.0,
      "step": 22310
    },
    {
      "entropy": 5.385932922363281,
      "epoch": 2.506205424832931,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043798751352229227,
      "loss": 5.072,
      "mean_token_accuracy": 0.21231390535831451,
      "num_tokens": 49827851.0,
      "step": 22315
    },
    {
      "entropy": 5.299003839492798,
      "epoch": 2.5067670017408883,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043795983274171806,
      "loss": 5.086,
      "mean_token_accuracy": 0.1997688814997673,
      "num_tokens": 49840492.0,
      "step": 22320
    },
    {
      "entropy": 5.226717567443847,
      "epoch": 2.507328578648846,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043793214677243967,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21143775582313537,
      "num_tokens": 49851427.0,
      "step": 22325
    },
    {
      "entropy": 5.343773984909058,
      "epoch": 2.5078901555568036,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004379044556153384,
      "loss": 5.0357,
      "mean_token_accuracy": 0.21013484448194503,
      "num_tokens": 49862883.0,
      "step": 22330
    },
    {
      "entropy": 5.332253694534302,
      "epoch": 2.508451732464761,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043787675927129625,
      "loss": 5.0386,
      "mean_token_accuracy": 0.2046024590730667,
      "num_tokens": 49874120.0,
      "step": 22335
    },
    {
      "entropy": 5.426383972167969,
      "epoch": 2.5090133093727185,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043784905774119483,
      "loss": 5.1084,
      "mean_token_accuracy": 0.20945753306150436,
      "num_tokens": 49885961.0,
      "step": 22340
    },
    {
      "entropy": 5.2519114971160885,
      "epoch": 2.509574886280676,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043782135102591626,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2134530246257782,
      "num_tokens": 49896821.0,
      "step": 22345
    },
    {
      "entropy": 5.205495929718017,
      "epoch": 2.510136463188634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004377936391263427,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21247723549604416,
      "num_tokens": 49908148.0,
      "step": 22350
    },
    {
      "entropy": 5.378243112564087,
      "epoch": 2.5106980400965915,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043776592204335655,
      "loss": 5.0911,
      "mean_token_accuracy": 0.20830341428518295,
      "num_tokens": 49919487.0,
      "step": 22355
    },
    {
      "entropy": 5.340674877166748,
      "epoch": 2.5112596170045487,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004377381997778403,
      "loss": 5.0451,
      "mean_token_accuracy": 0.21256323903799057,
      "num_tokens": 49930722.0,
      "step": 22360
    },
    {
      "entropy": 5.377729034423828,
      "epoch": 2.5118211939125064,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004377104723306767,
      "loss": 5.1523,
      "mean_token_accuracy": 0.19856999665498734,
      "num_tokens": 49941917.0,
      "step": 22365
    },
    {
      "entropy": 5.395837736129761,
      "epoch": 2.512382770820464,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004376827397027485,
      "loss": 5.0924,
      "mean_token_accuracy": 0.2015238881111145,
      "num_tokens": 49952681.0,
      "step": 22370
    },
    {
      "entropy": 5.393644094467163,
      "epoch": 2.5129443477284212,
      "grad_norm": 1.0,
      "learning_rate": 0.0004376550018949388,
      "loss": 5.0872,
      "mean_token_accuracy": 0.20748354792594909,
      "num_tokens": 49965542.0,
      "step": 22375
    },
    {
      "entropy": 5.311571216583252,
      "epoch": 2.513505924636379,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043762725890813083,
      "loss": 5.1372,
      "mean_token_accuracy": 0.20329391360282897,
      "num_tokens": 49977506.0,
      "step": 22380
    },
    {
      "entropy": 5.3263765335083,
      "epoch": 2.5140675015443366,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004375995107432077,
      "loss": 5.0844,
      "mean_token_accuracy": 0.20371172875165938,
      "num_tokens": 49988674.0,
      "step": 22385
    },
    {
      "entropy": 5.334761333465576,
      "epoch": 2.5146290784522938,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043757175740105316,
      "loss": 5.1035,
      "mean_token_accuracy": 0.20265783369541168,
      "num_tokens": 50001618.0,
      "step": 22390
    },
    {
      "entropy": 5.338435459136963,
      "epoch": 2.5151906553602514,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004375439988825508,
      "loss": 5.0353,
      "mean_token_accuracy": 0.21192275881767272,
      "num_tokens": 50012556.0,
      "step": 22395
    },
    {
      "entropy": 5.368850135803223,
      "epoch": 2.515752232268209,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004375162351885845,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21421469151973724,
      "num_tokens": 50023399.0,
      "step": 22400
    },
    {
      "entropy": 5.265899181365967,
      "epoch": 2.5163138091761668,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043748846632003823,
      "loss": 5.0237,
      "mean_token_accuracy": 0.20859778225421904,
      "num_tokens": 50035209.0,
      "step": 22405
    },
    {
      "entropy": 5.2753252506256105,
      "epoch": 2.5168753860841244,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043746069227779615,
      "loss": 5.0692,
      "mean_token_accuracy": 0.20776693671941757,
      "num_tokens": 50046383.0,
      "step": 22410
    },
    {
      "entropy": 5.352760362625122,
      "epoch": 2.5174369629920816,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004374329130627427,
      "loss": 5.0941,
      "mean_token_accuracy": 0.2113133564591408,
      "num_tokens": 50058785.0,
      "step": 22415
    },
    {
      "entropy": 5.398253297805786,
      "epoch": 2.5179985399000393,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043740512867576224,
      "loss": 5.0775,
      "mean_token_accuracy": 0.2068310335278511,
      "num_tokens": 50069629.0,
      "step": 22420
    },
    {
      "entropy": 5.404817390441894,
      "epoch": 2.518560116807997,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004373773391177395,
      "loss": 5.0604,
      "mean_token_accuracy": 0.2067250519990921,
      "num_tokens": 50080923.0,
      "step": 22425
    },
    {
      "entropy": 5.308856582641601,
      "epoch": 2.519121693715954,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004373495443895593,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21209421157836914,
      "num_tokens": 50092133.0,
      "step": 22430
    },
    {
      "entropy": 5.228975439071656,
      "epoch": 2.519683270623912,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004373217444921066,
      "loss": 4.9935,
      "mean_token_accuracy": 0.21204879134893417,
      "num_tokens": 50104001.0,
      "step": 22435
    },
    {
      "entropy": 5.351526117324829,
      "epoch": 2.5202448475318695,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043729393942626665,
      "loss": 5.0962,
      "mean_token_accuracy": 0.21241909861564637,
      "num_tokens": 50115908.0,
      "step": 22440
    },
    {
      "entropy": 5.361494874954223,
      "epoch": 2.520806424439827,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043726612919292476,
      "loss": 5.0897,
      "mean_token_accuracy": 0.20498345792293549,
      "num_tokens": 50127380.0,
      "step": 22445
    },
    {
      "entropy": 5.39303035736084,
      "epoch": 2.521368001347785,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004372383137929663,
      "loss": 5.0217,
      "mean_token_accuracy": 0.2097218871116638,
      "num_tokens": 50138458.0,
      "step": 22450
    },
    {
      "entropy": 5.41278600692749,
      "epoch": 2.521929578255742,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043721049322727705,
      "loss": 5.1698,
      "mean_token_accuracy": 0.20354342162609101,
      "num_tokens": 50148814.0,
      "step": 22455
    },
    {
      "entropy": 5.4267950534820555,
      "epoch": 2.5224911551636997,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004371826674967428,
      "loss": 5.1511,
      "mean_token_accuracy": 0.1972817525267601,
      "num_tokens": 50159233.0,
      "step": 22460
    },
    {
      "entropy": 5.341180992126465,
      "epoch": 2.5230527320716574,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043715483660224943,
      "loss": 5.0576,
      "mean_token_accuracy": 0.2070496693253517,
      "num_tokens": 50170752.0,
      "step": 22465
    },
    {
      "entropy": 5.276272106170654,
      "epoch": 2.5236143089796146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004371270005446833,
      "loss": 5.0456,
      "mean_token_accuracy": 0.20848852396011353,
      "num_tokens": 50183759.0,
      "step": 22470
    },
    {
      "entropy": 5.328631830215454,
      "epoch": 2.5241758858875722,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004370991593249305,
      "loss": 4.985,
      "mean_token_accuracy": 0.21318092495203017,
      "num_tokens": 50194091.0,
      "step": 22475
    },
    {
      "entropy": 5.359835815429688,
      "epoch": 2.52473746279553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043707131294387766,
      "loss": 5.0664,
      "mean_token_accuracy": 0.20873967558145523,
      "num_tokens": 50204591.0,
      "step": 22480
    },
    {
      "entropy": 5.268936824798584,
      "epoch": 2.525299039703487,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004370434614024113,
      "loss": 5.0009,
      "mean_token_accuracy": 0.208926922082901,
      "num_tokens": 50215809.0,
      "step": 22485
    },
    {
      "entropy": 5.307646131515503,
      "epoch": 2.525860616611445,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043701560470141824,
      "loss": 5.0976,
      "mean_token_accuracy": 0.20810645669698716,
      "num_tokens": 50227038.0,
      "step": 22490
    },
    {
      "entropy": 5.341178035736084,
      "epoch": 2.5264221935194024,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004369877428417856,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20956032127141952,
      "num_tokens": 50237001.0,
      "step": 22495
    },
    {
      "entropy": 5.368230104446411,
      "epoch": 2.52698377042736,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004369598758244003,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21598497927188873,
      "num_tokens": 50248389.0,
      "step": 22500
    },
    {
      "entropy": 5.269075965881347,
      "epoch": 2.5275453473353178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043693200365014985,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21188065260648728,
      "num_tokens": 50259676.0,
      "step": 22505
    },
    {
      "entropy": 5.391644811630249,
      "epoch": 2.528106924243275,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043690412631992154,
      "loss": 5.1618,
      "mean_token_accuracy": 0.20563230961561202,
      "num_tokens": 50270707.0,
      "step": 22510
    },
    {
      "entropy": 5.3331098556518555,
      "epoch": 2.5286685011512327,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043687624383460306,
      "loss": 4.9415,
      "mean_token_accuracy": 0.21821001172065735,
      "num_tokens": 50282187.0,
      "step": 22515
    },
    {
      "entropy": 5.368975591659546,
      "epoch": 2.5292300780591903,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043684835619508215,
      "loss": 5.0991,
      "mean_token_accuracy": 0.21162377297878265,
      "num_tokens": 50293672.0,
      "step": 22520
    },
    {
      "entropy": 5.298235177993774,
      "epoch": 2.5297916549671475,
      "grad_norm": 1.0,
      "learning_rate": 0.0004368204634022469,
      "loss": 4.9424,
      "mean_token_accuracy": 0.20941642224788665,
      "num_tokens": 50305235.0,
      "step": 22525
    },
    {
      "entropy": 5.35415506362915,
      "epoch": 2.530353231875105,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004367925654569852,
      "loss": 5.119,
      "mean_token_accuracy": 0.2026696652173996,
      "num_tokens": 50316178.0,
      "step": 22530
    },
    {
      "entropy": 5.320326423645019,
      "epoch": 2.530914808783063,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004367646623601856,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21312112212181092,
      "num_tokens": 50326713.0,
      "step": 22535
    },
    {
      "entropy": 5.362862253189087,
      "epoch": 2.5314763856910205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004367367541127364,
      "loss": 5.18,
      "mean_token_accuracy": 0.20388969033956528,
      "num_tokens": 50338648.0,
      "step": 22540
    },
    {
      "entropy": 5.378815746307373,
      "epoch": 2.532037962598978,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004367088407155261,
      "loss": 5.0337,
      "mean_token_accuracy": 0.21517971754074097,
      "num_tokens": 50350134.0,
      "step": 22545
    },
    {
      "entropy": 5.276973104476928,
      "epoch": 2.5325995395069354,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004366809221694437,
      "loss": 5.0072,
      "mean_token_accuracy": 0.21745716035366058,
      "num_tokens": 50361101.0,
      "step": 22550
    },
    {
      "entropy": 5.246868324279785,
      "epoch": 2.533161116414893,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043665299847537805,
      "loss": 5.0074,
      "mean_token_accuracy": 0.20485426783561705,
      "num_tokens": 50372376.0,
      "step": 22555
    },
    {
      "entropy": 5.364133548736572,
      "epoch": 2.5337226933228507,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043662506963421815,
      "loss": 5.0412,
      "mean_token_accuracy": 0.21227156221866608,
      "num_tokens": 50383494.0,
      "step": 22560
    },
    {
      "entropy": 5.314332675933838,
      "epoch": 2.534284270230808,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004365971356468534,
      "loss": 5.0758,
      "mean_token_accuracy": 0.20457465946674347,
      "num_tokens": 50394720.0,
      "step": 22565
    },
    {
      "entropy": 5.416807365417481,
      "epoch": 2.5348458471387656,
      "grad_norm": 1.09375,
      "learning_rate": 0.00043656919651417314,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20735086649656295,
      "num_tokens": 50407109.0,
      "step": 22570
    },
    {
      "entropy": 5.464645004272461,
      "epoch": 2.5354074240467233,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000436541252237067,
      "loss": 5.1814,
      "mean_token_accuracy": 0.20354954302310943,
      "num_tokens": 50418664.0,
      "step": 22575
    },
    {
      "entropy": 5.289036083221435,
      "epoch": 2.5359690009546805,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004365133028164248,
      "loss": 5.0008,
      "mean_token_accuracy": 0.20987916141748428,
      "num_tokens": 50431069.0,
      "step": 22580
    },
    {
      "entropy": 5.340060567855835,
      "epoch": 2.536530577862638,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004364853482531363,
      "loss": 5.0551,
      "mean_token_accuracy": 0.20651256591081618,
      "num_tokens": 50441484.0,
      "step": 22585
    },
    {
      "entropy": 5.269764709472656,
      "epoch": 2.537092154770596,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043645738854809177,
      "loss": 5.0003,
      "mean_token_accuracy": 0.2142564058303833,
      "num_tokens": 50452594.0,
      "step": 22590
    },
    {
      "entropy": 5.371035766601563,
      "epoch": 2.5376537316785535,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004364294237021813,
      "loss": 5.1126,
      "mean_token_accuracy": 0.20243385136127473,
      "num_tokens": 50463342.0,
      "step": 22595
    },
    {
      "entropy": 5.345622777938843,
      "epoch": 2.538215308586511,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043640145371629544,
      "loss": 5.0991,
      "mean_token_accuracy": 0.20302331298589707,
      "num_tokens": 50475164.0,
      "step": 22600
    },
    {
      "entropy": 5.408536195755005,
      "epoch": 2.5387768854944683,
      "grad_norm": 1.0,
      "learning_rate": 0.0004363734785913246,
      "loss": 5.1324,
      "mean_token_accuracy": 0.20276249945163727,
      "num_tokens": 50486793.0,
      "step": 22605
    },
    {
      "entropy": 5.303237199783325,
      "epoch": 2.539338462402426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004363454983281597,
      "loss": 5.0544,
      "mean_token_accuracy": 0.2128853678703308,
      "num_tokens": 50499202.0,
      "step": 22610
    },
    {
      "entropy": 5.479809284210205,
      "epoch": 2.5399000393103837,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004363175129276915,
      "loss": 5.1656,
      "mean_token_accuracy": 0.19808687269687653,
      "num_tokens": 50511801.0,
      "step": 22615
    },
    {
      "entropy": 5.3208949089050295,
      "epoch": 2.540461616218341,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043628952239081113,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2010200634598732,
      "num_tokens": 50522918.0,
      "step": 22620
    },
    {
      "entropy": 5.307263803482056,
      "epoch": 2.5410231931262985,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043626152671840984,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2135640189051628,
      "num_tokens": 50534027.0,
      "step": 22625
    },
    {
      "entropy": 5.321642827987671,
      "epoch": 2.541584770034256,
      "grad_norm": 1.0625,
      "learning_rate": 0.000436233525911379,
      "loss": 4.9933,
      "mean_token_accuracy": 0.21286804527044295,
      "num_tokens": 50543814.0,
      "step": 22630
    },
    {
      "entropy": 5.237502241134644,
      "epoch": 2.542146346942214,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004362055199706101,
      "loss": 5.0083,
      "mean_token_accuracy": 0.21432412564754486,
      "num_tokens": 50553818.0,
      "step": 22635
    },
    {
      "entropy": 5.362822341918945,
      "epoch": 2.5427079238501715,
      "grad_norm": 0.953125,
      "learning_rate": 0.00043617750889699496,
      "loss": 5.1644,
      "mean_token_accuracy": 0.20072179436683654,
      "num_tokens": 50564526.0,
      "step": 22640
    },
    {
      "entropy": 5.329212379455567,
      "epoch": 2.5432695007581287,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043614949269142546,
      "loss": 5.0159,
      "mean_token_accuracy": 0.2117571234703064,
      "num_tokens": 50576410.0,
      "step": 22645
    },
    {
      "entropy": 5.322167158126831,
      "epoch": 2.5438310776660864,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043612147135479356,
      "loss": 5.0235,
      "mean_token_accuracy": 0.21330979615449905,
      "num_tokens": 50586378.0,
      "step": 22650
    },
    {
      "entropy": 5.367131900787354,
      "epoch": 2.544392654574044,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004360934448879916,
      "loss": 5.0863,
      "mean_token_accuracy": 0.20704689025878906,
      "num_tokens": 50597157.0,
      "step": 22655
    },
    {
      "entropy": 5.318352699279785,
      "epoch": 2.5449542314820013,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004360654132919118,
      "loss": 5.0109,
      "mean_token_accuracy": 0.21472912281751633,
      "num_tokens": 50607390.0,
      "step": 22660
    },
    {
      "entropy": 5.255326843261718,
      "epoch": 2.545515808389959,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004360373765674467,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21899894773960113,
      "num_tokens": 50618397.0,
      "step": 22665
    },
    {
      "entropy": 5.363900089263916,
      "epoch": 2.5460773852979166,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00043600933471548927,
      "loss": 5.1899,
      "mean_token_accuracy": 0.20013949275016785,
      "num_tokens": 50631186.0,
      "step": 22670
    },
    {
      "entropy": 5.401368618011475,
      "epoch": 2.546638962205874,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004359812877369321,
      "loss": 5.0899,
      "mean_token_accuracy": 0.20725426077842712,
      "num_tokens": 50642761.0,
      "step": 22675
    },
    {
      "entropy": 5.31261510848999,
      "epoch": 2.5472005391138315,
      "grad_norm": 0.984375,
      "learning_rate": 0.00043595323563266815,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22120859622955322,
      "num_tokens": 50653374.0,
      "step": 22680
    },
    {
      "entropy": 5.273932456970215,
      "epoch": 2.547762116021789,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00043592517840359093,
      "loss": 5.0307,
      "mean_token_accuracy": 0.21363044977188111,
      "num_tokens": 50664331.0,
      "step": 22685
    },
    {
      "entropy": 5.377091979980468,
      "epoch": 2.548323692929747,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004358971160505936,
      "loss": 5.0601,
      "mean_token_accuracy": 0.20813425183296203,
      "num_tokens": 50675011.0,
      "step": 22690
    },
    {
      "entropy": 5.1913818359375,
      "epoch": 2.5488852698377045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043586904857456963,
      "loss": 4.9249,
      "mean_token_accuracy": 0.2179551303386688,
      "num_tokens": 50685139.0,
      "step": 22695
    },
    {
      "entropy": 5.328714752197266,
      "epoch": 2.5494468467456617,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043584097597641275,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21174869388341905,
      "num_tokens": 50696610.0,
      "step": 22700
    },
    {
      "entropy": 5.374369955062866,
      "epoch": 2.5500084236536193,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043581289825701683,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2128370150923729,
      "num_tokens": 50708662.0,
      "step": 22705
    },
    {
      "entropy": 5.231051731109619,
      "epoch": 2.550570000561577,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004357848154172759,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21558868736028672,
      "num_tokens": 50719457.0,
      "step": 22710
    },
    {
      "entropy": 5.307621145248413,
      "epoch": 2.5511315774695342,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043575672745808404,
      "loss": 5.026,
      "mean_token_accuracy": 0.20970945805311203,
      "num_tokens": 50730038.0,
      "step": 22715
    },
    {
      "entropy": 5.358652639389038,
      "epoch": 2.551693154377492,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043572863438033556,
      "loss": 5.058,
      "mean_token_accuracy": 0.2125213474035263,
      "num_tokens": 50742327.0,
      "step": 22720
    },
    {
      "entropy": 5.403436613082886,
      "epoch": 2.5522547312854496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004357005361849251,
      "loss": 5.1022,
      "mean_token_accuracy": 0.2077429175376892,
      "num_tokens": 50753056.0,
      "step": 22725
    },
    {
      "entropy": 5.3420013904571535,
      "epoch": 2.552816308193407,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043567243287274726,
      "loss": 5.0835,
      "mean_token_accuracy": 0.20546364039182663,
      "num_tokens": 50763961.0,
      "step": 22730
    },
    {
      "entropy": 5.324488258361816,
      "epoch": 2.553377885101365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043564432444469686,
      "loss": 4.9493,
      "mean_token_accuracy": 0.2195589378476143,
      "num_tokens": 50774978.0,
      "step": 22735
    },
    {
      "entropy": 5.325985908508301,
      "epoch": 2.553939462009322,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043561621090166885,
      "loss": 5.0807,
      "mean_token_accuracy": 0.20742491483688355,
      "num_tokens": 50786198.0,
      "step": 22740
    },
    {
      "entropy": 5.27057957649231,
      "epoch": 2.5545010389172798,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004355880922445584,
      "loss": 5.049,
      "mean_token_accuracy": 0.199573415517807,
      "num_tokens": 50797685.0,
      "step": 22745
    },
    {
      "entropy": 5.230537748336792,
      "epoch": 2.5550626158252374,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043555996847426077,
      "loss": 4.9594,
      "mean_token_accuracy": 0.220105642080307,
      "num_tokens": 50808518.0,
      "step": 22750
    },
    {
      "entropy": 5.341438722610474,
      "epoch": 2.5556241927331946,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043553183959167154,
      "loss": 5.073,
      "mean_token_accuracy": 0.2038519188761711,
      "num_tokens": 50819241.0,
      "step": 22755
    },
    {
      "entropy": 5.418996715545655,
      "epoch": 2.5561857696411523,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043550370559768614,
      "loss": 5.1263,
      "mean_token_accuracy": 0.19807977825403214,
      "num_tokens": 50830334.0,
      "step": 22760
    },
    {
      "entropy": 5.451842737197876,
      "epoch": 2.55674734654911,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004354755664932006,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20743948072195054,
      "num_tokens": 50840738.0,
      "step": 22765
    },
    {
      "entropy": 5.376601457595825,
      "epoch": 2.557308923457067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043544742227911076,
      "loss": 5.2,
      "mean_token_accuracy": 0.19879178702831268,
      "num_tokens": 50853397.0,
      "step": 22770
    },
    {
      "entropy": 5.378722238540649,
      "epoch": 2.557870500365025,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043541927295631276,
      "loss": 5.0638,
      "mean_token_accuracy": 0.20542369335889815,
      "num_tokens": 50865786.0,
      "step": 22775
    },
    {
      "entropy": 5.375649070739746,
      "epoch": 2.5584320772729825,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043539111852570296,
      "loss": 5.0794,
      "mean_token_accuracy": 0.21416058838367463,
      "num_tokens": 50876029.0,
      "step": 22780
    },
    {
      "entropy": 5.314967298507691,
      "epoch": 2.55899365418094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043536295898817767,
      "loss": 4.9858,
      "mean_token_accuracy": 0.21300610154867172,
      "num_tokens": 50886941.0,
      "step": 22785
    },
    {
      "entropy": 5.296543025970459,
      "epoch": 2.559555231088898,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004353347943446336,
      "loss": 5.0145,
      "mean_token_accuracy": 0.21018967479467393,
      "num_tokens": 50898231.0,
      "step": 22790
    },
    {
      "entropy": 5.335650825500489,
      "epoch": 2.560116807996855,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043530662459596744,
      "loss": 5.0751,
      "mean_token_accuracy": 0.20648989528417588,
      "num_tokens": 50908282.0,
      "step": 22795
    },
    {
      "entropy": 5.314947938919067,
      "epoch": 2.5606783849048127,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0004352784497430762,
      "loss": 4.9629,
      "mean_token_accuracy": 0.2164526477456093,
      "num_tokens": 50918462.0,
      "step": 22800
    },
    {
      "entropy": 5.302904033660889,
      "epoch": 2.5612399618127704,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043525026978685706,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20813806653022765,
      "num_tokens": 50929160.0,
      "step": 22805
    },
    {
      "entropy": 5.414872694015503,
      "epoch": 2.5618015387207276,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00043522208472820706,
      "loss": 5.2412,
      "mean_token_accuracy": 0.19988039135932922,
      "num_tokens": 50943361.0,
      "step": 22810
    },
    {
      "entropy": 5.432463312149048,
      "epoch": 2.5623631156286852,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004351938945680238,
      "loss": 5.1426,
      "mean_token_accuracy": 0.20509083718061447,
      "num_tokens": 50954515.0,
      "step": 22815
    },
    {
      "entropy": 5.446000242233277,
      "epoch": 2.562924692536643,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004351656993072048,
      "loss": 5.1329,
      "mean_token_accuracy": 0.2009390488266945,
      "num_tokens": 50965237.0,
      "step": 22820
    },
    {
      "entropy": 5.293050861358642,
      "epoch": 2.5634862694446006,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004351374989466477,
      "loss": 4.9974,
      "mean_token_accuracy": 0.21018233597278596,
      "num_tokens": 50976406.0,
      "step": 22825
    },
    {
      "entropy": 5.202159690856933,
      "epoch": 2.5640478463525582,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043510929348725065,
      "loss": 4.952,
      "mean_token_accuracy": 0.2074650138616562,
      "num_tokens": 50987214.0,
      "step": 22830
    },
    {
      "entropy": 5.343889474868774,
      "epoch": 2.5646094232605154,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043508108292991156,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21359646022319795,
      "num_tokens": 50997668.0,
      "step": 22835
    },
    {
      "entropy": 5.371627330780029,
      "epoch": 2.565171000168473,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004350528672755287,
      "loss": 5.0222,
      "mean_token_accuracy": 0.213147833943367,
      "num_tokens": 51008358.0,
      "step": 22840
    },
    {
      "entropy": 5.301675033569336,
      "epoch": 2.5657325770764308,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004350246465250005,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21169646382331847,
      "num_tokens": 51019410.0,
      "step": 22845
    },
    {
      "entropy": 5.417147970199585,
      "epoch": 2.566294153984388,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00043499642067922544,
      "loss": 5.1962,
      "mean_token_accuracy": 0.20229731649160385,
      "num_tokens": 51032307.0,
      "step": 22850
    },
    {
      "entropy": 5.459134912490844,
      "epoch": 2.5668557308923456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043496818973910235,
      "loss": 5.0711,
      "mean_token_accuracy": 0.20236650556325914,
      "num_tokens": 51043176.0,
      "step": 22855
    },
    {
      "entropy": 5.368083763122558,
      "epoch": 2.5674173078003033,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043493995370553,
      "loss": 5.0889,
      "mean_token_accuracy": 0.20394199192523957,
      "num_tokens": 51053977.0,
      "step": 22860
    },
    {
      "entropy": 5.317160940170288,
      "epoch": 2.5679788847082605,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043491171257940755,
      "loss": 5.0649,
      "mean_token_accuracy": 0.21018093675374985,
      "num_tokens": 51064801.0,
      "step": 22865
    },
    {
      "entropy": 5.389581871032715,
      "epoch": 2.568540461616218,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043488346636163405,
      "loss": 5.0273,
      "mean_token_accuracy": 0.21663236916065215,
      "num_tokens": 51075400.0,
      "step": 22870
    },
    {
      "entropy": 5.371446228027343,
      "epoch": 2.569102038524176,
      "grad_norm": 1.0,
      "learning_rate": 0.0004348552150531091,
      "loss": 5.1668,
      "mean_token_accuracy": 0.20104426592588426,
      "num_tokens": 51087077.0,
      "step": 22875
    },
    {
      "entropy": 5.2992103099823,
      "epoch": 2.5696636154321335,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043482695865473207,
      "loss": 5.0536,
      "mean_token_accuracy": 0.2045610472559929,
      "num_tokens": 51099798.0,
      "step": 22880
    },
    {
      "entropy": 5.444314908981323,
      "epoch": 2.570225192340091,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004347986971674026,
      "loss": 5.1832,
      "mean_token_accuracy": 0.19880043119192123,
      "num_tokens": 51111880.0,
      "step": 22885
    },
    {
      "entropy": 5.361950635910034,
      "epoch": 2.5707867692480484,
      "grad_norm": 1.0,
      "learning_rate": 0.00043477043059202077,
      "loss": 5.0586,
      "mean_token_accuracy": 0.20542476028203965,
      "num_tokens": 51122358.0,
      "step": 22890
    },
    {
      "entropy": 5.342841625213623,
      "epoch": 2.571348346156006,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004347421589294864,
      "loss": 5.0807,
      "mean_token_accuracy": 0.2151031196117401,
      "num_tokens": 51132775.0,
      "step": 22895
    },
    {
      "entropy": 5.34867901802063,
      "epoch": 2.5719099230639637,
      "grad_norm": 1.0,
      "learning_rate": 0.0004347138821806997,
      "loss": 5.1407,
      "mean_token_accuracy": 0.20563014149665831,
      "num_tokens": 51144974.0,
      "step": 22900
    },
    {
      "entropy": 5.3553080558776855,
      "epoch": 2.572471499971921,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000434685600346561,
      "loss": 4.9677,
      "mean_token_accuracy": 0.2211044669151306,
      "num_tokens": 51154700.0,
      "step": 22905
    },
    {
      "entropy": 5.368930816650391,
      "epoch": 2.5730330768798786,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000434657313427971,
      "loss": 5.0974,
      "mean_token_accuracy": 0.20454032272100447,
      "num_tokens": 51166910.0,
      "step": 22910
    },
    {
      "entropy": 5.293558597564697,
      "epoch": 2.5735946537878363,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004346290214258301,
      "loss": 4.9829,
      "mean_token_accuracy": 0.2078662022948265,
      "num_tokens": 51177620.0,
      "step": 22915
    },
    {
      "entropy": 5.3800530433654785,
      "epoch": 2.574156230695794,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043460072434103934,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21333712488412857,
      "num_tokens": 51187894.0,
      "step": 22920
    },
    {
      "entropy": 5.304312372207642,
      "epoch": 2.5747178076037516,
      "grad_norm": 1.0,
      "learning_rate": 0.0004345724221744996,
      "loss": 5.1028,
      "mean_token_accuracy": 0.20223884731531144,
      "num_tokens": 51199625.0,
      "step": 22925
    },
    {
      "entropy": 5.2669758796691895,
      "epoch": 2.575279384511709,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043454411492711197,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21084279865026473,
      "num_tokens": 51211932.0,
      "step": 22930
    },
    {
      "entropy": 5.374780559539795,
      "epoch": 2.5758409614196665,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043451580259977793,
      "loss": 5.114,
      "mean_token_accuracy": 0.20703873336315154,
      "num_tokens": 51223196.0,
      "step": 22935
    },
    {
      "entropy": 5.3962812423706055,
      "epoch": 2.576402538327624,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043448748519339875,
      "loss": 5.0602,
      "mean_token_accuracy": 0.20327444076538087,
      "num_tokens": 51234246.0,
      "step": 22940
    },
    {
      "entropy": 5.346772956848144,
      "epoch": 2.5769641152355813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043445916270887626,
      "loss": 5.018,
      "mean_token_accuracy": 0.21081008166074752,
      "num_tokens": 51245056.0,
      "step": 22945
    },
    {
      "entropy": 5.341243505477905,
      "epoch": 2.577525692143539,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004344308351471123,
      "loss": 5.0369,
      "mean_token_accuracy": 0.20924060493707658,
      "num_tokens": 51256513.0,
      "step": 22950
    },
    {
      "entropy": 5.265821409225464,
      "epoch": 2.5780872690514967,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043440250250900854,
      "loss": 5.0415,
      "mean_token_accuracy": 0.20694104880094527,
      "num_tokens": 51267609.0,
      "step": 22955
    },
    {
      "entropy": 5.4054553508758545,
      "epoch": 2.578648845959454,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004343741647954673,
      "loss": 5.161,
      "mean_token_accuracy": 0.20443835556507112,
      "num_tokens": 51277940.0,
      "step": 22960
    },
    {
      "entropy": 5.511870431900024,
      "epoch": 2.5792104228674115,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004343458220073909,
      "loss": 5.1615,
      "mean_token_accuracy": 0.20982973724603654,
      "num_tokens": 51288853.0,
      "step": 22965
    },
    {
      "entropy": 5.35790319442749,
      "epoch": 2.579771999775369,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004343174741456817,
      "loss": 5.1657,
      "mean_token_accuracy": 0.1988702669739723,
      "num_tokens": 51299197.0,
      "step": 22970
    },
    {
      "entropy": 5.376319313049317,
      "epoch": 2.580333576683327,
      "grad_norm": 1.140625,
      "learning_rate": 0.0004342891212112423,
      "loss": 5.1712,
      "mean_token_accuracy": 0.19607665836811067,
      "num_tokens": 51310821.0,
      "step": 22975
    },
    {
      "entropy": 5.491288852691651,
      "epoch": 2.5808951535912845,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00043426076320497557,
      "loss": 5.0862,
      "mean_token_accuracy": 0.19908646494150162,
      "num_tokens": 51321805.0,
      "step": 22980
    },
    {
      "entropy": 5.260193586349487,
      "epoch": 2.5814567304992417,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043423240012778437,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21433060765266418,
      "num_tokens": 51332647.0,
      "step": 22985
    },
    {
      "entropy": 5.257340288162231,
      "epoch": 2.5820183074071994,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004342040319805717,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2167609840631485,
      "num_tokens": 51344715.0,
      "step": 22990
    },
    {
      "entropy": 5.39791841506958,
      "epoch": 2.582579884315157,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000434175658764241,
      "loss": 5.1513,
      "mean_token_accuracy": 0.2072545751929283,
      "num_tokens": 51355481.0,
      "step": 22995
    },
    {
      "entropy": 5.334388875961304,
      "epoch": 2.5831414612231143,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043414728047969553,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21886058449745177,
      "num_tokens": 51367098.0,
      "step": 23000
    },
    {
      "entropy": 5.383289194107055,
      "epoch": 2.583703038131072,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043411889712783897,
      "loss": 5.1764,
      "mean_token_accuracy": 0.201360122859478,
      "num_tokens": 51379948.0,
      "step": 23005
    },
    {
      "entropy": 5.315848016738892,
      "epoch": 2.5842646150390296,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043409050870957505,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21220746487379075,
      "num_tokens": 51391675.0,
      "step": 23010
    },
    {
      "entropy": 5.305407476425171,
      "epoch": 2.5848261919469873,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004340621152258075,
      "loss": 4.9528,
      "mean_token_accuracy": 0.2144648849964142,
      "num_tokens": 51401809.0,
      "step": 23015
    },
    {
      "entropy": 5.3098883628845215,
      "epoch": 2.585387768854945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043403371667744055,
      "loss": 5.0378,
      "mean_token_accuracy": 0.2089108482003212,
      "num_tokens": 51413184.0,
      "step": 23020
    },
    {
      "entropy": 5.331208848953247,
      "epoch": 2.585949345762902,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004340053130653785,
      "loss": 5.0914,
      "mean_token_accuracy": 0.2029418557882309,
      "num_tokens": 51424751.0,
      "step": 23025
    },
    {
      "entropy": 5.294845867156982,
      "epoch": 2.58651092267086,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004339769043905255,
      "loss": 5.023,
      "mean_token_accuracy": 0.20772373527288437,
      "num_tokens": 51436550.0,
      "step": 23030
    },
    {
      "entropy": 5.379385185241699,
      "epoch": 2.5870724995788175,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043394849065378624,
      "loss": 5.047,
      "mean_token_accuracy": 0.20946955382823945,
      "num_tokens": 51447691.0,
      "step": 23035
    },
    {
      "entropy": 5.298296642303467,
      "epoch": 2.5876340764867747,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043392007185606533,
      "loss": 5.0365,
      "mean_token_accuracy": 0.21058009713888168,
      "num_tokens": 51459715.0,
      "step": 23040
    },
    {
      "entropy": 5.308298587799072,
      "epoch": 2.5881956533947323,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004338916479982678,
      "loss": 4.9795,
      "mean_token_accuracy": 0.2164639487862587,
      "num_tokens": 51470541.0,
      "step": 23045
    },
    {
      "entropy": 5.332243347167969,
      "epoch": 2.58875723030269,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043386321908129856,
      "loss": 4.9711,
      "mean_token_accuracy": 0.21471761912107468,
      "num_tokens": 51480724.0,
      "step": 23050
    },
    {
      "entropy": 5.43809700012207,
      "epoch": 2.589318807210647,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00043383478510606276,
      "loss": 5.2342,
      "mean_token_accuracy": 0.20005029886960984,
      "num_tokens": 51492204.0,
      "step": 23055
    },
    {
      "entropy": 5.363996791839599,
      "epoch": 2.589880384118605,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004338063460734658,
      "loss": 5.0769,
      "mean_token_accuracy": 0.20128879249095916,
      "num_tokens": 51503354.0,
      "step": 23060
    },
    {
      "entropy": 5.361293125152588,
      "epoch": 2.5904419610265625,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004337779019844132,
      "loss": 5.0845,
      "mean_token_accuracy": 0.21567115187644958,
      "num_tokens": 51513563.0,
      "step": 23065
    },
    {
      "entropy": 5.36114912033081,
      "epoch": 2.59100353793452,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043374945283981066,
      "loss": 5.0542,
      "mean_token_accuracy": 0.20668838620185853,
      "num_tokens": 51524996.0,
      "step": 23070
    },
    {
      "entropy": 5.248710775375367,
      "epoch": 2.591565114842478,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004337209986405639,
      "loss": 4.9409,
      "mean_token_accuracy": 0.21769076734781265,
      "num_tokens": 51536822.0,
      "step": 23075
    },
    {
      "entropy": 5.26055212020874,
      "epoch": 2.592126691750435,
      "grad_norm": 1.125,
      "learning_rate": 0.000433692539387579,
      "loss": 5.0049,
      "mean_token_accuracy": 0.21578723937273026,
      "num_tokens": 51547545.0,
      "step": 23080
    },
    {
      "entropy": 5.341302156448364,
      "epoch": 2.5926882686583927,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004336640750817621,
      "loss": 5.089,
      "mean_token_accuracy": 0.2108192190527916,
      "num_tokens": 51560721.0,
      "step": 23085
    },
    {
      "entropy": 5.306601333618164,
      "epoch": 2.5932498455663504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043363560572401953,
      "loss": 4.9359,
      "mean_token_accuracy": 0.2124685376882553,
      "num_tokens": 51573226.0,
      "step": 23090
    },
    {
      "entropy": 5.321824789047241,
      "epoch": 2.5938114224743076,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043360713131525774,
      "loss": 5.0641,
      "mean_token_accuracy": 0.20712223052978515,
      "num_tokens": 51584338.0,
      "step": 23095
    },
    {
      "entropy": 5.367713165283203,
      "epoch": 2.5943729993822653,
      "grad_norm": 1.2578125,
      "learning_rate": 0.0004335786518563834,
      "loss": 5.0473,
      "mean_token_accuracy": 0.20937429666519164,
      "num_tokens": 51593979.0,
      "step": 23100
    },
    {
      "entropy": 5.296347284317017,
      "epoch": 2.594934576290223,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004335501673483033,
      "loss": 5.0142,
      "mean_token_accuracy": 0.21255873441696166,
      "num_tokens": 51604713.0,
      "step": 23105
    },
    {
      "entropy": 5.339443731307983,
      "epoch": 2.5954961531981806,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043352167779192423,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20973604619503022,
      "num_tokens": 51616034.0,
      "step": 23110
    },
    {
      "entropy": 5.229994106292724,
      "epoch": 2.5960577301061383,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043349318318815363,
      "loss": 4.956,
      "mean_token_accuracy": 0.20722443759441375,
      "num_tokens": 51627551.0,
      "step": 23115
    },
    {
      "entropy": 5.373639917373657,
      "epoch": 2.5966193070140955,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043346468353789853,
      "loss": 5.0418,
      "mean_token_accuracy": 0.2119785100221634,
      "num_tokens": 51639625.0,
      "step": 23120
    },
    {
      "entropy": 5.3134795188903805,
      "epoch": 2.597180883922053,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004334361788420664,
      "loss": 4.9774,
      "mean_token_accuracy": 0.21390149742364883,
      "num_tokens": 51649918.0,
      "step": 23125
    },
    {
      "entropy": 5.234883451461792,
      "epoch": 2.597742460830011,
      "grad_norm": 1.0,
      "learning_rate": 0.00043340766910156497,
      "loss": 5.0082,
      "mean_token_accuracy": 0.22567882835865022,
      "num_tokens": 51659527.0,
      "step": 23130
    },
    {
      "entropy": 5.287597894668579,
      "epoch": 2.598304037737968,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043337915431730195,
      "loss": 5.1289,
      "mean_token_accuracy": 0.2006644368171692,
      "num_tokens": 51670704.0,
      "step": 23135
    },
    {
      "entropy": 5.326800012588501,
      "epoch": 2.5988656146459257,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043335063449018515,
      "loss": 5.0292,
      "mean_token_accuracy": 0.21450041979551315,
      "num_tokens": 51681606.0,
      "step": 23140
    },
    {
      "entropy": 5.3945794105529785,
      "epoch": 2.5994271915538834,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004333221096211227,
      "loss": 5.115,
      "mean_token_accuracy": 0.20639398992061614,
      "num_tokens": 51692592.0,
      "step": 23145
    },
    {
      "entropy": 5.318304538726807,
      "epoch": 2.599988768461841,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043329357971102296,
      "loss": 5.048,
      "mean_token_accuracy": 0.21101107448339462,
      "num_tokens": 51704990.0,
      "step": 23150
    },
    {
      "entropy": 5.353518295288086,
      "epoch": 2.6005503453697982,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043326504476079425,
      "loss": 5.051,
      "mean_token_accuracy": 0.2111370638012886,
      "num_tokens": 51715934.0,
      "step": 23155
    },
    {
      "entropy": 5.334613513946533,
      "epoch": 2.601111922277756,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004332365047713451,
      "loss": 5.1541,
      "mean_token_accuracy": 0.20242172479629517,
      "num_tokens": 51727314.0,
      "step": 23160
    },
    {
      "entropy": 5.317201948165893,
      "epoch": 2.6016734991857136,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004332079597435842,
      "loss": 4.9628,
      "mean_token_accuracy": 0.21933379620313645,
      "num_tokens": 51738322.0,
      "step": 23165
    },
    {
      "entropy": 5.259560775756836,
      "epoch": 2.602235076093671,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004331794096784206,
      "loss": 4.9745,
      "mean_token_accuracy": 0.21421266347169876,
      "num_tokens": 51749221.0,
      "step": 23170
    },
    {
      "entropy": 5.271308946609497,
      "epoch": 2.6027966530016284,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004331508545767633,
      "loss": 5.0221,
      "mean_token_accuracy": 0.20627715736627578,
      "num_tokens": 51760568.0,
      "step": 23175
    },
    {
      "entropy": 5.345342493057251,
      "epoch": 2.603358229909586,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004331222944395213,
      "loss": 5.0753,
      "mean_token_accuracy": 0.21563337445259095,
      "num_tokens": 51772652.0,
      "step": 23180
    },
    {
      "entropy": 5.357407855987549,
      "epoch": 2.6039198068175438,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004330937292676042,
      "loss": 5.0567,
      "mean_token_accuracy": 0.21271584928035736,
      "num_tokens": 51783739.0,
      "step": 23185
    },
    {
      "entropy": 5.312764263153076,
      "epoch": 2.604481383725501,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004330651590619214,
      "loss": 5.0587,
      "mean_token_accuracy": 0.20981205403804778,
      "num_tokens": 51795521.0,
      "step": 23190
    },
    {
      "entropy": 5.258847284317016,
      "epoch": 2.6050429606334586,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004330365838233827,
      "loss": 4.8819,
      "mean_token_accuracy": 0.21689916253089905,
      "num_tokens": 51805041.0,
      "step": 23195
    },
    {
      "entropy": 5.378132915496826,
      "epoch": 2.6056045375414163,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043300800355289786,
      "loss": 5.1121,
      "mean_token_accuracy": 0.20650954246520997,
      "num_tokens": 51815157.0,
      "step": 23200
    },
    {
      "entropy": 5.413678121566773,
      "epoch": 2.606166114449374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043297941825137684,
      "loss": 5.1319,
      "mean_token_accuracy": 0.20830687433481215,
      "num_tokens": 51826760.0,
      "step": 23205
    },
    {
      "entropy": 5.4535524368286135,
      "epoch": 2.6067276913573316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043295082791972993,
      "loss": 5.2406,
      "mean_token_accuracy": 0.20192149877548218,
      "num_tokens": 51838018.0,
      "step": 23210
    },
    {
      "entropy": 5.41840090751648,
      "epoch": 2.607289268265289,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043292223255886737,
      "loss": 5.117,
      "mean_token_accuracy": 0.20307396501302719,
      "num_tokens": 51848217.0,
      "step": 23215
    },
    {
      "entropy": 5.31718487739563,
      "epoch": 2.6078508451732465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043289363216969973,
      "loss": 4.992,
      "mean_token_accuracy": 0.21767829209566117,
      "num_tokens": 51857537.0,
      "step": 23220
    },
    {
      "entropy": 5.308617639541626,
      "epoch": 2.608412422081204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043286502675313757,
      "loss": 5.0314,
      "mean_token_accuracy": 0.21636929214000702,
      "num_tokens": 51867961.0,
      "step": 23225
    },
    {
      "entropy": 5.332007217407226,
      "epoch": 2.6089739989891614,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004328364163100917,
      "loss": 5.1382,
      "mean_token_accuracy": 0.19954214990139008,
      "num_tokens": 51879868.0,
      "step": 23230
    },
    {
      "entropy": 5.373329210281372,
      "epoch": 2.609535575897119,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004328078008414731,
      "loss": 5.0787,
      "mean_token_accuracy": 0.19881752729415894,
      "num_tokens": 51890849.0,
      "step": 23235
    },
    {
      "entropy": 5.375471544265747,
      "epoch": 2.6100971528050767,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000432779180348193,
      "loss": 5.0421,
      "mean_token_accuracy": 0.21202234774827958,
      "num_tokens": 51901193.0,
      "step": 23240
    },
    {
      "entropy": 5.362505626678467,
      "epoch": 2.6106587297130344,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043275055483116254,
      "loss": 5.0441,
      "mean_token_accuracy": 0.21033971905708312,
      "num_tokens": 51912551.0,
      "step": 23245
    },
    {
      "entropy": 5.460479259490967,
      "epoch": 2.6112203066209916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004327219242912933,
      "loss": 5.2428,
      "mean_token_accuracy": 0.19244807213544846,
      "num_tokens": 51923970.0,
      "step": 23250
    },
    {
      "entropy": 5.246540307998657,
      "epoch": 2.6117818835289492,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004326932887294967,
      "loss": 4.9353,
      "mean_token_accuracy": 0.21528396457433702,
      "num_tokens": 51936667.0,
      "step": 23255
    },
    {
      "entropy": 5.28762469291687,
      "epoch": 2.612343460436907,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004326646481466848,
      "loss": 4.9513,
      "mean_token_accuracy": 0.21467649787664414,
      "num_tokens": 51947479.0,
      "step": 23260
    },
    {
      "entropy": 5.284732627868652,
      "epoch": 2.6129050373448646,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004326360025437692,
      "loss": 5.1002,
      "mean_token_accuracy": 0.2164476439356804,
      "num_tokens": 51958669.0,
      "step": 23265
    },
    {
      "entropy": 5.249891233444214,
      "epoch": 2.613466614252822,
      "grad_norm": 1.0,
      "learning_rate": 0.00043260735192166217,
      "loss": 4.9468,
      "mean_token_accuracy": 0.22005217671394348,
      "num_tokens": 51969273.0,
      "step": 23270
    },
    {
      "entropy": 5.368363571166992,
      "epoch": 2.6140281911607794,
      "grad_norm": 0.91796875,
      "learning_rate": 0.000432578696281276,
      "loss": 5.1237,
      "mean_token_accuracy": 0.20743368119001387,
      "num_tokens": 51981220.0,
      "step": 23275
    },
    {
      "entropy": 5.337776374816895,
      "epoch": 2.614589768068737,
      "grad_norm": 1.09375,
      "learning_rate": 0.000432550035623523,
      "loss": 5.1089,
      "mean_token_accuracy": 0.20474644005298615,
      "num_tokens": 51992768.0,
      "step": 23280
    },
    {
      "entropy": 5.292093133926391,
      "epoch": 2.6151513449766943,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004325213699493157,
      "loss": 5.0058,
      "mean_token_accuracy": 0.20383877009153367,
      "num_tokens": 52002754.0,
      "step": 23285
    },
    {
      "entropy": 5.370396709442138,
      "epoch": 2.615712921884652,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004324926992595669,
      "loss": 5.0097,
      "mean_token_accuracy": 0.20936553329229354,
      "num_tokens": 52014503.0,
      "step": 23290
    },
    {
      "entropy": 5.363729000091553,
      "epoch": 2.6162744987926096,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0004324640235551895,
      "loss": 5.0842,
      "mean_token_accuracy": 0.20791612416505814,
      "num_tokens": 52026151.0,
      "step": 23295
    },
    {
      "entropy": 5.2663969039917,
      "epoch": 2.6168360757005673,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004324353428370965,
      "loss": 5.043,
      "mean_token_accuracy": 0.2031331777572632,
      "num_tokens": 52037704.0,
      "step": 23300
    },
    {
      "entropy": 5.333244323730469,
      "epoch": 2.617397652608525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004324066571062012,
      "loss": 5.0727,
      "mean_token_accuracy": 0.20923311561346053,
      "num_tokens": 52049326.0,
      "step": 23305
    },
    {
      "entropy": 5.4121815204620365,
      "epoch": 2.617959229516482,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004323779663634168,
      "loss": 5.0484,
      "mean_token_accuracy": 0.21584014147520064,
      "num_tokens": 52060030.0,
      "step": 23310
    },
    {
      "entropy": 5.403695487976075,
      "epoch": 2.61852080642444,
      "grad_norm": 1.0625,
      "learning_rate": 0.000432349270609657,
      "loss": 5.207,
      "mean_token_accuracy": 0.20441460162401198,
      "num_tokens": 52071744.0,
      "step": 23315
    },
    {
      "entropy": 5.395181512832641,
      "epoch": 2.6190823833323975,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043232056984583527,
      "loss": 5.0386,
      "mean_token_accuracy": 0.20882814079523088,
      "num_tokens": 52082443.0,
      "step": 23320
    },
    {
      "entropy": 5.318614339828491,
      "epoch": 2.6196439602403547,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004322918640728656,
      "loss": 5.0219,
      "mean_token_accuracy": 0.21076486259698868,
      "num_tokens": 52094211.0,
      "step": 23325
    },
    {
      "entropy": 5.262171983718872,
      "epoch": 2.6202055371483124,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043226315329166203,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2160695716738701,
      "num_tokens": 52104802.0,
      "step": 23330
    },
    {
      "entropy": 5.411321973800659,
      "epoch": 2.62076711405627,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043223443750313873,
      "loss": 5.0886,
      "mean_token_accuracy": 0.20482709705829621,
      "num_tokens": 52115459.0,
      "step": 23335
    },
    {
      "entropy": 5.352807283401489,
      "epoch": 2.6213286909642277,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004322057167082099,
      "loss": 5.1118,
      "mean_token_accuracy": 0.20402692258358002,
      "num_tokens": 52125572.0,
      "step": 23340
    },
    {
      "entropy": 5.298005485534668,
      "epoch": 2.621890267872185,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00043217699090779006,
      "loss": 5.0365,
      "mean_token_accuracy": 0.20902466177940368,
      "num_tokens": 52137405.0,
      "step": 23345
    },
    {
      "entropy": 5.381075525283814,
      "epoch": 2.6224518447801426,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043214826010279383,
      "loss": 5.1307,
      "mean_token_accuracy": 0.2000174567103386,
      "num_tokens": 52148131.0,
      "step": 23350
    },
    {
      "entropy": 5.360813474655151,
      "epoch": 2.6230134216881003,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004321195242941361,
      "loss": 5.0487,
      "mean_token_accuracy": 0.20936367213726043,
      "num_tokens": 52159361.0,
      "step": 23355
    },
    {
      "entropy": 5.2887636661529545,
      "epoch": 2.623574998596058,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043209078348273175,
      "loss": 5.0342,
      "mean_token_accuracy": 0.20622579753398895,
      "num_tokens": 52170521.0,
      "step": 23360
    },
    {
      "entropy": 5.429940986633301,
      "epoch": 2.624136575504015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004320620376694959,
      "loss": 5.1405,
      "mean_token_accuracy": 0.2106136530637741,
      "num_tokens": 52182069.0,
      "step": 23365
    },
    {
      "entropy": 5.28065595626831,
      "epoch": 2.624698152411973,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00043203328685534396,
      "loss": 4.951,
      "mean_token_accuracy": 0.21966911852359772,
      "num_tokens": 52191867.0,
      "step": 23370
    },
    {
      "entropy": 5.396372556686401,
      "epoch": 2.6252597293199305,
      "grad_norm": 1.0,
      "learning_rate": 0.0004320045310411912,
      "loss": 5.2129,
      "mean_token_accuracy": 0.1931976094841957,
      "num_tokens": 52204511.0,
      "step": 23375
    },
    {
      "entropy": 5.383773279190064,
      "epoch": 2.6258213062278877,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043197577022795327,
      "loss": 4.9909,
      "mean_token_accuracy": 0.2126855283975601,
      "num_tokens": 52214276.0,
      "step": 23380
    },
    {
      "entropy": 5.3106895923614506,
      "epoch": 2.6263828831358453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004319470044165459,
      "loss": 5.037,
      "mean_token_accuracy": 0.21392443627119065,
      "num_tokens": 52225388.0,
      "step": 23385
    },
    {
      "entropy": 5.315458059310913,
      "epoch": 2.626944460043803,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004319182336078852,
      "loss": 4.9501,
      "mean_token_accuracy": 0.2162473037838936,
      "num_tokens": 52235300.0,
      "step": 23390
    },
    {
      "entropy": 5.395591163635254,
      "epoch": 2.6275060369517607,
      "grad_norm": 1.03125,
      "learning_rate": 0.000431889457802887,
      "loss": 5.2002,
      "mean_token_accuracy": 0.19759428799152373,
      "num_tokens": 52248392.0,
      "step": 23395
    },
    {
      "entropy": 5.35674638748169,
      "epoch": 2.6280676138597183,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004318606770024675,
      "loss": 5.0238,
      "mean_token_accuracy": 0.20323655605316163,
      "num_tokens": 52259839.0,
      "step": 23400
    },
    {
      "entropy": 5.364160490036011,
      "epoch": 2.6286291907676755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004318318912075433,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20067032277584076,
      "num_tokens": 52271195.0,
      "step": 23405
    },
    {
      "entropy": 5.3357116222381595,
      "epoch": 2.629190767675633,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004318031004190308,
      "loss": 5.0201,
      "mean_token_accuracy": 0.2141626477241516,
      "num_tokens": 52282445.0,
      "step": 23410
    },
    {
      "entropy": 5.402815866470337,
      "epoch": 2.629752344583591,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00043177430463784685,
      "loss": 5.0537,
      "mean_token_accuracy": 0.20307561606168748,
      "num_tokens": 52294144.0,
      "step": 23415
    },
    {
      "entropy": 5.279774713516235,
      "epoch": 2.630313921491548,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043174550386490816,
      "loss": 5.0575,
      "mean_token_accuracy": 0.21180805563926697,
      "num_tokens": 52304510.0,
      "step": 23420
    },
    {
      "entropy": 5.328264427185059,
      "epoch": 2.6308754983995057,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004317166981011319,
      "loss": 4.9816,
      "mean_token_accuracy": 0.21206202507019042,
      "num_tokens": 52314507.0,
      "step": 23425
    },
    {
      "entropy": 5.2640355110168455,
      "epoch": 2.6314370753074634,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043168788734743516,
      "loss": 5.0084,
      "mean_token_accuracy": 0.21649845242500304,
      "num_tokens": 52325104.0,
      "step": 23430
    },
    {
      "entropy": 5.409167814254761,
      "epoch": 2.631998652215421,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043165907160473525,
      "loss": 5.1047,
      "mean_token_accuracy": 0.19847842305898666,
      "num_tokens": 52336037.0,
      "step": 23435
    },
    {
      "entropy": 5.335485744476318,
      "epoch": 2.6325602291233783,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004316302508739499,
      "loss": 4.9614,
      "mean_token_accuracy": 0.2141459584236145,
      "num_tokens": 52346901.0,
      "step": 23440
    },
    {
      "entropy": 5.313931655883789,
      "epoch": 2.633121806031336,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004316014251559965,
      "loss": 5.0531,
      "mean_token_accuracy": 0.21456570774316788,
      "num_tokens": 52359159.0,
      "step": 23445
    },
    {
      "entropy": 5.271452617645264,
      "epoch": 2.6336833829392936,
      "grad_norm": 1.125,
      "learning_rate": 0.00043157259445179304,
      "loss": 4.94,
      "mean_token_accuracy": 0.2165892615914345,
      "num_tokens": 52369274.0,
      "step": 23450
    },
    {
      "entropy": 5.235901880264282,
      "epoch": 2.6342449598472513,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004315437587622574,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21726230084896087,
      "num_tokens": 52381360.0,
      "step": 23455
    },
    {
      "entropy": 5.320491123199463,
      "epoch": 2.6348065367552085,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00043151491808830777,
      "loss": 5.06,
      "mean_token_accuracy": 0.20404604375362395,
      "num_tokens": 52391550.0,
      "step": 23460
    },
    {
      "entropy": 5.322268056869507,
      "epoch": 2.635368113663166,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043148607243086255,
      "loss": 4.991,
      "mean_token_accuracy": 0.21521255373954773,
      "num_tokens": 52401339.0,
      "step": 23465
    },
    {
      "entropy": 5.35680832862854,
      "epoch": 2.635929690571124,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004314572217908399,
      "loss": 5.0686,
      "mean_token_accuracy": 0.20575239956378938,
      "num_tokens": 52412062.0,
      "step": 23470
    },
    {
      "entropy": 5.3641399383544925,
      "epoch": 2.636491267479081,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004314283661691587,
      "loss": 5.0641,
      "mean_token_accuracy": 0.2086193561553955,
      "num_tokens": 52422524.0,
      "step": 23475
    },
    {
      "entropy": 5.422532653808593,
      "epoch": 2.6370528443870387,
      "grad_norm": 1.328125,
      "learning_rate": 0.0004313995055667377,
      "loss": 5.1417,
      "mean_token_accuracy": 0.20664166510105134,
      "num_tokens": 52433422.0,
      "step": 23480
    },
    {
      "entropy": 5.397749996185302,
      "epoch": 2.6376144212949963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004313706399844958,
      "loss": 5.1247,
      "mean_token_accuracy": 0.2059051811695099,
      "num_tokens": 52443879.0,
      "step": 23485
    },
    {
      "entropy": 5.3578352451324465,
      "epoch": 2.638175998202954,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000431341769423352,
      "loss": 5.045,
      "mean_token_accuracy": 0.20758167058229446,
      "num_tokens": 52455610.0,
      "step": 23490
    },
    {
      "entropy": 5.4327699661254885,
      "epoch": 2.6387375751109117,
      "grad_norm": 1.0,
      "learning_rate": 0.0004313128938842257,
      "loss": 5.1257,
      "mean_token_accuracy": 0.20834770649671555,
      "num_tokens": 52467392.0,
      "step": 23495
    },
    {
      "entropy": 5.370547199249268,
      "epoch": 2.639299152018869,
      "grad_norm": 1.2109375,
      "learning_rate": 0.00043128401336803616,
      "loss": 4.9787,
      "mean_token_accuracy": 0.20553670227527618,
      "num_tokens": 52478812.0,
      "step": 23500
    },
    {
      "entropy": 5.355093574523925,
      "epoch": 2.6398607289268265,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043125512787570305,
      "loss": 5.087,
      "mean_token_accuracy": 0.21375200748443604,
      "num_tokens": 52488837.0,
      "step": 23505
    },
    {
      "entropy": 5.356348943710327,
      "epoch": 2.640422305834784,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004312262374081461,
      "loss": 5.1359,
      "mean_token_accuracy": 0.2005271703004837,
      "num_tokens": 52501656.0,
      "step": 23510
    },
    {
      "entropy": 5.301506996154785,
      "epoch": 2.6409838827427414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00043119734196628507,
      "loss": 5.009,
      "mean_token_accuracy": 0.2089068189263344,
      "num_tokens": 52512280.0,
      "step": 23515
    },
    {
      "entropy": 5.299967050552368,
      "epoch": 2.641545459650699,
      "grad_norm": 1.078125,
      "learning_rate": 0.00043116844155104025,
      "loss": 5.0661,
      "mean_token_accuracy": 0.21407205611467361,
      "num_tokens": 52523664.0,
      "step": 23520
    },
    {
      "entropy": 5.343350744247436,
      "epoch": 2.6421070365586568,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004311395361633316,
      "loss": 5.0927,
      "mean_token_accuracy": 0.20679408311843872,
      "num_tokens": 52534913.0,
      "step": 23525
    },
    {
      "entropy": 5.285246753692627,
      "epoch": 2.6426686134666144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004311106258040795,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21468174010515212,
      "num_tokens": 52545576.0,
      "step": 23530
    },
    {
      "entropy": 5.4022541522979735,
      "epoch": 2.643230190374572,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043108171047420463,
      "loss": 5.0733,
      "mean_token_accuracy": 0.20498761981725694,
      "num_tokens": 52556136.0,
      "step": 23535
    },
    {
      "entropy": 5.352645826339722,
      "epoch": 2.6437917672825293,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004310527901746275,
      "loss": 5.1158,
      "mean_token_accuracy": 0.20155650824308396,
      "num_tokens": 52567425.0,
      "step": 23540
    },
    {
      "entropy": 5.303713607788086,
      "epoch": 2.644353344190487,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000431023864906269,
      "loss": 5.0136,
      "mean_token_accuracy": 0.2118804469704628,
      "num_tokens": 52578040.0,
      "step": 23545
    },
    {
      "entropy": 5.298139858245849,
      "epoch": 2.6449149210984446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00043099493467005015,
      "loss": 4.9352,
      "mean_token_accuracy": 0.2156015396118164,
      "num_tokens": 52588261.0,
      "step": 23550
    },
    {
      "entropy": 5.261673831939698,
      "epoch": 2.645476498006402,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000430965999466892,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2112935096025467,
      "num_tokens": 52599522.0,
      "step": 23555
    },
    {
      "entropy": 5.370843601226807,
      "epoch": 2.6460380749143595,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00043093705929771604,
      "loss": 5.1799,
      "mean_token_accuracy": 0.1982271373271942,
      "num_tokens": 52610244.0,
      "step": 23560
    },
    {
      "entropy": 5.472581672668457,
      "epoch": 2.646599651822317,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004309081141634436,
      "loss": 5.1346,
      "mean_token_accuracy": 0.20787825733423232,
      "num_tokens": 52621745.0,
      "step": 23565
    },
    {
      "entropy": 5.371363401412964,
      "epoch": 2.6471612287302744,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00043087916406499635,
      "loss": 5.0733,
      "mean_token_accuracy": 0.212897190451622,
      "num_tokens": 52633089.0,
      "step": 23570
    },
    {
      "entropy": 5.362775611877441,
      "epoch": 2.647722805638232,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043085020900329604,
      "loss": 5.0982,
      "mean_token_accuracy": 0.20771678537130356,
      "num_tokens": 52644911.0,
      "step": 23575
    },
    {
      "entropy": 5.276778841018677,
      "epoch": 2.6482843825461897,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00043082124897926463,
      "loss": 4.8858,
      "mean_token_accuracy": 0.21941453367471694,
      "num_tokens": 52656958.0,
      "step": 23580
    },
    {
      "entropy": 5.348625707626343,
      "epoch": 2.6488459594541474,
      "grad_norm": 1.0,
      "learning_rate": 0.0004307922839938242,
      "loss": 5.1074,
      "mean_token_accuracy": 0.20620821118354798,
      "num_tokens": 52668988.0,
      "step": 23585
    },
    {
      "entropy": 5.323052501678466,
      "epoch": 2.649407536362105,
      "grad_norm": 0.96875,
      "learning_rate": 0.00043076331404789703,
      "loss": 5.0372,
      "mean_token_accuracy": 0.20957735627889634,
      "num_tokens": 52679948.0,
      "step": 23590
    },
    {
      "entropy": 5.3480224609375,
      "epoch": 2.6499691132700622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004307343391424055,
      "loss": 5.0035,
      "mean_token_accuracy": 0.2131124407052994,
      "num_tokens": 52690249.0,
      "step": 23595
    },
    {
      "entropy": 5.3486532211303714,
      "epoch": 2.65053069017802,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004307053592782723,
      "loss": 5.1148,
      "mean_token_accuracy": 0.2070415124297142,
      "num_tokens": 52702766.0,
      "step": 23600
    },
    {
      "entropy": 5.36473822593689,
      "epoch": 2.6510922670859776,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043067637445642004,
      "loss": 5.0748,
      "mean_token_accuracy": 0.21122482419013977,
      "num_tokens": 52714252.0,
      "step": 23605
    },
    {
      "entropy": 5.392329406738281,
      "epoch": 2.651653843993935,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00043064738467777166,
      "loss": 5.1358,
      "mean_token_accuracy": 0.20708727091550827,
      "num_tokens": 52724748.0,
      "step": 23610
    },
    {
      "entropy": 5.2881182670593265,
      "epoch": 2.6522154209018924,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043061838994325007,
      "loss": 5.0937,
      "mean_token_accuracy": 0.20903701335191727,
      "num_tokens": 52735636.0,
      "step": 23615
    },
    {
      "entropy": 5.453711128234863,
      "epoch": 2.65277699780985,
      "grad_norm": 1.0625,
      "learning_rate": 0.00043058939025377867,
      "loss": 5.2301,
      "mean_token_accuracy": 0.20017748326063156,
      "num_tokens": 52747812.0,
      "step": 23620
    },
    {
      "entropy": 5.358028411865234,
      "epoch": 2.6533385747178078,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004305603856102808,
      "loss": 5.0149,
      "mean_token_accuracy": 0.21469873636960984,
      "num_tokens": 52759322.0,
      "step": 23625
    },
    {
      "entropy": 5.327148342132569,
      "epoch": 2.6539001516257654,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004305313760136799,
      "loss": 5.0067,
      "mean_token_accuracy": 0.21212183982133864,
      "num_tokens": 52770238.0,
      "step": 23630
    },
    {
      "entropy": 5.371646928787231,
      "epoch": 2.6544617285337226,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00043050236146489955,
      "loss": 5.0289,
      "mean_token_accuracy": 0.2124057799577713,
      "num_tokens": 52781243.0,
      "step": 23635
    },
    {
      "entropy": 5.348789548873901,
      "epoch": 2.6550233054416803,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004304733419648638,
      "loss": 4.9943,
      "mean_token_accuracy": 0.2102017492055893,
      "num_tokens": 52791341.0,
      "step": 23640
    },
    {
      "entropy": 5.348156309127807,
      "epoch": 2.655584882349638,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00043044431751449656,
      "loss": 5.1068,
      "mean_token_accuracy": 0.20545905828475952,
      "num_tokens": 52803322.0,
      "step": 23645
    },
    {
      "entropy": 5.236481618881226,
      "epoch": 2.656146459257595,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0004304152881147219,
      "loss": 4.9126,
      "mean_token_accuracy": 0.2181096076965332,
      "num_tokens": 52813561.0,
      "step": 23650
    },
    {
      "entropy": 5.326382160186768,
      "epoch": 2.656708036165553,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004303862537664642,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20802494883537292,
      "num_tokens": 52824150.0,
      "step": 23655
    },
    {
      "entropy": 5.23766598701477,
      "epoch": 2.6572696130735105,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000430357214470648,
      "loss": 4.89,
      "mean_token_accuracy": 0.22090789824724197,
      "num_tokens": 52835013.0,
      "step": 23660
    },
    {
      "entropy": 5.35734806060791,
      "epoch": 2.6578311899814677,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004303281702281977,
      "loss": 5.1214,
      "mean_token_accuracy": 0.20988199561834336,
      "num_tokens": 52846682.0,
      "step": 23665
    },
    {
      "entropy": 5.240550422668457,
      "epoch": 2.6583927668894254,
      "grad_norm": 1.015625,
      "learning_rate": 0.00043029912104003827,
      "loss": 5.0021,
      "mean_token_accuracy": 0.20321325063705445,
      "num_tokens": 52857561.0,
      "step": 23670
    },
    {
      "entropy": 5.341693258285522,
      "epoch": 2.658954343797383,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00043027006690709457,
      "loss": 4.9684,
      "mean_token_accuracy": 0.2123485967516899,
      "num_tokens": 52868833.0,
      "step": 23675
    },
    {
      "entropy": 5.451469373703003,
      "epoch": 2.6595159207053407,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004302410078302918,
      "loss": 5.1151,
      "mean_token_accuracy": 0.2035816490650177,
      "num_tokens": 52879725.0,
      "step": 23680
    },
    {
      "entropy": 5.391719055175781,
      "epoch": 2.6600774976132984,
      "grad_norm": 1.046875,
      "learning_rate": 0.00043021194381055505,
      "loss": 5.1524,
      "mean_token_accuracy": 0.20112592428922654,
      "num_tokens": 52890585.0,
      "step": 23685
    },
    {
      "entropy": 5.391796016693116,
      "epoch": 2.6606390745212556,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00043018287484880986,
      "loss": 5.1632,
      "mean_token_accuracy": 0.20103542357683182,
      "num_tokens": 52902816.0,
      "step": 23690
    },
    {
      "entropy": 5.481287336349487,
      "epoch": 2.6612006514292132,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004301538009459817,
      "loss": 5.1934,
      "mean_token_accuracy": 0.2051415801048279,
      "num_tokens": 52914523.0,
      "step": 23695
    },
    {
      "entropy": 5.368372678756714,
      "epoch": 2.661762228337171,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004301247221029964,
      "loss": 5.0903,
      "mean_token_accuracy": 0.21363114714622497,
      "num_tokens": 52924953.0,
      "step": 23700
    },
    {
      "entropy": 5.260449171066284,
      "epoch": 2.662323805245128,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00043009563832077977,
      "loss": 4.9455,
      "mean_token_accuracy": 0.21240502446889878,
      "num_tokens": 52935435.0,
      "step": 23705
    },
    {
      "entropy": 5.334786796569825,
      "epoch": 2.662885382153086,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004300665496002578,
      "loss": 5.0798,
      "mean_token_accuracy": 0.21222809702157974,
      "num_tokens": 52946969.0,
      "step": 23710
    },
    {
      "entropy": 5.29271445274353,
      "epoch": 2.6634469590610435,
      "grad_norm": 1.03125,
      "learning_rate": 0.00043003745594235687,
      "loss": 5.0924,
      "mean_token_accuracy": 0.20423001646995545,
      "num_tokens": 52958634.0,
      "step": 23715
    },
    {
      "entropy": 5.326714754104614,
      "epoch": 2.664008535969001,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004300083573480031,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21710915118455887,
      "num_tokens": 52968412.0,
      "step": 23720
    },
    {
      "entropy": 5.257176446914673,
      "epoch": 2.6645701128769588,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042997925381812313,
      "loss": 4.9533,
      "mean_token_accuracy": 0.20609032809734346,
      "num_tokens": 52978653.0,
      "step": 23725
    },
    {
      "entropy": 5.305527067184448,
      "epoch": 2.665131689784916,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042995014535364363,
      "loss": 5.0183,
      "mean_token_accuracy": 0.2103234127163887,
      "num_tokens": 52988518.0,
      "step": 23730
    },
    {
      "entropy": 5.305840396881104,
      "epoch": 2.6656932666928737,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004299210319554914,
      "loss": 4.9812,
      "mean_token_accuracy": 0.208907787501812,
      "num_tokens": 52999271.0,
      "step": 23735
    },
    {
      "entropy": 5.32987232208252,
      "epoch": 2.6662548436008313,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042989191362459347,
      "loss": 5.089,
      "mean_token_accuracy": 0.21576597839593886,
      "num_tokens": 53010530.0,
      "step": 23740
    },
    {
      "entropy": 5.303563737869263,
      "epoch": 2.6668164205087885,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004298627903618769,
      "loss": 5.0768,
      "mean_token_accuracy": 0.206349977850914,
      "num_tokens": 53021631.0,
      "step": 23745
    },
    {
      "entropy": 5.355739402770996,
      "epoch": 2.667377997416746,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042983366216826907,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21247007101774215,
      "num_tokens": 53033763.0,
      "step": 23750
    },
    {
      "entropy": 5.295595407485962,
      "epoch": 2.667939574324704,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004298045290446973,
      "loss": 5.0775,
      "mean_token_accuracy": 0.20784296244382858,
      "num_tokens": 53045839.0,
      "step": 23755
    },
    {
      "entropy": 5.365728664398193,
      "epoch": 2.668501151232661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042977539099208937,
      "loss": 5.0712,
      "mean_token_accuracy": 0.20375602096319198,
      "num_tokens": 53057295.0,
      "step": 23760
    },
    {
      "entropy": 5.483901739120483,
      "epoch": 2.6690627281406187,
      "grad_norm": 1.0,
      "learning_rate": 0.00042974624801137283,
      "loss": 5.276,
      "mean_token_accuracy": 0.19350766092538835,
      "num_tokens": 53068595.0,
      "step": 23765
    },
    {
      "entropy": 5.330386400222778,
      "epoch": 2.6696243050485764,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042971710010347594,
      "loss": 5.0314,
      "mean_token_accuracy": 0.2106096088886261,
      "num_tokens": 53080170.0,
      "step": 23770
    },
    {
      "entropy": 5.268099975585938,
      "epoch": 2.670185881956534,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042968794726932634,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21193926334381102,
      "num_tokens": 53090894.0,
      "step": 23775
    },
    {
      "entropy": 5.34975757598877,
      "epoch": 2.6707474588644917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004296587895098527,
      "loss": 5.1504,
      "mean_token_accuracy": 0.20355963557958603,
      "num_tokens": 53103732.0,
      "step": 23780
    },
    {
      "entropy": 5.445427274703979,
      "epoch": 2.671309035772449,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042962962682598316,
      "loss": 5.0648,
      "mean_token_accuracy": 0.20785606205463408,
      "num_tokens": 53115237.0,
      "step": 23785
    },
    {
      "entropy": 5.359000968933105,
      "epoch": 2.6718706126804066,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042960045921864627,
      "loss": 5.0148,
      "mean_token_accuracy": 0.21048446893692016,
      "num_tokens": 53125638.0,
      "step": 23790
    },
    {
      "entropy": 5.248629426956176,
      "epoch": 2.6724321895883643,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004295712866887709,
      "loss": 4.9896,
      "mean_token_accuracy": 0.21710642129182817,
      "num_tokens": 53136153.0,
      "step": 23795
    },
    {
      "entropy": 5.338393878936768,
      "epoch": 2.6729937664963215,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042954210923728565,
      "loss": 5.0764,
      "mean_token_accuracy": 0.20773268043994902,
      "num_tokens": 53146562.0,
      "step": 23800
    },
    {
      "entropy": 5.469605493545532,
      "epoch": 2.673555343404279,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004295129268651198,
      "loss": 5.172,
      "mean_token_accuracy": 0.20251040309667587,
      "num_tokens": 53158460.0,
      "step": 23805
    },
    {
      "entropy": 5.341962814331055,
      "epoch": 2.674116920312237,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004294837395732025,
      "loss": 5.0518,
      "mean_token_accuracy": 0.21168941408395767,
      "num_tokens": 53170576.0,
      "step": 23810
    },
    {
      "entropy": 5.328232955932617,
      "epoch": 2.6746784972201945,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042945454736246293,
      "loss": 4.9587,
      "mean_token_accuracy": 0.21866464167833327,
      "num_tokens": 53181658.0,
      "step": 23815
    },
    {
      "entropy": 5.3662437915802,
      "epoch": 2.675240074128152,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042942535023383075,
      "loss": 5.1052,
      "mean_token_accuracy": 0.2026084065437317,
      "num_tokens": 53193983.0,
      "step": 23820
    },
    {
      "entropy": 5.288748455047608,
      "epoch": 2.6758016510361093,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042939614818823547,
      "loss": 5.0338,
      "mean_token_accuracy": 0.20363822877407073,
      "num_tokens": 53205603.0,
      "step": 23825
    },
    {
      "entropy": 5.251203489303589,
      "epoch": 2.676363227944067,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004293669412266069,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21425586342811584,
      "num_tokens": 53216295.0,
      "step": 23830
    },
    {
      "entropy": 5.347871780395508,
      "epoch": 2.6769248048520247,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004293377293498751,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21420655399560928,
      "num_tokens": 53227055.0,
      "step": 23835
    },
    {
      "entropy": 5.351640748977661,
      "epoch": 2.677486381759982,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00042930851255897014,
      "loss": 5.0578,
      "mean_token_accuracy": 0.21233672201633452,
      "num_tokens": 53239050.0,
      "step": 23840
    },
    {
      "entropy": 5.2699038028717045,
      "epoch": 2.6780479586679395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004292792908548223,
      "loss": 4.9385,
      "mean_token_accuracy": 0.2129247948527336,
      "num_tokens": 53249551.0,
      "step": 23845
    },
    {
      "entropy": 5.380059862136841,
      "epoch": 2.678609535575897,
      "grad_norm": 1.0,
      "learning_rate": 0.00042925006423836197,
      "loss": 5.1353,
      "mean_token_accuracy": 0.20429541319608688,
      "num_tokens": 53261645.0,
      "step": 23850
    },
    {
      "entropy": 5.420942640304565,
      "epoch": 2.6791711124838544,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042922083271051975,
      "loss": 5.1258,
      "mean_token_accuracy": 0.20256736278533935,
      "num_tokens": 53272601.0,
      "step": 23855
    },
    {
      "entropy": 5.328851795196533,
      "epoch": 2.679732689391812,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004291915962722265,
      "loss": 5.069,
      "mean_token_accuracy": 0.21110014766454696,
      "num_tokens": 53283190.0,
      "step": 23860
    },
    {
      "entropy": 5.317771053314209,
      "epoch": 2.6802942662997697,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004291623549244129,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21311812549829484,
      "num_tokens": 53294374.0,
      "step": 23865
    },
    {
      "entropy": 5.398582553863525,
      "epoch": 2.6808558432077274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042913310866801024,
      "loss": 5.1683,
      "mean_token_accuracy": 0.20171941369771956,
      "num_tokens": 53305609.0,
      "step": 23870
    },
    {
      "entropy": 5.232536697387696,
      "epoch": 2.681417420115685,
      "grad_norm": 0.9375,
      "learning_rate": 0.00042910385750394964,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21547982841730118,
      "num_tokens": 53316251.0,
      "step": 23875
    },
    {
      "entropy": 5.2967016220092775,
      "epoch": 2.6819789970236423,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004290746014331623,
      "loss": 4.9699,
      "mean_token_accuracy": 0.21411479860544205,
      "num_tokens": 53326798.0,
      "step": 23880
    },
    {
      "entropy": 5.365123176574707,
      "epoch": 2.6825405739316,
      "grad_norm": 1.0,
      "learning_rate": 0.00042904534045658,
      "loss": 5.076,
      "mean_token_accuracy": 0.20420823246240616,
      "num_tokens": 53337586.0,
      "step": 23885
    },
    {
      "entropy": 5.351949167251587,
      "epoch": 2.6831021508395576,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004290160745751343,
      "loss": 5.1025,
      "mean_token_accuracy": 0.2028505876660347,
      "num_tokens": 53348517.0,
      "step": 23890
    },
    {
      "entropy": 5.374739027023315,
      "epoch": 2.683663727747515,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004289868037897571,
      "loss": 5.0414,
      "mean_token_accuracy": 0.21362991631031036,
      "num_tokens": 53358937.0,
      "step": 23895
    },
    {
      "entropy": 5.3121216773986815,
      "epoch": 2.6842253046554725,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004289575281013802,
      "loss": 5.0303,
      "mean_token_accuracy": 0.21353475898504257,
      "num_tokens": 53370007.0,
      "step": 23900
    },
    {
      "entropy": 5.279910612106323,
      "epoch": 2.68478688156343,
      "grad_norm": 1.09375,
      "learning_rate": 0.000428928247510936,
      "loss": 4.9154,
      "mean_token_accuracy": 0.21412036567926407,
      "num_tokens": 53380913.0,
      "step": 23905
    },
    {
      "entropy": 5.286054706573486,
      "epoch": 2.685348458471388,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004288989620193567,
      "loss": 5.0349,
      "mean_token_accuracy": 0.20693573951721192,
      "num_tokens": 53392526.0,
      "step": 23910
    },
    {
      "entropy": 5.33939118385315,
      "epoch": 2.6859100353793455,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042886967162757475,
      "loss": 5.065,
      "mean_token_accuracy": 0.2144103690981865,
      "num_tokens": 53404522.0,
      "step": 23915
    },
    {
      "entropy": 5.314584970474243,
      "epoch": 2.6864716122873027,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042884037633652276,
      "loss": 5.0053,
      "mean_token_accuracy": 0.21183351427316666,
      "num_tokens": 53415156.0,
      "step": 23920
    },
    {
      "entropy": 5.354341316223144,
      "epoch": 2.6870331891952604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00042881107614713355,
      "loss": 5.0682,
      "mean_token_accuracy": 0.1974719375371933,
      "num_tokens": 53426297.0,
      "step": 23925
    },
    {
      "entropy": 5.468953704833984,
      "epoch": 2.687594766103218,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042878177106034,
      "loss": 5.159,
      "mean_token_accuracy": 0.2000969648361206,
      "num_tokens": 53438415.0,
      "step": 23930
    },
    {
      "entropy": 5.297367858886719,
      "epoch": 2.6881563430111752,
      "grad_norm": 1.0,
      "learning_rate": 0.0004287524610770752,
      "loss": 5.1227,
      "mean_token_accuracy": 0.203294338285923,
      "num_tokens": 53450103.0,
      "step": 23935
    },
    {
      "entropy": 5.305332136154175,
      "epoch": 2.688717919919133,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042872314619827234,
      "loss": 4.9646,
      "mean_token_accuracy": 0.2108446776866913,
      "num_tokens": 53461440.0,
      "step": 23940
    },
    {
      "entropy": 5.471339225769043,
      "epoch": 2.6892794968270906,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000428693826424865,
      "loss": 5.1729,
      "mean_token_accuracy": 0.1993219166994095,
      "num_tokens": 53472369.0,
      "step": 23945
    },
    {
      "entropy": 5.368491363525391,
      "epoch": 2.6898410737350478,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004286645017577866,
      "loss": 5.085,
      "mean_token_accuracy": 0.21478749215602874,
      "num_tokens": 53483739.0,
      "step": 23950
    },
    {
      "entropy": 5.378449296951294,
      "epoch": 2.6904026506430054,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042863517219797085,
      "loss": 5.1169,
      "mean_token_accuracy": 0.20360804051160813,
      "num_tokens": 53494131.0,
      "step": 23955
    },
    {
      "entropy": 5.398599338531494,
      "epoch": 2.690964227550963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042860583774635165,
      "loss": 5.1195,
      "mean_token_accuracy": 0.20506985038518905,
      "num_tokens": 53504923.0,
      "step": 23960
    },
    {
      "entropy": 5.403127670288086,
      "epoch": 2.6915258044589208,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000428576498403863,
      "loss": 5.1268,
      "mean_token_accuracy": 0.1986025959253311,
      "num_tokens": 53516660.0,
      "step": 23965
    },
    {
      "entropy": 5.280955648422241,
      "epoch": 2.6920873813668784,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000428547154171439,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21741465330123902,
      "num_tokens": 53527617.0,
      "step": 23970
    },
    {
      "entropy": 5.328863477706909,
      "epoch": 2.6926489582748356,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004285178050500142,
      "loss": 5.0625,
      "mean_token_accuracy": 0.20478272438049316,
      "num_tokens": 53537830.0,
      "step": 23975
    },
    {
      "entropy": 5.444429492950439,
      "epoch": 2.6932105351827933,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004284884510405228,
      "loss": 5.1806,
      "mean_token_accuracy": 0.2014629527926445,
      "num_tokens": 53549981.0,
      "step": 23980
    },
    {
      "entropy": 5.4127280712127686,
      "epoch": 2.693772112090751,
      "grad_norm": 1.125,
      "learning_rate": 0.0004284590921438997,
      "loss": 5.0584,
      "mean_token_accuracy": 0.21297961622476577,
      "num_tokens": 53559941.0,
      "step": 23985
    },
    {
      "entropy": 5.234043264389038,
      "epoch": 2.694333688998708,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042842972836107954,
      "loss": 4.9321,
      "mean_token_accuracy": 0.2121164694428444,
      "num_tokens": 53571611.0,
      "step": 23990
    },
    {
      "entropy": 5.386085128784179,
      "epoch": 2.694895265906666,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004284003596929973,
      "loss": 5.0023,
      "mean_token_accuracy": 0.20832203775644303,
      "num_tokens": 53582425.0,
      "step": 23995
    },
    {
      "entropy": 5.383046436309814,
      "epoch": 2.6954568428146235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042837098614058815,
      "loss": 5.1163,
      "mean_token_accuracy": 0.2039183035492897,
      "num_tokens": 53593576.0,
      "step": 24000
    },
    {
      "epoch": 2.6954568428146235,
      "eval_entropy": 5.147169432321529,
      "eval_loss": 5.184617519378662,
      "eval_mean_token_accuracy": 0.21068443574043996,
      "eval_num_tokens": 53593576.0,
      "eval_runtime": 23.9278,
      "eval_samples_per_second": 1296.108,
      "eval_steps_per_second": 162.029,
      "step": 24000
    },
    {
      "entropy": 5.303683948516846,
      "epoch": 2.696018419722581,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042834160770478726,
      "loss": 5.0687,
      "mean_token_accuracy": 0.20835837423801423,
      "num_tokens": 53603768.0,
      "step": 24005
    },
    {
      "entropy": 5.317247486114502,
      "epoch": 2.696579996630539,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042831222438653016,
      "loss": 4.9793,
      "mean_token_accuracy": 0.2124404489994049,
      "num_tokens": 53614464.0,
      "step": 24010
    },
    {
      "entropy": 5.365368986129761,
      "epoch": 2.697141573538496,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004282828361867524,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20209968090057373,
      "num_tokens": 53625641.0,
      "step": 24015
    },
    {
      "entropy": 5.374537992477417,
      "epoch": 2.6977031504464537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042825344310638966,
      "loss": 5.1362,
      "mean_token_accuracy": 0.20645972937345505,
      "num_tokens": 53636741.0,
      "step": 24020
    },
    {
      "entropy": 5.368784475326538,
      "epoch": 2.6982647273544114,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042822404514637785,
      "loss": 5.0278,
      "mean_token_accuracy": 0.20842444598674775,
      "num_tokens": 53647295.0,
      "step": 24025
    },
    {
      "entropy": 5.337030267715454,
      "epoch": 2.6988263042623686,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000428194642307653,
      "loss": 5.0322,
      "mean_token_accuracy": 0.21799342930316926,
      "num_tokens": 53658371.0,
      "step": 24030
    },
    {
      "entropy": 5.316513013839722,
      "epoch": 2.6993878811703262,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004281652345911513,
      "loss": 5.0727,
      "mean_token_accuracy": 0.21146162003278732,
      "num_tokens": 53669534.0,
      "step": 24035
    },
    {
      "entropy": 5.350559854507447,
      "epoch": 2.699949458078284,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042813582199780913,
      "loss": 5.091,
      "mean_token_accuracy": 0.2071331784129143,
      "num_tokens": 53681247.0,
      "step": 24040
    },
    {
      "entropy": 5.351489686965943,
      "epoch": 2.700511034986241,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004281064045285631,
      "loss": 5.1196,
      "mean_token_accuracy": 0.19908246546983718,
      "num_tokens": 53692197.0,
      "step": 24045
    },
    {
      "entropy": 5.312388134002686,
      "epoch": 2.701072611894199,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042807698218434964,
      "loss": 5.0446,
      "mean_token_accuracy": 0.20646736174821853,
      "num_tokens": 53703734.0,
      "step": 24050
    },
    {
      "entropy": 5.384285974502563,
      "epoch": 2.7016341888021564,
      "grad_norm": 0.96875,
      "learning_rate": 0.00042804755496610576,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20638844668865203,
      "num_tokens": 53714888.0,
      "step": 24055
    },
    {
      "entropy": 5.3299723148345945,
      "epoch": 2.702195765710114,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042801812287476836,
      "loss": 5.0204,
      "mean_token_accuracy": 0.209059277176857,
      "num_tokens": 53726231.0,
      "step": 24060
    },
    {
      "entropy": 5.355964851379395,
      "epoch": 2.7027573426180718,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004279886859112746,
      "loss": 5.0772,
      "mean_token_accuracy": 0.2067788764834404,
      "num_tokens": 53737777.0,
      "step": 24065
    },
    {
      "entropy": 5.3608767032623295,
      "epoch": 2.703318919526029,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042795924407656175,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20884765237569808,
      "num_tokens": 53748277.0,
      "step": 24070
    },
    {
      "entropy": 5.31551513671875,
      "epoch": 2.7038804964339866,
      "grad_norm": 1.0,
      "learning_rate": 0.00042792979737156717,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21192601770162584,
      "num_tokens": 53759205.0,
      "step": 24075
    },
    {
      "entropy": 5.3343969821929935,
      "epoch": 2.7044420733419443,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004279003457972286,
      "loss": 5.0914,
      "mean_token_accuracy": 0.19732648879289627,
      "num_tokens": 53770984.0,
      "step": 24080
    },
    {
      "entropy": 5.290869951248169,
      "epoch": 2.7050036502499015,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004278708893544837,
      "loss": 4.971,
      "mean_token_accuracy": 0.21295663565397263,
      "num_tokens": 53782092.0,
      "step": 24085
    },
    {
      "entropy": 5.323276424407959,
      "epoch": 2.705565227157859,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004278414280442704,
      "loss": 5.0355,
      "mean_token_accuracy": 0.2055908441543579,
      "num_tokens": 53794136.0,
      "step": 24090
    },
    {
      "entropy": 5.396450328826904,
      "epoch": 2.706126804065817,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042781196186752676,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20272564589977266,
      "num_tokens": 53804612.0,
      "step": 24095
    },
    {
      "entropy": 5.297765159606934,
      "epoch": 2.7066883809737745,
      "grad_norm": 1.0,
      "learning_rate": 0.0004277824908251909,
      "loss": 5.0424,
      "mean_token_accuracy": 0.20423139929771422,
      "num_tokens": 53815850.0,
      "step": 24100
    },
    {
      "entropy": 5.292834997177124,
      "epoch": 2.707249957881732,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004277530149182014,
      "loss": 5.0148,
      "mean_token_accuracy": 0.2112646222114563,
      "num_tokens": 53827050.0,
      "step": 24105
    },
    {
      "entropy": 5.4006969928741455,
      "epoch": 2.7078115347896894,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042772353414749655,
      "loss": 5.1044,
      "mean_token_accuracy": 0.2024707615375519,
      "num_tokens": 53837697.0,
      "step": 24110
    },
    {
      "entropy": 5.29635968208313,
      "epoch": 2.708373111697647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042769404851401524,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2122233182191849,
      "num_tokens": 53849278.0,
      "step": 24115
    },
    {
      "entropy": 5.33439040184021,
      "epoch": 2.7089346886056047,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004276645580186961,
      "loss": 5.098,
      "mean_token_accuracy": 0.20332104861736297,
      "num_tokens": 53861951.0,
      "step": 24120
    },
    {
      "entropy": 5.376227235794067,
      "epoch": 2.709496265513562,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042763506266247834,
      "loss": 4.9578,
      "mean_token_accuracy": 0.22051245272159575,
      "num_tokens": 53873647.0,
      "step": 24125
    },
    {
      "entropy": 5.275733470916748,
      "epoch": 2.7100578424215196,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004276055624463009,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2077844724059105,
      "num_tokens": 53885063.0,
      "step": 24130
    },
    {
      "entropy": 5.291540479660034,
      "epoch": 2.7106194193294773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042757605737110324,
      "loss": 5.0778,
      "mean_token_accuracy": 0.2086331754922867,
      "num_tokens": 53896955.0,
      "step": 24135
    },
    {
      "entropy": 5.286900806427002,
      "epoch": 2.7111809962374345,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004275465474378246,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21677857488393784,
      "num_tokens": 53907862.0,
      "step": 24140
    },
    {
      "entropy": 5.355318164825439,
      "epoch": 2.711742573145392,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004275170326474049,
      "loss": 5.069,
      "mean_token_accuracy": 0.21115175485610962,
      "num_tokens": 53918918.0,
      "step": 24145
    },
    {
      "entropy": 5.328081703186035,
      "epoch": 2.71230415005335,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00042748751300078367,
      "loss": 5.0776,
      "mean_token_accuracy": 0.2090213656425476,
      "num_tokens": 53930488.0,
      "step": 24150
    },
    {
      "entropy": 5.407615661621094,
      "epoch": 2.7128657269613075,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004274579884989008,
      "loss": 5.0593,
      "mean_token_accuracy": 0.21274656802415848,
      "num_tokens": 53942314.0,
      "step": 24155
    },
    {
      "entropy": 5.343293380737305,
      "epoch": 2.713427303869265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042742845914269645,
      "loss": 5.0926,
      "mean_token_accuracy": 0.22173751443624495,
      "num_tokens": 53951915.0,
      "step": 24160
    },
    {
      "entropy": 5.302781581878662,
      "epoch": 2.7139888807772223,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042739892493311095,
      "loss": 5.0884,
      "mean_token_accuracy": 0.21447746455669403,
      "num_tokens": 53962815.0,
      "step": 24165
    },
    {
      "entropy": 5.288300657272339,
      "epoch": 2.71455045768518,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042736938587108447,
      "loss": 5.0105,
      "mean_token_accuracy": 0.20835032165050507,
      "num_tokens": 53972740.0,
      "step": 24170
    },
    {
      "entropy": 5.375681209564209,
      "epoch": 2.7151120345931377,
      "grad_norm": 1.0,
      "learning_rate": 0.0004273398419575577,
      "loss": 5.0294,
      "mean_token_accuracy": 0.20849468261003495,
      "num_tokens": 53983632.0,
      "step": 24175
    },
    {
      "entropy": 5.420884037017823,
      "epoch": 2.715673611501095,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004273102931934713,
      "loss": 5.1214,
      "mean_token_accuracy": 0.2076072096824646,
      "num_tokens": 53994085.0,
      "step": 24180
    },
    {
      "entropy": 5.342078018188476,
      "epoch": 2.7162351884090525,
      "grad_norm": 1.078125,
      "learning_rate": 0.000427280739579766,
      "loss": 5.0409,
      "mean_token_accuracy": 0.21353916078805923,
      "num_tokens": 54004106.0,
      "step": 24185
    },
    {
      "entropy": 5.268125391006469,
      "epoch": 2.71679676531701,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042725118111738305,
      "loss": 4.9165,
      "mean_token_accuracy": 0.21636734455823897,
      "num_tokens": 54014582.0,
      "step": 24190
    },
    {
      "entropy": 5.35806450843811,
      "epoch": 2.717358342224968,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004272216178072633,
      "loss": 5.0652,
      "mean_token_accuracy": 0.20815996676683426,
      "num_tokens": 54025626.0,
      "step": 24195
    },
    {
      "entropy": 5.327173376083374,
      "epoch": 2.7179199191329255,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004271920496503483,
      "loss": 5.0811,
      "mean_token_accuracy": 0.20943896621465682,
      "num_tokens": 54036974.0,
      "step": 24200
    },
    {
      "entropy": 5.390891075134277,
      "epoch": 2.7184814960408827,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042716247664757934,
      "loss": 5.0617,
      "mean_token_accuracy": 0.20888594686985015,
      "num_tokens": 54048404.0,
      "step": 24205
    },
    {
      "entropy": 5.370880365371704,
      "epoch": 2.7190430729488404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042713289879989816,
      "loss": 5.112,
      "mean_token_accuracy": 0.2097370982170105,
      "num_tokens": 54059210.0,
      "step": 24210
    },
    {
      "entropy": 5.371309995651245,
      "epoch": 2.719604649856798,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004271033161082464,
      "loss": 5.1007,
      "mean_token_accuracy": 0.20762156397104264,
      "num_tokens": 54070406.0,
      "step": 24215
    },
    {
      "entropy": 5.282644128799438,
      "epoch": 2.7201662267647553,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042707372857356615,
      "loss": 5.0209,
      "mean_token_accuracy": 0.20549097508192063,
      "num_tokens": 54081497.0,
      "step": 24220
    },
    {
      "entropy": 5.197215890884399,
      "epoch": 2.720727803672713,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004270441361967994,
      "loss": 4.9037,
      "mean_token_accuracy": 0.22109153717756272,
      "num_tokens": 54090934.0,
      "step": 24225
    },
    {
      "entropy": 5.229683256149292,
      "epoch": 2.7212893805806706,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004270145389788883,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2165642186999321,
      "num_tokens": 54102011.0,
      "step": 24230
    },
    {
      "entropy": 5.43311915397644,
      "epoch": 2.721850957488628,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042698493692077533,
      "loss": 5.1101,
      "mean_token_accuracy": 0.20836647748947143,
      "num_tokens": 54113066.0,
      "step": 24235
    },
    {
      "entropy": 5.240908622741699,
      "epoch": 2.7224125343965855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004269553300234031,
      "loss": 4.9326,
      "mean_token_accuracy": 0.21689112782478331,
      "num_tokens": 54124291.0,
      "step": 24240
    },
    {
      "entropy": 5.3498091220855715,
      "epoch": 2.722974111304543,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042692571828771423,
      "loss": 5.1765,
      "mean_token_accuracy": 0.20289139449596405,
      "num_tokens": 54136147.0,
      "step": 24245
    },
    {
      "entropy": 5.405992698669434,
      "epoch": 2.723535688212501,
      "grad_norm": 1.21875,
      "learning_rate": 0.00042689610171465147,
      "loss": 5.0692,
      "mean_token_accuracy": 0.2023770421743393,
      "num_tokens": 54147953.0,
      "step": 24250
    },
    {
      "entropy": 5.3296812057495115,
      "epoch": 2.7240972651204585,
      "grad_norm": 1.046875,
      "learning_rate": 0.000426866480305158,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21206077933311462,
      "num_tokens": 54159498.0,
      "step": 24255
    },
    {
      "entropy": 5.318127202987671,
      "epoch": 2.7246588420284157,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004268368540601769,
      "loss": 5.0254,
      "mean_token_accuracy": 0.2107198268175125,
      "num_tokens": 54170380.0,
      "step": 24260
    },
    {
      "entropy": 5.324195766448975,
      "epoch": 2.7252204189363733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004268072229806514,
      "loss": 5.116,
      "mean_token_accuracy": 0.20388455241918563,
      "num_tokens": 54181105.0,
      "step": 24265
    },
    {
      "entropy": 5.492804622650146,
      "epoch": 2.725781995844331,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004267775870675251,
      "loss": 5.1326,
      "mean_token_accuracy": 0.19865497052669526,
      "num_tokens": 54192142.0,
      "step": 24270
    },
    {
      "entropy": 5.2816626071929935,
      "epoch": 2.7263435727522882,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004267479463217415,
      "loss": 5.0635,
      "mean_token_accuracy": 0.2035185158252716,
      "num_tokens": 54202716.0,
      "step": 24275
    },
    {
      "entropy": 5.242980051040649,
      "epoch": 2.726905149660246,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042671830074424453,
      "loss": 4.9468,
      "mean_token_accuracy": 0.21777301877737046,
      "num_tokens": 54213035.0,
      "step": 24280
    },
    {
      "entropy": 5.311383867263794,
      "epoch": 2.7274667265682035,
      "grad_norm": 1.03125,
      "learning_rate": 0.000426688650335978,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21017018109560012,
      "num_tokens": 54223999.0,
      "step": 24285
    },
    {
      "entropy": 5.304384565353393,
      "epoch": 2.728028303476161,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00042665899509788593,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21400246769189835,
      "num_tokens": 54234393.0,
      "step": 24290
    },
    {
      "entropy": 5.24182538986206,
      "epoch": 2.728589880384119,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004266293350309128,
      "loss": 5.0172,
      "mean_token_accuracy": 0.21179965138435364,
      "num_tokens": 54244787.0,
      "step": 24295
    },
    {
      "entropy": 5.358412599563598,
      "epoch": 2.729151457292076,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004265996701360026,
      "loss": 5.0745,
      "mean_token_accuracy": 0.20280406326055528,
      "num_tokens": 54256033.0,
      "step": 24300
    },
    {
      "entropy": 5.38976788520813,
      "epoch": 2.7297130342000337,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042657000041410026,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2115352213382721,
      "num_tokens": 54266796.0,
      "step": 24305
    },
    {
      "entropy": 5.385937356948853,
      "epoch": 2.7302746111079914,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004265403258661503,
      "loss": 5.1248,
      "mean_token_accuracy": 0.2007625639438629,
      "num_tokens": 54279049.0,
      "step": 24310
    },
    {
      "entropy": 5.321739435195923,
      "epoch": 2.7308361880159486,
      "grad_norm": 1.0,
      "learning_rate": 0.0004265106464930976,
      "loss": 5.0618,
      "mean_token_accuracy": 0.206312857568264,
      "num_tokens": 54290544.0,
      "step": 24315
    },
    {
      "entropy": 5.39624924659729,
      "epoch": 2.7313977649239063,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00042648096229588715,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21245255768299104,
      "num_tokens": 54300905.0,
      "step": 24320
    },
    {
      "entropy": 5.322198677062988,
      "epoch": 2.731959341831864,
      "grad_norm": 1.0,
      "learning_rate": 0.00042645127327546405,
      "loss": 4.9558,
      "mean_token_accuracy": 0.21144359409809113,
      "num_tokens": 54313019.0,
      "step": 24325
    },
    {
      "entropy": 5.359595441818238,
      "epoch": 2.732520918739821,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042642157943277375,
      "loss": 5.1238,
      "mean_token_accuracy": 0.20542749166488647,
      "num_tokens": 54324034.0,
      "step": 24330
    },
    {
      "entropy": 5.271572303771973,
      "epoch": 2.733082495647779,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004263918807687616,
      "loss": 4.9947,
      "mean_token_accuracy": 0.21264663636684417,
      "num_tokens": 54335335.0,
      "step": 24335
    },
    {
      "entropy": 5.383876514434815,
      "epoch": 2.7336440725557365,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004263621772843733,
      "loss": 5.1137,
      "mean_token_accuracy": 0.20981981307268144,
      "num_tokens": 54345659.0,
      "step": 24340
    },
    {
      "entropy": 5.397920227050781,
      "epoch": 2.734205649463694,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004263324689805545,
      "loss": 5.1053,
      "mean_token_accuracy": 0.2050393357872963,
      "num_tokens": 54357959.0,
      "step": 24345
    },
    {
      "entropy": 5.348810720443725,
      "epoch": 2.734767226371652,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004263027558582512,
      "loss": 5.0483,
      "mean_token_accuracy": 0.2081203818321228,
      "num_tokens": 54368779.0,
      "step": 24350
    },
    {
      "entropy": 5.3485723495483395,
      "epoch": 2.735328803279609,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004262730379184095,
      "loss": 5.088,
      "mean_token_accuracy": 0.20457210391759872,
      "num_tokens": 54381726.0,
      "step": 24355
    },
    {
      "entropy": 5.380208492279053,
      "epoch": 2.7358903801875667,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042624331516197554,
      "loss": 5.0734,
      "mean_token_accuracy": 0.21135687530040742,
      "num_tokens": 54393201.0,
      "step": 24360
    },
    {
      "entropy": 5.373672389984131,
      "epoch": 2.7364519570955244,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042621358758989584,
      "loss": 5.0432,
      "mean_token_accuracy": 0.21532459259033204,
      "num_tokens": 54404178.0,
      "step": 24365
    },
    {
      "entropy": 5.225181627273559,
      "epoch": 2.7370135340034816,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004261838552031169,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21809635907411576,
      "num_tokens": 54414963.0,
      "step": 24370
    },
    {
      "entropy": 5.224365472793579,
      "epoch": 2.7375751109114392,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00042615411800258523,
      "loss": 4.9228,
      "mean_token_accuracy": 0.2211331382393837,
      "num_tokens": 54425333.0,
      "step": 24375
    },
    {
      "entropy": 5.32285966873169,
      "epoch": 2.738136687819397,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042612437598924793,
      "loss": 5.1397,
      "mean_token_accuracy": 0.2027912437915802,
      "num_tokens": 54436326.0,
      "step": 24380
    },
    {
      "entropy": 5.338959121704102,
      "epoch": 2.7386982647273546,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004260946291640519,
      "loss": 5.0104,
      "mean_token_accuracy": 0.2113371953368187,
      "num_tokens": 54446664.0,
      "step": 24385
    },
    {
      "entropy": 5.385915899276734,
      "epoch": 2.739259841635312,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042606487752794417,
      "loss": 5.0305,
      "mean_token_accuracy": 0.21417651921510697,
      "num_tokens": 54457135.0,
      "step": 24390
    },
    {
      "entropy": 5.252908611297608,
      "epoch": 2.7398214185432694,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042603512108187224,
      "loss": 4.9569,
      "mean_token_accuracy": 0.20809108316898345,
      "num_tokens": 54468248.0,
      "step": 24395
    },
    {
      "entropy": 5.300257873535156,
      "epoch": 2.740382995451227,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042600535982678347,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22223865389823913,
      "num_tokens": 54478676.0,
      "step": 24400
    },
    {
      "entropy": 5.248467016220093,
      "epoch": 2.7409445723591848,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042597559376362544,
      "loss": 5.0058,
      "mean_token_accuracy": 0.2023841619491577,
      "num_tokens": 54490342.0,
      "step": 24405
    },
    {
      "entropy": 5.2956455707550045,
      "epoch": 2.741506149267142,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004259458228933459,
      "loss": 5.0085,
      "mean_token_accuracy": 0.21476971358060837,
      "num_tokens": 54501543.0,
      "step": 24410
    },
    {
      "entropy": 5.310221385955811,
      "epoch": 2.7420677261750996,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042591604721689297,
      "loss": 5.011,
      "mean_token_accuracy": 0.21097420006990433,
      "num_tokens": 54511777.0,
      "step": 24415
    },
    {
      "entropy": 5.365600347518921,
      "epoch": 2.7426293030830573,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004258862667352145,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21666189581155776,
      "num_tokens": 54523255.0,
      "step": 24420
    },
    {
      "entropy": 5.340653324127198,
      "epoch": 2.7431908799910145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004258564814492587,
      "loss": 5.067,
      "mean_token_accuracy": 0.20161346793174745,
      "num_tokens": 54534158.0,
      "step": 24425
    },
    {
      "entropy": 5.4229789733886715,
      "epoch": 2.743752456898972,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000425826691359974,
      "loss": 5.1097,
      "mean_token_accuracy": 0.2083696499466896,
      "num_tokens": 54545735.0,
      "step": 24430
    },
    {
      "entropy": 5.390956926345825,
      "epoch": 2.74431403380693,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004257968964683091,
      "loss": 5.0831,
      "mean_token_accuracy": 0.210528427362442,
      "num_tokens": 54557112.0,
      "step": 24435
    },
    {
      "entropy": 5.430970048904419,
      "epoch": 2.7448756107148875,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004257670967752125,
      "loss": 5.0998,
      "mean_token_accuracy": 0.20253605246543885,
      "num_tokens": 54568600.0,
      "step": 24440
    },
    {
      "entropy": 5.362000846862793,
      "epoch": 2.745437187622845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000425737292281633,
      "loss": 4.992,
      "mean_token_accuracy": 0.21306524574756622,
      "num_tokens": 54579008.0,
      "step": 24445
    },
    {
      "entropy": 5.265348291397094,
      "epoch": 2.7459987645308024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004257074829885196,
      "loss": 5.1083,
      "mean_token_accuracy": 0.2049977406859398,
      "num_tokens": 54590306.0,
      "step": 24450
    },
    {
      "entropy": 5.361380863189697,
      "epoch": 2.74656034143876,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004256776688968216,
      "loss": 5.0855,
      "mean_token_accuracy": 0.2032221332192421,
      "num_tokens": 54600999.0,
      "step": 24455
    },
    {
      "entropy": 5.409546756744385,
      "epoch": 2.7471219183467177,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042564785000748814,
      "loss": 5.0536,
      "mean_token_accuracy": 0.21017140299081802,
      "num_tokens": 54612977.0,
      "step": 24460
    },
    {
      "entropy": 5.317591571807862,
      "epoch": 2.747683495254675,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004256180263214687,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2225753292441368,
      "num_tokens": 54623597.0,
      "step": 24465
    },
    {
      "entropy": 5.222692251205444,
      "epoch": 2.7482450721626326,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042558819783971295,
      "loss": 4.9395,
      "mean_token_accuracy": 0.2249891936779022,
      "num_tokens": 54634618.0,
      "step": 24470
    },
    {
      "entropy": 5.311006784439087,
      "epoch": 2.7488066490705902,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004255583645631705,
      "loss": 5.0555,
      "mean_token_accuracy": 0.2116137519478798,
      "num_tokens": 54645291.0,
      "step": 24475
    },
    {
      "entropy": 5.358775806427002,
      "epoch": 2.749368225978548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004255285264927913,
      "loss": 5.0959,
      "mean_token_accuracy": 0.20713599175214767,
      "num_tokens": 54657294.0,
      "step": 24480
    },
    {
      "entropy": 5.328173017501831,
      "epoch": 2.7499298028865056,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042549868362952556,
      "loss": 5.023,
      "mean_token_accuracy": 0.20414148569107055,
      "num_tokens": 54668627.0,
      "step": 24485
    },
    {
      "entropy": 5.330658721923828,
      "epoch": 2.750491379794463,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042546883597432327,
      "loss": 5.0325,
      "mean_token_accuracy": 0.20852886885404587,
      "num_tokens": 54679763.0,
      "step": 24490
    },
    {
      "entropy": 5.3544416427612305,
      "epoch": 2.7510529567024204,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00042543898352813485,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2123015969991684,
      "num_tokens": 54690476.0,
      "step": 24495
    },
    {
      "entropy": 5.357709217071533,
      "epoch": 2.751614533610378,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000425409126291911,
      "loss": 5.0639,
      "mean_token_accuracy": 0.21063308864831926,
      "num_tokens": 54701383.0,
      "step": 24500
    },
    {
      "entropy": 5.326432085037231,
      "epoch": 2.7521761105183353,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004253792642666021,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21831102818250656,
      "num_tokens": 54712268.0,
      "step": 24505
    },
    {
      "entropy": 5.276725435256958,
      "epoch": 2.752737687426293,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042534939745315914,
      "loss": 4.968,
      "mean_token_accuracy": 0.2131206661462784,
      "num_tokens": 54722834.0,
      "step": 24510
    },
    {
      "entropy": 5.327757549285889,
      "epoch": 2.7532992643342507,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000425319525852533,
      "loss": 5.0094,
      "mean_token_accuracy": 0.21324769854545594,
      "num_tokens": 54734628.0,
      "step": 24515
    },
    {
      "entropy": 5.185222101211548,
      "epoch": 2.753860841242208,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004252896494656749,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21986861526966095,
      "num_tokens": 54745927.0,
      "step": 24520
    },
    {
      "entropy": 5.246461200714111,
      "epoch": 2.7544224181501655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004252597682935362,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2143947422504425,
      "num_tokens": 54757822.0,
      "step": 24525
    },
    {
      "entropy": 5.28954062461853,
      "epoch": 2.754983995058123,
      "grad_norm": 1.046875,
      "learning_rate": 0.000425229882337068,
      "loss": 5.0237,
      "mean_token_accuracy": 0.2164092928171158,
      "num_tokens": 54768964.0,
      "step": 24530
    },
    {
      "entropy": 5.313607311248779,
      "epoch": 2.755545571966081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00042519999159722216,
      "loss": 5.0055,
      "mean_token_accuracy": 0.2093656361103058,
      "num_tokens": 54779471.0,
      "step": 24535
    },
    {
      "entropy": 5.301028442382813,
      "epoch": 2.7561071488740385,
      "grad_norm": 1.15625,
      "learning_rate": 0.00042517009607495035,
      "loss": 4.9651,
      "mean_token_accuracy": 0.21065238863229752,
      "num_tokens": 54790426.0,
      "step": 24540
    },
    {
      "entropy": 5.252620363235474,
      "epoch": 2.7566687257819957,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042514019577120444,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21448414772748947,
      "num_tokens": 54801765.0,
      "step": 24545
    },
    {
      "entropy": 5.325947570800781,
      "epoch": 2.7572303026899534,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004251102906869364,
      "loss": 5.0095,
      "mean_token_accuracy": 0.20403648018836976,
      "num_tokens": 54814272.0,
      "step": 24550
    },
    {
      "entropy": 5.331180191040039,
      "epoch": 2.757791879597911,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004250803808230986,
      "loss": 5.0623,
      "mean_token_accuracy": 0.21213953346014022,
      "num_tokens": 54825610.0,
      "step": 24555
    },
    {
      "entropy": 5.4249303340911865,
      "epoch": 2.7583534565058683,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042505046618064314,
      "loss": 5.038,
      "mean_token_accuracy": 0.21582036167383195,
      "num_tokens": 54835546.0,
      "step": 24560
    },
    {
      "entropy": 5.3595630645751955,
      "epoch": 2.758915033413826,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00042502054676052273,
      "loss": 5.0829,
      "mean_token_accuracy": 0.21065393537282945,
      "num_tokens": 54845653.0,
      "step": 24565
    },
    {
      "entropy": 5.325079250335693,
      "epoch": 2.7594766103217836,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042499062256368987,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21290736943483352,
      "num_tokens": 54855514.0,
      "step": 24570
    },
    {
      "entropy": 5.316353368759155,
      "epoch": 2.7600381872297413,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004249606935910975,
      "loss": 5.0584,
      "mean_token_accuracy": 0.2105695441365242,
      "num_tokens": 54866131.0,
      "step": 24575
    },
    {
      "entropy": 5.2915702819824215,
      "epoch": 2.760599764137699,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004249307598436984,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22027980387210847,
      "num_tokens": 54876842.0,
      "step": 24580
    },
    {
      "entropy": 5.362304830551148,
      "epoch": 2.761161341045656,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004249008213224458,
      "loss": 5.132,
      "mean_token_accuracy": 0.2034139171242714,
      "num_tokens": 54888036.0,
      "step": 24585
    },
    {
      "entropy": 5.356868171691895,
      "epoch": 2.761722917953614,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042487087802829275,
      "loss": 4.9967,
      "mean_token_accuracy": 0.21266153156757356,
      "num_tokens": 54898501.0,
      "step": 24590
    },
    {
      "entropy": 5.342863082885742,
      "epoch": 2.7622844948615715,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004248409299621931,
      "loss": 4.9991,
      "mean_token_accuracy": 0.21231875717639923,
      "num_tokens": 54909054.0,
      "step": 24595
    },
    {
      "entropy": 5.347270917892456,
      "epoch": 2.7628460717695287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004248109771251,
      "loss": 5.0895,
      "mean_token_accuracy": 0.2041510060429573,
      "num_tokens": 54919572.0,
      "step": 24600
    },
    {
      "entropy": 5.426667070388794,
      "epoch": 2.7634076486774863,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042478101951796716,
      "loss": 5.0644,
      "mean_token_accuracy": 0.21054824739694594,
      "num_tokens": 54930929.0,
      "step": 24605
    },
    {
      "entropy": 5.298181009292603,
      "epoch": 2.763969225585444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004247510571417487,
      "loss": 4.9607,
      "mean_token_accuracy": 0.21439721435308456,
      "num_tokens": 54941317.0,
      "step": 24610
    },
    {
      "entropy": 5.303405618667602,
      "epoch": 2.764530802493401,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004247210899973985,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21985895037651063,
      "num_tokens": 54951203.0,
      "step": 24615
    },
    {
      "entropy": 5.308162832260132,
      "epoch": 2.765092379401359,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004246911180858707,
      "loss": 5.0364,
      "mean_token_accuracy": 0.2122126743197441,
      "num_tokens": 54961485.0,
      "step": 24620
    },
    {
      "entropy": 5.280770254135132,
      "epoch": 2.7656539563093165,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004246611414081197,
      "loss": 5.0239,
      "mean_token_accuracy": 0.20873398035764695,
      "num_tokens": 54973152.0,
      "step": 24625
    },
    {
      "entropy": 5.333557510375977,
      "epoch": 2.766215533217274,
      "grad_norm": 1.0,
      "learning_rate": 0.00042463115996509984,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21048093140125274,
      "num_tokens": 54984863.0,
      "step": 24630
    },
    {
      "entropy": 5.3518318176269535,
      "epoch": 2.766777110125232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004246011737577658,
      "loss": 5.029,
      "mean_token_accuracy": 0.20855274498462678,
      "num_tokens": 54997382.0,
      "step": 24635
    },
    {
      "entropy": 5.386217308044434,
      "epoch": 2.767338687033189,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004245711827870724,
      "loss": 5.0639,
      "mean_token_accuracy": 0.20998713374137878,
      "num_tokens": 55007686.0,
      "step": 24640
    },
    {
      "entropy": 5.330079555511475,
      "epoch": 2.7679002639411467,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042454118705397454,
      "loss": 5.1187,
      "mean_token_accuracy": 0.20726052969694136,
      "num_tokens": 55018776.0,
      "step": 24645
    },
    {
      "entropy": 5.3041757106781,
      "epoch": 2.7684618408491044,
      "grad_norm": 1.0,
      "learning_rate": 0.0004245111865594273,
      "loss": 5.0838,
      "mean_token_accuracy": 0.2061333805322647,
      "num_tokens": 55031566.0,
      "step": 24650
    },
    {
      "entropy": 5.293343877792358,
      "epoch": 2.7690234177570616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042448118130438583,
      "loss": 4.94,
      "mean_token_accuracy": 0.22153064757585525,
      "num_tokens": 55043321.0,
      "step": 24655
    },
    {
      "entropy": 5.300153732299805,
      "epoch": 2.7695849946650193,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042445117128980553,
      "loss": 4.9211,
      "mean_token_accuracy": 0.21422286778688432,
      "num_tokens": 55052655.0,
      "step": 24660
    },
    {
      "entropy": 5.30465989112854,
      "epoch": 2.770146571572977,
      "grad_norm": 1.0,
      "learning_rate": 0.00042442115651664217,
      "loss": 5.0491,
      "mean_token_accuracy": 0.2089325577020645,
      "num_tokens": 55063317.0,
      "step": 24665
    },
    {
      "entropy": 5.259474420547486,
      "epoch": 2.7707081484809346,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00042439113698585103,
      "loss": 4.9661,
      "mean_token_accuracy": 0.20934724807739258,
      "num_tokens": 55074916.0,
      "step": 24670
    },
    {
      "entropy": 5.319598960876465,
      "epoch": 2.7712697253888923,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004243611126983882,
      "loss": 4.9761,
      "mean_token_accuracy": 0.2138664573431015,
      "num_tokens": 55086456.0,
      "step": 24675
    },
    {
      "entropy": 5.4429549217224125,
      "epoch": 2.7718313022968495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042433108365520967,
      "loss": 5.2038,
      "mean_token_accuracy": 0.19819564521312713,
      "num_tokens": 55097268.0,
      "step": 24680
    },
    {
      "entropy": 5.328191947937012,
      "epoch": 2.772392879204807,
      "grad_norm": 1.109375,
      "learning_rate": 0.00042430104985727155,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21188465207815171,
      "num_tokens": 55107451.0,
      "step": 24685
    },
    {
      "entropy": 5.313832426071167,
      "epoch": 2.772954456112765,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042427101130553,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21061184406280517,
      "num_tokens": 55117781.0,
      "step": 24690
    },
    {
      "entropy": 5.40861291885376,
      "epoch": 2.773516033020722,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004242409680009416,
      "loss": 5.0791,
      "mean_token_accuracy": 0.20741066634654998,
      "num_tokens": 55129519.0,
      "step": 24695
    },
    {
      "entropy": 5.250806474685669,
      "epoch": 2.7740776099286797,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004242109199444629,
      "loss": 4.9546,
      "mean_token_accuracy": 0.21118432581424712,
      "num_tokens": 55138964.0,
      "step": 24700
    },
    {
      "entropy": 5.361993789672852,
      "epoch": 2.7746391868366374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004241808671370506,
      "loss": 5.0461,
      "mean_token_accuracy": 0.20968980640172957,
      "num_tokens": 55149953.0,
      "step": 24705
    },
    {
      "entropy": 5.3568675994873045,
      "epoch": 2.7752007637445946,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042415080957966176,
      "loss": 5.1097,
      "mean_token_accuracy": 0.20591331720352174,
      "num_tokens": 55162208.0,
      "step": 24710
    },
    {
      "entropy": 5.355731964111328,
      "epoch": 2.7757623406525522,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004241207472732531,
      "loss": 5.0244,
      "mean_token_accuracy": 0.20786578506231307,
      "num_tokens": 55174014.0,
      "step": 24715
    },
    {
      "entropy": 5.402897024154663,
      "epoch": 2.77632391756051,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0004240906802187821,
      "loss": 5.1221,
      "mean_token_accuracy": 0.19890672713518143,
      "num_tokens": 55185173.0,
      "step": 24720
    },
    {
      "entropy": 5.354981708526611,
      "epoch": 2.7768854944684676,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004240606084172061,
      "loss": 5.0942,
      "mean_token_accuracy": 0.21359160989522935,
      "num_tokens": 55198404.0,
      "step": 24725
    },
    {
      "entropy": 5.257541656494141,
      "epoch": 2.777447071376425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042403053186948237,
      "loss": 5.0065,
      "mean_token_accuracy": 0.20848805755376815,
      "num_tokens": 55209126.0,
      "step": 24730
    },
    {
      "entropy": 5.304691505432129,
      "epoch": 2.7780086482843824,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004240004505765688,
      "loss": 5.0132,
      "mean_token_accuracy": 0.2119477152824402,
      "num_tokens": 55220311.0,
      "step": 24735
    },
    {
      "entropy": 5.295843553543091,
      "epoch": 2.77857022519234,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004239703645394231,
      "loss": 4.9814,
      "mean_token_accuracy": 0.21605071425437927,
      "num_tokens": 55230525.0,
      "step": 24740
    },
    {
      "entropy": 5.432407379150391,
      "epoch": 2.7791318021002978,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004239402737590032,
      "loss": 5.1576,
      "mean_token_accuracy": 0.20329540222883224,
      "num_tokens": 55242084.0,
      "step": 24745
    },
    {
      "entropy": 5.345238828659058,
      "epoch": 2.779693379008255,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00042391017823626716,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21063298135995864,
      "num_tokens": 55253677.0,
      "step": 24750
    },
    {
      "entropy": 5.217318773269653,
      "epoch": 2.7802549559162126,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004238800779721734,
      "loss": 5.0071,
      "mean_token_accuracy": 0.21726589947938918,
      "num_tokens": 55265118.0,
      "step": 24755
    },
    {
      "entropy": 5.243190765380859,
      "epoch": 2.7808165328241703,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042384997296768007,
      "loss": 5.02,
      "mean_token_accuracy": 0.21715293675661088,
      "num_tokens": 55277119.0,
      "step": 24760
    },
    {
      "entropy": 5.381906318664551,
      "epoch": 2.781378109732128,
      "grad_norm": 1.0625,
      "learning_rate": 0.000423819863223746,
      "loss": 4.9948,
      "mean_token_accuracy": 0.21411103159189224,
      "num_tokens": 55287828.0,
      "step": 24765
    },
    {
      "entropy": 5.3967711448669435,
      "epoch": 2.7819396866400856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042378974874132967,
      "loss": 5.1395,
      "mean_token_accuracy": 0.20540174841880798,
      "num_tokens": 55299115.0,
      "step": 24770
    },
    {
      "entropy": 5.238689804077149,
      "epoch": 2.782501263548043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004237596295213901,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20580557435750962,
      "num_tokens": 55310532.0,
      "step": 24775
    },
    {
      "entropy": 5.367662191390991,
      "epoch": 2.7830628404560005,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004237295055648862,
      "loss": 5.0904,
      "mean_token_accuracy": 0.20538880676031113,
      "num_tokens": 55321127.0,
      "step": 24780
    },
    {
      "entropy": 5.462488269805908,
      "epoch": 2.783624417363958,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042369937687277716,
      "loss": 5.0439,
      "mean_token_accuracy": 0.2205798491835594,
      "num_tokens": 55331791.0,
      "step": 24785
    },
    {
      "entropy": 5.354202604293823,
      "epoch": 2.7841859942719154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004236692434460223,
      "loss": 5.0969,
      "mean_token_accuracy": 0.20510492324829102,
      "num_tokens": 55342724.0,
      "step": 24790
    },
    {
      "entropy": 5.267935562133789,
      "epoch": 2.784747571179873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004236391052855811,
      "loss": 4.9949,
      "mean_token_accuracy": 0.21175975054502488,
      "num_tokens": 55354471.0,
      "step": 24795
    },
    {
      "entropy": 5.334197616577148,
      "epoch": 2.7853091480878307,
      "grad_norm": 1.0,
      "learning_rate": 0.00042360896239241306,
      "loss": 5.0255,
      "mean_token_accuracy": 0.20884050875902177,
      "num_tokens": 55366273.0,
      "step": 24800
    },
    {
      "entropy": 5.286634683609009,
      "epoch": 2.785870724995788,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042357881476747807,
      "loss": 5.0064,
      "mean_token_accuracy": 0.21183134019374847,
      "num_tokens": 55377384.0,
      "step": 24805
    },
    {
      "entropy": 5.320109796524048,
      "epoch": 2.7864323019037456,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000423548662411736,
      "loss": 5.0261,
      "mean_token_accuracy": 0.20936318784952163,
      "num_tokens": 55387348.0,
      "step": 24810
    },
    {
      "entropy": 5.3199280261993405,
      "epoch": 2.7869938788117032,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004235185053261469,
      "loss": 4.9541,
      "mean_token_accuracy": 0.22172591090202332,
      "num_tokens": 55398484.0,
      "step": 24815
    },
    {
      "entropy": 5.2666233539581295,
      "epoch": 2.787555455719661,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000423488343511671,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21881017982959747,
      "num_tokens": 55409347.0,
      "step": 24820
    },
    {
      "entropy": 5.259686040878296,
      "epoch": 2.7881170326276186,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004234581769692686,
      "loss": 5.0312,
      "mean_token_accuracy": 0.21097774058580399,
      "num_tokens": 55421048.0,
      "step": 24825
    },
    {
      "entropy": 5.279538822174072,
      "epoch": 2.788678609535576,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004234280056999003,
      "loss": 4.9764,
      "mean_token_accuracy": 0.22023516595363618,
      "num_tokens": 55431189.0,
      "step": 24830
    },
    {
      "entropy": 5.388339948654175,
      "epoch": 2.7892401864435334,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004233978297045267,
      "loss": 5.0341,
      "mean_token_accuracy": 0.2040510892868042,
      "num_tokens": 55442695.0,
      "step": 24835
    },
    {
      "entropy": 5.426276922225952,
      "epoch": 2.789801763351491,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004233676489841087,
      "loss": 5.0552,
      "mean_token_accuracy": 0.2074578359723091,
      "num_tokens": 55454954.0,
      "step": 24840
    },
    {
      "entropy": 5.344195032119751,
      "epoch": 2.7903633402594483,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042333746353960723,
      "loss": 5.1136,
      "mean_token_accuracy": 0.2098468080163002,
      "num_tokens": 55466270.0,
      "step": 24845
    },
    {
      "entropy": 5.382609510421753,
      "epoch": 2.790924917167406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004233072733719833,
      "loss": 5.0984,
      "mean_token_accuracy": 0.2131660982966423,
      "num_tokens": 55476925.0,
      "step": 24850
    },
    {
      "entropy": 5.300843715667725,
      "epoch": 2.7914864940753636,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004232770784821984,
      "loss": 5.0014,
      "mean_token_accuracy": 0.20784088522195815,
      "num_tokens": 55489705.0,
      "step": 24855
    },
    {
      "entropy": 5.385250806808472,
      "epoch": 2.7920480709833213,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042324687887121366,
      "loss": 5.1385,
      "mean_token_accuracy": 0.2026788890361786,
      "num_tokens": 55501362.0,
      "step": 24860
    },
    {
      "entropy": 5.341915321350098,
      "epoch": 2.792609647891279,
      "grad_norm": 1.125,
      "learning_rate": 0.0004232166745399909,
      "loss": 5.1003,
      "mean_token_accuracy": 0.20874485671520232,
      "num_tokens": 55512754.0,
      "step": 24865
    },
    {
      "entropy": 5.353347444534302,
      "epoch": 2.793171224799236,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042318646548949177,
      "loss": 5.021,
      "mean_token_accuracy": 0.2045422151684761,
      "num_tokens": 55524057.0,
      "step": 24870
    },
    {
      "entropy": 5.330910634994507,
      "epoch": 2.793732801707194,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042315625172067794,
      "loss": 5.0262,
      "mean_token_accuracy": 0.21335629373788834,
      "num_tokens": 55534768.0,
      "step": 24875
    },
    {
      "entropy": 5.402321100234985,
      "epoch": 2.7942943786151515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042312603323451173,
      "loss": 5.0953,
      "mean_token_accuracy": 0.20379077643156052,
      "num_tokens": 55547135.0,
      "step": 24880
    },
    {
      "entropy": 5.358650732040405,
      "epoch": 2.7948559555231087,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0004230958100319551,
      "loss": 5.1062,
      "mean_token_accuracy": 0.20704581141471862,
      "num_tokens": 55559638.0,
      "step": 24885
    },
    {
      "entropy": 5.3304027080535885,
      "epoch": 2.7954175324310664,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004230655821139705,
      "loss": 5.021,
      "mean_token_accuracy": 0.21345882266759872,
      "num_tokens": 55570985.0,
      "step": 24890
    },
    {
      "entropy": 5.272723913192749,
      "epoch": 2.795979109339024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004230353494815203,
      "loss": 4.9954,
      "mean_token_accuracy": 0.2107615813612938,
      "num_tokens": 55581701.0,
      "step": 24895
    },
    {
      "entropy": 5.3093650341033936,
      "epoch": 2.7965406862469813,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004230051121355672,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21493848264217377,
      "num_tokens": 55592891.0,
      "step": 24900
    },
    {
      "entropy": 5.2516429901123045,
      "epoch": 2.797102263154939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042297487007707387,
      "loss": 5.0185,
      "mean_token_accuracy": 0.2216321587562561,
      "num_tokens": 55603762.0,
      "step": 24905
    },
    {
      "entropy": 5.392943620681763,
      "epoch": 2.7976638400628966,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042294462330700336,
      "loss": 5.0458,
      "mean_token_accuracy": 0.215898397564888,
      "num_tokens": 55614450.0,
      "step": 24910
    },
    {
      "entropy": 5.275924730300903,
      "epoch": 2.7982254169708543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004229143718263186,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2129613384604454,
      "num_tokens": 55625240.0,
      "step": 24915
    },
    {
      "entropy": 5.236031818389892,
      "epoch": 2.798786993878812,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004228841156359829,
      "loss": 4.968,
      "mean_token_accuracy": 0.2195337727665901,
      "num_tokens": 55635646.0,
      "step": 24920
    },
    {
      "entropy": 5.306163167953491,
      "epoch": 2.799348570786769,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004228538547369596,
      "loss": 4.9264,
      "mean_token_accuracy": 0.21893587410449983,
      "num_tokens": 55646669.0,
      "step": 24925
    },
    {
      "entropy": 5.4360980033874515,
      "epoch": 2.799910147694727,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042282358913021224,
      "loss": 5.2598,
      "mean_token_accuracy": 0.201321479678154,
      "num_tokens": 55658792.0,
      "step": 24930
    },
    {
      "entropy": 5.321544170379639,
      "epoch": 2.8004717246026845,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042279331881670446,
      "loss": 5.0516,
      "mean_token_accuracy": 0.21301648020744324,
      "num_tokens": 55670037.0,
      "step": 24935
    },
    {
      "entropy": 5.346176719665527,
      "epoch": 2.8010333015106417,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004227630437974,
      "loss": 5.0572,
      "mean_token_accuracy": 0.20424896776676177,
      "num_tokens": 55680405.0,
      "step": 24940
    },
    {
      "entropy": 5.35882978439331,
      "epoch": 2.8015948784185993,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042273276407326304,
      "loss": 5.0087,
      "mean_token_accuracy": 0.2113076329231262,
      "num_tokens": 55692199.0,
      "step": 24945
    },
    {
      "entropy": 5.285901737213135,
      "epoch": 2.802156455326557,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00042270247964525755,
      "loss": 4.9866,
      "mean_token_accuracy": 0.21307801306247712,
      "num_tokens": 55702654.0,
      "step": 24950
    },
    {
      "entropy": 5.30672173500061,
      "epoch": 2.8027180322345147,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004226721905143477,
      "loss": 5.0244,
      "mean_token_accuracy": 0.21592682600021362,
      "num_tokens": 55712710.0,
      "step": 24955
    },
    {
      "entropy": 5.35311484336853,
      "epoch": 2.8032796091424723,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00042264189668149817,
      "loss": 5.0185,
      "mean_token_accuracy": 0.20870975255966187,
      "num_tokens": 55723429.0,
      "step": 24960
    },
    {
      "entropy": 5.310142517089844,
      "epoch": 2.8038411860504295,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004226115981476733,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21471438705921173,
      "num_tokens": 55735113.0,
      "step": 24965
    },
    {
      "entropy": 5.194744682312011,
      "epoch": 2.804402762958387,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000422581294913838,
      "loss": 4.8793,
      "mean_token_accuracy": 0.21801006495952607,
      "num_tokens": 55745857.0,
      "step": 24970
    },
    {
      "entropy": 5.283068227767944,
      "epoch": 2.804964339866345,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000422550986980957,
      "loss": 4.9289,
      "mean_token_accuracy": 0.21277132481336594,
      "num_tokens": 55756856.0,
      "step": 24975
    },
    {
      "entropy": 5.374700450897217,
      "epoch": 2.805525916774302,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042252067434999524,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20436871200799941,
      "num_tokens": 55768827.0,
      "step": 24980
    },
    {
      "entropy": 5.312294769287109,
      "epoch": 2.8060874936822597,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004224903570219181,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2071317195892334,
      "num_tokens": 55780360.0,
      "step": 24985
    },
    {
      "entropy": 5.32460241317749,
      "epoch": 2.8066490705902174,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042246003499769077,
      "loss": 5.0415,
      "mean_token_accuracy": 0.2087404429912567,
      "num_tokens": 55791874.0,
      "step": 24990
    },
    {
      "entropy": 5.309207010269165,
      "epoch": 2.8072106474981746,
      "grad_norm": 0.953125,
      "learning_rate": 0.00042242970827827874,
      "loss": 5.0135,
      "mean_token_accuracy": 0.21685759127140045,
      "num_tokens": 55802689.0,
      "step": 24995
    },
    {
      "entropy": 5.313310050964356,
      "epoch": 2.8077722244061323,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004223993768646476,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21302051842212677,
      "num_tokens": 55813945.0,
      "step": 25000
    },
    {
      "entropy": 5.301851844787597,
      "epoch": 2.80833380131409,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004223690407577631,
      "loss": 4.9916,
      "mean_token_accuracy": 0.21230896562337875,
      "num_tokens": 55826403.0,
      "step": 25005
    },
    {
      "entropy": 5.245838594436646,
      "epoch": 2.8088953782220476,
      "grad_norm": 0.984375,
      "learning_rate": 0.00042233869995859127,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21785952597856523,
      "num_tokens": 55836858.0,
      "step": 25010
    },
    {
      "entropy": 5.339810228347778,
      "epoch": 2.8094569551300053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042230835446809803,
      "loss": 4.9883,
      "mean_token_accuracy": 0.2197006016969681,
      "num_tokens": 55847664.0,
      "step": 25015
    },
    {
      "entropy": 5.266009902954101,
      "epoch": 2.8100185320379625,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004222780042872496,
      "loss": 4.9781,
      "mean_token_accuracy": 0.21395136266946793,
      "num_tokens": 55858333.0,
      "step": 25020
    },
    {
      "entropy": 5.275211143493652,
      "epoch": 2.81058010894592,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004222476494170125,
      "loss": 4.9886,
      "mean_token_accuracy": 0.2153901293873787,
      "num_tokens": 55868819.0,
      "step": 25025
    },
    {
      "entropy": 5.296229696273803,
      "epoch": 2.811141685853878,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004222172898583531,
      "loss": 4.9738,
      "mean_token_accuracy": 0.21946696192026138,
      "num_tokens": 55879133.0,
      "step": 25030
    },
    {
      "entropy": 5.284844636917114,
      "epoch": 2.811703262761835,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004221869256122381,
      "loss": 4.9521,
      "mean_token_accuracy": 0.21601525098085403,
      "num_tokens": 55890362.0,
      "step": 25035
    },
    {
      "entropy": 5.314735794067383,
      "epoch": 2.8122648396697927,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004221565566796343,
      "loss": 5.0012,
      "mean_token_accuracy": 0.2110033929347992,
      "num_tokens": 55899920.0,
      "step": 25040
    },
    {
      "entropy": 5.279232120513916,
      "epoch": 2.8128264165777503,
      "grad_norm": 1.0,
      "learning_rate": 0.0004221261830615086,
      "loss": 5.0264,
      "mean_token_accuracy": 0.21454062312841415,
      "num_tokens": 55911598.0,
      "step": 25045
    },
    {
      "entropy": 5.206222248077393,
      "epoch": 2.813387993485708,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004220958047588282,
      "loss": 4.9321,
      "mean_token_accuracy": 0.22180513739585878,
      "num_tokens": 55922514.0,
      "step": 25050
    },
    {
      "entropy": 5.31848292350769,
      "epoch": 2.8139495703936657,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004220654217725603,
      "loss": 4.9642,
      "mean_token_accuracy": 0.22169241160154343,
      "num_tokens": 55932796.0,
      "step": 25055
    },
    {
      "entropy": 5.338256692886352,
      "epoch": 2.814511147301623,
      "grad_norm": 1.140625,
      "learning_rate": 0.00042203503410367244,
      "loss": 5.0514,
      "mean_token_accuracy": 0.21228789687156677,
      "num_tokens": 55944337.0,
      "step": 25060
    },
    {
      "entropy": 5.404289054870605,
      "epoch": 2.8150727242095805,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000422004641753132,
      "loss": 5.1624,
      "mean_token_accuracy": 0.2018483802676201,
      "num_tokens": 55955871.0,
      "step": 25065
    },
    {
      "entropy": 5.40138807296753,
      "epoch": 2.815634301117538,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004219742447219068,
      "loss": 5.0475,
      "mean_token_accuracy": 0.2087547719478607,
      "num_tokens": 55966091.0,
      "step": 25070
    },
    {
      "entropy": 5.3155660152435305,
      "epoch": 2.8161958780254954,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00042194384301096457,
      "loss": 4.994,
      "mean_token_accuracy": 0.2160571426153183,
      "num_tokens": 55976755.0,
      "step": 25075
    },
    {
      "entropy": 5.294264078140259,
      "epoch": 2.816757454933453,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042191343662127337,
      "loss": 5.0774,
      "mean_token_accuracy": 0.21273763924837114,
      "num_tokens": 55987579.0,
      "step": 25080
    },
    {
      "entropy": 5.350923013687134,
      "epoch": 2.8173190318414107,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042188302555380144,
      "loss": 5.0152,
      "mean_token_accuracy": 0.20944810807704925,
      "num_tokens": 55997027.0,
      "step": 25085
    },
    {
      "entropy": 5.326024866104126,
      "epoch": 2.817880608749368,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000421852609809517,
      "loss": 5.0236,
      "mean_token_accuracy": 0.2085127964615822,
      "num_tokens": 56008928.0,
      "step": 25090
    },
    {
      "entropy": 5.308360433578491,
      "epoch": 2.8184421856573256,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004218221893893884,
      "loss": 5.0257,
      "mean_token_accuracy": 0.21272192448377608,
      "num_tokens": 56021313.0,
      "step": 25095
    },
    {
      "entropy": 5.352844524383545,
      "epoch": 2.8190037625652833,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042179176429438454,
      "loss": 5.0108,
      "mean_token_accuracy": 0.21170289516448976,
      "num_tokens": 56032002.0,
      "step": 25100
    },
    {
      "entropy": 5.341112089157105,
      "epoch": 2.819565339473241,
      "grad_norm": 1.0,
      "learning_rate": 0.0004217613345254739,
      "loss": 4.9806,
      "mean_token_accuracy": 0.2232983261346817,
      "num_tokens": 56042905.0,
      "step": 25105
    },
    {
      "entropy": 5.303177976608277,
      "epoch": 2.8201269163811986,
      "grad_norm": 1.09375,
      "learning_rate": 0.00042173090008362535,
      "loss": 4.967,
      "mean_token_accuracy": 0.21113592982292176,
      "num_tokens": 56052507.0,
      "step": 25110
    },
    {
      "entropy": 5.230526494979858,
      "epoch": 2.820688493289156,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042170046096980807,
      "loss": 5.0558,
      "mean_token_accuracy": 0.20810750424861907,
      "num_tokens": 56065250.0,
      "step": 25115
    },
    {
      "entropy": 5.311351108551025,
      "epoch": 2.8212500701971135,
      "grad_norm": 1.0,
      "learning_rate": 0.00042167001718499124,
      "loss": 4.9289,
      "mean_token_accuracy": 0.2132183164358139,
      "num_tokens": 56075858.0,
      "step": 25120
    },
    {
      "entropy": 5.360123825073242,
      "epoch": 2.821811647105071,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004216395687301442,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20865718722343446,
      "num_tokens": 56086647.0,
      "step": 25125
    },
    {
      "entropy": 5.347607755661011,
      "epoch": 2.8223732240130284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004216091156062364,
      "loss": 5.2279,
      "mean_token_accuracy": 0.19683048278093337,
      "num_tokens": 56098124.0,
      "step": 25130
    },
    {
      "entropy": 5.344070148468018,
      "epoch": 2.822934800920986,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004215786578142375,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21030884087085724,
      "num_tokens": 56108505.0,
      "step": 25135
    },
    {
      "entropy": 5.314936017990112,
      "epoch": 2.8234963778289437,
      "grad_norm": 1.0625,
      "learning_rate": 0.00042154819535511733,
      "loss": 4.9822,
      "mean_token_accuracy": 0.21955758482217788,
      "num_tokens": 56118423.0,
      "step": 25140
    },
    {
      "entropy": 5.3612419128417965,
      "epoch": 2.8240579547369014,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042151772822984567,
      "loss": 5.0356,
      "mean_token_accuracy": 0.21584475338459014,
      "num_tokens": 56130346.0,
      "step": 25145
    },
    {
      "entropy": 5.298059320449829,
      "epoch": 2.824619531644859,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004214872564393928,
      "loss": 5.0689,
      "mean_token_accuracy": 0.2154630944132805,
      "num_tokens": 56141044.0,
      "step": 25150
    },
    {
      "entropy": 5.278464603424072,
      "epoch": 2.8251811085528162,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004214567799847289,
      "loss": 4.9537,
      "mean_token_accuracy": 0.2204898178577423,
      "num_tokens": 56151786.0,
      "step": 25155
    },
    {
      "entropy": 5.355052423477173,
      "epoch": 2.825742685460774,
      "grad_norm": 1.0,
      "learning_rate": 0.0004214262988668242,
      "loss": 5.0022,
      "mean_token_accuracy": 0.2083557739853859,
      "num_tokens": 56164196.0,
      "step": 25160
    },
    {
      "entropy": 5.320587491989135,
      "epoch": 2.8263042623687316,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004213958130866494,
      "loss": 5.047,
      "mean_token_accuracy": 0.21349860727787018,
      "num_tokens": 56176259.0,
      "step": 25165
    },
    {
      "entropy": 5.293961429595948,
      "epoch": 2.8268658392766888,
      "grad_norm": 1.03125,
      "learning_rate": 0.00042136532264517527,
      "loss": 5.0499,
      "mean_token_accuracy": 0.20524821430444717,
      "num_tokens": 56186617.0,
      "step": 25170
    },
    {
      "entropy": 5.21634578704834,
      "epoch": 2.8274274161846464,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042133482754337236,
      "loss": 4.9071,
      "mean_token_accuracy": 0.21691754311323166,
      "num_tokens": 56197609.0,
      "step": 25175
    },
    {
      "entropy": 5.264827823638916,
      "epoch": 2.827988993092604,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042130432778221186,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21833383589982985,
      "num_tokens": 56207490.0,
      "step": 25180
    },
    {
      "entropy": 5.286197185516357,
      "epoch": 2.8285505700005613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042127382336266467,
      "loss": 4.9957,
      "mean_token_accuracy": 0.21431188732385636,
      "num_tokens": 56218167.0,
      "step": 25185
    },
    {
      "entropy": 5.224011087417603,
      "epoch": 2.829112146908519,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004212433142857024,
      "loss": 4.8965,
      "mean_token_accuracy": 0.2187386840581894,
      "num_tokens": 56229199.0,
      "step": 25190
    },
    {
      "entropy": 5.346891927719116,
      "epoch": 2.8296737238164766,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042121280055229624,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20838897973299025,
      "num_tokens": 56240165.0,
      "step": 25195
    },
    {
      "entropy": 5.4392595291137695,
      "epoch": 2.8302353007244343,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004211822821634177,
      "loss": 5.1803,
      "mean_token_accuracy": 0.2025257647037506,
      "num_tokens": 56252447.0,
      "step": 25200
    },
    {
      "entropy": 5.360434436798096,
      "epoch": 2.830796877632392,
      "grad_norm": 1.0,
      "learning_rate": 0.0004211517591200387,
      "loss": 5.0256,
      "mean_token_accuracy": 0.216510571539402,
      "num_tokens": 56263116.0,
      "step": 25205
    },
    {
      "entropy": 5.322058296203613,
      "epoch": 2.831358454540349,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0004211212314231309,
      "loss": 5.0759,
      "mean_token_accuracy": 0.20728585571050645,
      "num_tokens": 56276257.0,
      "step": 25210
    },
    {
      "entropy": 5.2707000255584715,
      "epoch": 2.831920031448307,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042109069907366646,
      "loss": 4.9652,
      "mean_token_accuracy": 0.2123456507921219,
      "num_tokens": 56287865.0,
      "step": 25215
    },
    {
      "entropy": 5.32804856300354,
      "epoch": 2.8324816083562645,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042106016207261754,
      "loss": 4.9824,
      "mean_token_accuracy": 0.2173628032207489,
      "num_tokens": 56298122.0,
      "step": 25220
    },
    {
      "entropy": 5.3748664379119875,
      "epoch": 2.8330431852642217,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042102962042095643,
      "loss": 5.0973,
      "mean_token_accuracy": 0.21238444596529008,
      "num_tokens": 56309455.0,
      "step": 25225
    },
    {
      "entropy": 5.3704486846923825,
      "epoch": 2.8336047621721794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042099907411965553,
      "loss": 5.0612,
      "mean_token_accuracy": 0.20921215713024138,
      "num_tokens": 56319804.0,
      "step": 25230
    },
    {
      "entropy": 5.289084339141846,
      "epoch": 2.834166339080137,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042096852316968756,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22174431383609772,
      "num_tokens": 56330136.0,
      "step": 25235
    },
    {
      "entropy": 5.4456909656524655,
      "epoch": 2.8347279159880947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042093796757202513,
      "loss": 5.1934,
      "mean_token_accuracy": 0.20083945542573928,
      "num_tokens": 56341861.0,
      "step": 25240
    },
    {
      "entropy": 5.2559717178344725,
      "epoch": 2.8352894928960524,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004209074073276412,
      "loss": 5.0366,
      "mean_token_accuracy": 0.2102880746126175,
      "num_tokens": 56352016.0,
      "step": 25245
    },
    {
      "entropy": 5.233224582672119,
      "epoch": 2.8358510698040096,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004208768424375089,
      "loss": 4.9312,
      "mean_token_accuracy": 0.2155200645327568,
      "num_tokens": 56363740.0,
      "step": 25250
    },
    {
      "entropy": 5.350187158584594,
      "epoch": 2.8364126467119672,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042084627290260124,
      "loss": 4.958,
      "mean_token_accuracy": 0.21626657694578172,
      "num_tokens": 56375080.0,
      "step": 25255
    },
    {
      "entropy": 5.318953323364258,
      "epoch": 2.836974223619925,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004208156987238918,
      "loss": 5.0881,
      "mean_token_accuracy": 0.20558465421199798,
      "num_tokens": 56385858.0,
      "step": 25260
    },
    {
      "entropy": 5.359691715240478,
      "epoch": 2.837535800527882,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004207851199023539,
      "loss": 5.0363,
      "mean_token_accuracy": 0.21353203654289246,
      "num_tokens": 56394423.0,
      "step": 25265
    },
    {
      "entropy": 5.33716721534729,
      "epoch": 2.83809737743584,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042075453643896125,
      "loss": 5.0231,
      "mean_token_accuracy": 0.2080884248018265,
      "num_tokens": 56405843.0,
      "step": 25270
    },
    {
      "entropy": 5.319983577728271,
      "epoch": 2.8386589543437974,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00042072394833468765,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2082054540514946,
      "num_tokens": 56417757.0,
      "step": 25275
    },
    {
      "entropy": 5.297622871398926,
      "epoch": 2.8392205312517547,
      "grad_norm": 1.015625,
      "learning_rate": 0.000420693355590507,
      "loss": 5.0426,
      "mean_token_accuracy": 0.20967078059911728,
      "num_tokens": 56429937.0,
      "step": 25280
    },
    {
      "entropy": 5.252430009841919,
      "epoch": 2.8397821081597123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00042066275820739333,
      "loss": 5.0059,
      "mean_token_accuracy": 0.21328503489494324,
      "num_tokens": 56441114.0,
      "step": 25285
    },
    {
      "entropy": 5.26390118598938,
      "epoch": 2.84034368506767,
      "grad_norm": 1.078125,
      "learning_rate": 0.000420632156186321,
      "loss": 5.0009,
      "mean_token_accuracy": 0.20950732827186586,
      "num_tokens": 56451839.0,
      "step": 25290
    },
    {
      "entropy": 5.3275792598724365,
      "epoch": 2.8409052619756276,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004206015495282644,
      "loss": 5.0516,
      "mean_token_accuracy": 0.20955644696950912,
      "num_tokens": 56463384.0,
      "step": 25295
    },
    {
      "entropy": 5.3753650188446045,
      "epoch": 2.8414668388835853,
      "grad_norm": 1.015625,
      "learning_rate": 0.00042057093823419795,
      "loss": 4.9951,
      "mean_token_accuracy": 0.20609987080097197,
      "num_tokens": 56474905.0,
      "step": 25300
    },
    {
      "entropy": 5.316319942474365,
      "epoch": 2.8420284157915425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042054032230509636,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21159250885248185,
      "num_tokens": 56485319.0,
      "step": 25305
    },
    {
      "entropy": 5.276068687438965,
      "epoch": 2.8425899926995,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00042050970174193437,
      "loss": 5.0966,
      "mean_token_accuracy": 0.20598801523447036,
      "num_tokens": 56496064.0,
      "step": 25310
    },
    {
      "entropy": 5.279910516738892,
      "epoch": 2.843151569607458,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00042047907654568724,
      "loss": 5.0201,
      "mean_token_accuracy": 0.20724644362926484,
      "num_tokens": 56507167.0,
      "step": 25315
    },
    {
      "entropy": 5.386574602127075,
      "epoch": 2.843713146515415,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004204484467173297,
      "loss": 5.1125,
      "mean_token_accuracy": 0.21376283019781112,
      "num_tokens": 56518400.0,
      "step": 25320
    },
    {
      "entropy": 5.392175579071045,
      "epoch": 2.8442747234233727,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004204178122578373,
      "loss": 5.1329,
      "mean_token_accuracy": 0.20758781731128692,
      "num_tokens": 56529630.0,
      "step": 25325
    },
    {
      "entropy": 5.403842639923096,
      "epoch": 2.8448363003313304,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042038717316818534,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21114799678325652,
      "num_tokens": 56539520.0,
      "step": 25330
    },
    {
      "entropy": 5.33147783279419,
      "epoch": 2.845397877239288,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004203565294493494,
      "loss": 5.0016,
      "mean_token_accuracy": 0.214001801609993,
      "num_tokens": 56550677.0,
      "step": 25335
    },
    {
      "entropy": 5.291436243057251,
      "epoch": 2.8459594541472457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00042032588110230525,
      "loss": 5.0933,
      "mean_token_accuracy": 0.2076392501592636,
      "num_tokens": 56561694.0,
      "step": 25340
    },
    {
      "entropy": 5.354751777648926,
      "epoch": 2.846521031055203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004202952281280286,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20697802752256395,
      "num_tokens": 56572322.0,
      "step": 25345
    },
    {
      "entropy": 5.340733242034912,
      "epoch": 2.8470826079631606,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004202645705274956,
      "loss": 5.0471,
      "mean_token_accuracy": 0.2051721528172493,
      "num_tokens": 56583198.0,
      "step": 25350
    },
    {
      "entropy": 5.330936479568481,
      "epoch": 2.8476441848711183,
      "grad_norm": 1.046875,
      "learning_rate": 0.00042023390830168224,
      "loss": 5.0969,
      "mean_token_accuracy": 0.2106487661600113,
      "num_tokens": 56594732.0,
      "step": 25355
    },
    {
      "entropy": 5.384609794616699,
      "epoch": 2.8482057617790755,
      "grad_norm": 1.015625,
      "learning_rate": 0.000420203241451565,
      "loss": 5.0528,
      "mean_token_accuracy": 0.21501615047454833,
      "num_tokens": 56605907.0,
      "step": 25360
    },
    {
      "entropy": 5.407325887680054,
      "epoch": 2.848767338687033,
      "grad_norm": 1.078125,
      "learning_rate": 0.00042017256997812016,
      "loss": 5.1159,
      "mean_token_accuracy": 0.2133430853486061,
      "num_tokens": 56616841.0,
      "step": 25365
    },
    {
      "entropy": 5.358513212203979,
      "epoch": 2.849328915594991,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00042014189388232446,
      "loss": 5.0471,
      "mean_token_accuracy": 0.20450548380613326,
      "num_tokens": 56628044.0,
      "step": 25370
    },
    {
      "entropy": 5.344708871841431,
      "epoch": 2.849890492502948,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004201112131651545,
      "loss": 5.0667,
      "mean_token_accuracy": 0.206785649061203,
      "num_tokens": 56638350.0,
      "step": 25375
    },
    {
      "entropy": 5.344955015182495,
      "epoch": 2.8504520694109057,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004200805278275873,
      "loss": 5.0071,
      "mean_token_accuracy": 0.20835406929254532,
      "num_tokens": 56649923.0,
      "step": 25380
    },
    {
      "entropy": 5.2104727268219,
      "epoch": 2.8510136463188633,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004200498378705997,
      "loss": 4.9243,
      "mean_token_accuracy": 0.21556854993104935,
      "num_tokens": 56661210.0,
      "step": 25385
    },
    {
      "entropy": 5.3639427661895756,
      "epoch": 2.851575223226821,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004200191432951692,
      "loss": 5.1284,
      "mean_token_accuracy": 0.20294114500284194,
      "num_tokens": 56673106.0,
      "step": 25390
    },
    {
      "entropy": 5.30151252746582,
      "epoch": 2.8521368001347787,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004199884441022728,
      "loss": 5.0426,
      "mean_token_accuracy": 0.21145144402980803,
      "num_tokens": 56684703.0,
      "step": 25395
    },
    {
      "entropy": 5.370098638534546,
      "epoch": 2.852698377042736,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041995774029288807,
      "loss": 5.0266,
      "mean_token_accuracy": 0.2156658485531807,
      "num_tokens": 56696881.0,
      "step": 25400
    },
    {
      "entropy": 5.304983711242675,
      "epoch": 2.8532599539506935,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041992703186799273,
      "loss": 4.9824,
      "mean_token_accuracy": 0.21893545240163803,
      "num_tokens": 56707372.0,
      "step": 25405
    },
    {
      "entropy": 5.275336074829101,
      "epoch": 2.853821530858651,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041989631882856447,
      "loss": 4.9258,
      "mean_token_accuracy": 0.21718777120113372,
      "num_tokens": 56717954.0,
      "step": 25410
    },
    {
      "entropy": 5.325252389907837,
      "epoch": 2.8543831077666084,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004198656011755812,
      "loss": 5.0465,
      "mean_token_accuracy": 0.21063360571861267,
      "num_tokens": 56728798.0,
      "step": 25415
    },
    {
      "entropy": 5.304375076293946,
      "epoch": 2.854944684674566,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004198348789100211,
      "loss": 5.0162,
      "mean_token_accuracy": 0.211671644449234,
      "num_tokens": 56740095.0,
      "step": 25420
    },
    {
      "entropy": 5.256620931625366,
      "epoch": 2.8555062615825237,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041980415203286223,
      "loss": 4.974,
      "mean_token_accuracy": 0.20669995248317719,
      "num_tokens": 56751002.0,
      "step": 25425
    },
    {
      "entropy": 5.339339017868042,
      "epoch": 2.8560678384904814,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000419773420545083,
      "loss": 4.9729,
      "mean_token_accuracy": 0.21696695387363435,
      "num_tokens": 56762244.0,
      "step": 25430
    },
    {
      "entropy": 5.338521957397461,
      "epoch": 2.856629415398439,
      "grad_norm": 1.0,
      "learning_rate": 0.00041974268444766187,
      "loss": 5.0419,
      "mean_token_accuracy": 0.21011248081922532,
      "num_tokens": 56772918.0,
      "step": 25435
    },
    {
      "entropy": 5.218806648254395,
      "epoch": 2.8571909923063963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004197119437415776,
      "loss": 4.9241,
      "mean_token_accuracy": 0.2211347848176956,
      "num_tokens": 56784069.0,
      "step": 25440
    },
    {
      "entropy": 5.266423559188842,
      "epoch": 2.857752569214354,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004196811984278089,
      "loss": 4.9203,
      "mean_token_accuracy": 0.21744282245635987,
      "num_tokens": 56795187.0,
      "step": 25445
    },
    {
      "entropy": 5.306480550765992,
      "epoch": 2.8583141461223116,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004196504485073347,
      "loss": 5.0115,
      "mean_token_accuracy": 0.21071134954690934,
      "num_tokens": 56806766.0,
      "step": 25450
    },
    {
      "entropy": 5.326224946975708,
      "epoch": 2.858875723030269,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041961969398113404,
      "loss": 5.0722,
      "mean_token_accuracy": 0.2083122655749321,
      "num_tokens": 56817872.0,
      "step": 25455
    },
    {
      "entropy": 5.396213245391846,
      "epoch": 2.8594372999382265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041958893485018634,
      "loss": 5.1275,
      "mean_token_accuracy": 0.20319733172655105,
      "num_tokens": 56828291.0,
      "step": 25460
    },
    {
      "entropy": 5.316315698623657,
      "epoch": 2.859998876846184,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004195581711154709,
      "loss": 4.952,
      "mean_token_accuracy": 0.2176439568400383,
      "num_tokens": 56839165.0,
      "step": 25465
    },
    {
      "entropy": 5.215438365936279,
      "epoch": 2.8605604537541414,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041952740277796717,
      "loss": 4.934,
      "mean_token_accuracy": 0.22274965941905975,
      "num_tokens": 56850754.0,
      "step": 25470
    },
    {
      "entropy": 5.223601484298706,
      "epoch": 2.861122030662099,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041949662983865486,
      "loss": 5.0246,
      "mean_token_accuracy": 0.21332585960626602,
      "num_tokens": 56862798.0,
      "step": 25475
    },
    {
      "entropy": 5.2471240043640135,
      "epoch": 2.8616836075700567,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004194658522985139,
      "loss": 4.951,
      "mean_token_accuracy": 0.21707855463027953,
      "num_tokens": 56873479.0,
      "step": 25480
    },
    {
      "entropy": 5.340463066101075,
      "epoch": 2.8622451844780143,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004194350701585241,
      "loss": 4.9728,
      "mean_token_accuracy": 0.2148469030857086,
      "num_tokens": 56883924.0,
      "step": 25485
    },
    {
      "entropy": 5.33103232383728,
      "epoch": 2.862806761385972,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0004194042834196656,
      "loss": 5.0271,
      "mean_token_accuracy": 0.2156791567802429,
      "num_tokens": 56895622.0,
      "step": 25490
    },
    {
      "entropy": 5.284690189361572,
      "epoch": 2.8633683382939292,
      "grad_norm": 1.0,
      "learning_rate": 0.0004193734920829188,
      "loss": 4.9748,
      "mean_token_accuracy": 0.2066233977675438,
      "num_tokens": 56906504.0,
      "step": 25495
    },
    {
      "entropy": 5.23050537109375,
      "epoch": 2.863929915201887,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000419342696149264,
      "loss": 4.9891,
      "mean_token_accuracy": 0.21351407319307328,
      "num_tokens": 56917228.0,
      "step": 25500
    },
    {
      "entropy": 5.316056919097901,
      "epoch": 2.8644914921098446,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004193118956196818,
      "loss": 4.9999,
      "mean_token_accuracy": 0.2116966187953949,
      "num_tokens": 56927981.0,
      "step": 25505
    },
    {
      "entropy": 5.311671304702759,
      "epoch": 2.8650530690178018,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004192810904951528,
      "loss": 5.0019,
      "mean_token_accuracy": 0.20933346301317216,
      "num_tokens": 56938552.0,
      "step": 25510
    },
    {
      "entropy": 5.279219627380371,
      "epoch": 2.8656146459257594,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000419250280776658,
      "loss": 4.9311,
      "mean_token_accuracy": 0.21908278465270997,
      "num_tokens": 56949902.0,
      "step": 25515
    },
    {
      "entropy": 5.370410203933716,
      "epoch": 2.866176222833717,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004192194664651783,
      "loss": 5.1243,
      "mean_token_accuracy": 0.2105543076992035,
      "num_tokens": 56961335.0,
      "step": 25520
    },
    {
      "entropy": 5.2483453273773195,
      "epoch": 2.8667377997416748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004191886475616947,
      "loss": 4.9648,
      "mean_token_accuracy": 0.21825129240751268,
      "num_tokens": 56972542.0,
      "step": 25525
    },
    {
      "entropy": 5.308157968521118,
      "epoch": 2.8672993766496324,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041915782406718875,
      "loss": 5.0329,
      "mean_token_accuracy": 0.2143852725625038,
      "num_tokens": 56983838.0,
      "step": 25530
    },
    {
      "entropy": 5.372677421569824,
      "epoch": 2.8678609535575896,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004191269959826418,
      "loss": 5.0134,
      "mean_token_accuracy": 0.21496371328830718,
      "num_tokens": 56996125.0,
      "step": 25535
    },
    {
      "entropy": 5.294695901870727,
      "epoch": 2.8684225304655473,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041909616330903544,
      "loss": 5.0192,
      "mean_token_accuracy": 0.20843113511800765,
      "num_tokens": 57007599.0,
      "step": 25540
    },
    {
      "entropy": 5.3006445407867435,
      "epoch": 2.868984107373505,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041906532604735127,
      "loss": 5.0207,
      "mean_token_accuracy": 0.21018261164426805,
      "num_tokens": 57018102.0,
      "step": 25545
    },
    {
      "entropy": 5.349273777008056,
      "epoch": 2.869545684281462,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004190344841985713,
      "loss": 5.092,
      "mean_token_accuracy": 0.20800057351589202,
      "num_tokens": 57029253.0,
      "step": 25550
    },
    {
      "entropy": 5.398218011856079,
      "epoch": 2.87010726118942,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041900363776367747,
      "loss": 5.0598,
      "mean_token_accuracy": 0.20555667877197265,
      "num_tokens": 57040485.0,
      "step": 25555
    },
    {
      "entropy": 5.233086156845093,
      "epoch": 2.8706688380973775,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004189727867436519,
      "loss": 4.8913,
      "mean_token_accuracy": 0.22072126865386962,
      "num_tokens": 57051340.0,
      "step": 25560
    },
    {
      "entropy": 5.1781455993652346,
      "epoch": 2.8712304150053347,
      "grad_norm": 1.0,
      "learning_rate": 0.00041894193113947703,
      "loss": 4.948,
      "mean_token_accuracy": 0.21427869349718093,
      "num_tokens": 57061726.0,
      "step": 25565
    },
    {
      "entropy": 5.2830431938171385,
      "epoch": 2.8717919919132924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041891107095213523,
      "loss": 5.017,
      "mean_token_accuracy": 0.21070819050073625,
      "num_tokens": 57073367.0,
      "step": 25570
    },
    {
      "entropy": 5.397106170654297,
      "epoch": 2.87235356882125,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041888020618260913,
      "loss": 5.1252,
      "mean_token_accuracy": 0.21253031492233276,
      "num_tokens": 57085379.0,
      "step": 25575
    },
    {
      "entropy": 5.400806999206543,
      "epoch": 2.8729151457292077,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004188493368318814,
      "loss": 5.0567,
      "mean_token_accuracy": 0.21494307219982148,
      "num_tokens": 57096151.0,
      "step": 25580
    },
    {
      "entropy": 5.305035877227783,
      "epoch": 2.8734767226371654,
      "grad_norm": 1.046875,
      "learning_rate": 0.000418818462900935,
      "loss": 5.071,
      "mean_token_accuracy": 0.20705191493034364,
      "num_tokens": 57107378.0,
      "step": 25585
    },
    {
      "entropy": 5.310205793380737,
      "epoch": 2.8740382995451226,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004187875843907531,
      "loss": 5.0574,
      "mean_token_accuracy": 0.20689097940921783,
      "num_tokens": 57119814.0,
      "step": 25590
    },
    {
      "entropy": 5.507593011856079,
      "epoch": 2.8745998764530802,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041875670130231857,
      "loss": 5.1221,
      "mean_token_accuracy": 0.20951765924692153,
      "num_tokens": 57132070.0,
      "step": 25595
    },
    {
      "entropy": 5.437823677062989,
      "epoch": 2.875161453361038,
      "grad_norm": 1.09375,
      "learning_rate": 0.000418725813636615,
      "loss": 5.1237,
      "mean_token_accuracy": 0.20721967071294783,
      "num_tokens": 57143216.0,
      "step": 25600
    },
    {
      "entropy": 5.229108667373657,
      "epoch": 2.875723030268995,
      "grad_norm": 1.0,
      "learning_rate": 0.00041869492139462585,
      "loss": 5.0277,
      "mean_token_accuracy": 0.2135410413146019,
      "num_tokens": 57154485.0,
      "step": 25605
    },
    {
      "entropy": 5.312213611602783,
      "epoch": 2.876284607176953,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00041866402457733454,
      "loss": 5.0755,
      "mean_token_accuracy": 0.20682656913995742,
      "num_tokens": 57164738.0,
      "step": 25610
    },
    {
      "entropy": 5.402405548095703,
      "epoch": 2.8768461840849104,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041863312318572504,
      "loss": 5.0746,
      "mean_token_accuracy": 0.20976763665676118,
      "num_tokens": 57175790.0,
      "step": 25615
    },
    {
      "entropy": 5.329756069183349,
      "epoch": 2.877407760992868,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004186022172207812,
      "loss": 5.1179,
      "mean_token_accuracy": 0.20440957844257354,
      "num_tokens": 57187201.0,
      "step": 25620
    },
    {
      "entropy": 5.330894517898559,
      "epoch": 2.8779693379008258,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041857130668348705,
      "loss": 4.9562,
      "mean_token_accuracy": 0.21679462641477584,
      "num_tokens": 57197591.0,
      "step": 25625
    },
    {
      "entropy": 5.400057125091553,
      "epoch": 2.878530914808783,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004185403915748268,
      "loss": 5.0804,
      "mean_token_accuracy": 0.21287895292043685,
      "num_tokens": 57210194.0,
      "step": 25630
    },
    {
      "entropy": 5.334183073043823,
      "epoch": 2.8790924917167406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004185094718957848,
      "loss": 5.0233,
      "mean_token_accuracy": 0.20930223017930985,
      "num_tokens": 57220752.0,
      "step": 25635
    },
    {
      "entropy": 5.263636684417724,
      "epoch": 2.8796540686246983,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041847854764734556,
      "loss": 4.9444,
      "mean_token_accuracy": 0.21788402944803237,
      "num_tokens": 57233109.0,
      "step": 25640
    },
    {
      "entropy": 5.28462028503418,
      "epoch": 2.8802156455326555,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0004184476188304937,
      "loss": 5.1198,
      "mean_token_accuracy": 0.20864464193582535,
      "num_tokens": 57244548.0,
      "step": 25645
    },
    {
      "entropy": 5.334233617782592,
      "epoch": 2.880777222440613,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041841668544621396,
      "loss": 5.0048,
      "mean_token_accuracy": 0.2183662474155426,
      "num_tokens": 57255328.0,
      "step": 25650
    },
    {
      "entropy": 5.358094120025635,
      "epoch": 2.881338799348571,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004183857474954915,
      "loss": 5.0278,
      "mean_token_accuracy": 0.2057601273059845,
      "num_tokens": 57268890.0,
      "step": 25655
    },
    {
      "entropy": 5.397912311553955,
      "epoch": 2.881900376256528,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041835480497931114,
      "loss": 5.074,
      "mean_token_accuracy": 0.2116403415799141,
      "num_tokens": 57278505.0,
      "step": 25660
    },
    {
      "entropy": 5.397904586791992,
      "epoch": 2.8824619531644857,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004183238578986582,
      "loss": 5.0727,
      "mean_token_accuracy": 0.2044989913702011,
      "num_tokens": 57289499.0,
      "step": 25665
    },
    {
      "entropy": 5.435350465774536,
      "epoch": 2.8830235300724434,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041829290625451806,
      "loss": 5.2174,
      "mean_token_accuracy": 0.20547000765800477,
      "num_tokens": 57300970.0,
      "step": 25670
    },
    {
      "entropy": 5.326061201095581,
      "epoch": 2.883585106980401,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041826195004787616,
      "loss": 4.945,
      "mean_token_accuracy": 0.21690098196268082,
      "num_tokens": 57311997.0,
      "step": 25675
    },
    {
      "entropy": 5.368774700164795,
      "epoch": 2.8841466838883587,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041823098927971827,
      "loss": 5.0623,
      "mean_token_accuracy": 0.20946099162101744,
      "num_tokens": 57323385.0,
      "step": 25680
    },
    {
      "entropy": 5.327319240570068,
      "epoch": 2.884708260796316,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004182000239510301,
      "loss": 5.0242,
      "mean_token_accuracy": 0.20732742100954055,
      "num_tokens": 57335389.0,
      "step": 25685
    },
    {
      "entropy": 5.3751630783081055,
      "epoch": 2.8852698377042736,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041816905406279756,
      "loss": 5.0693,
      "mean_token_accuracy": 0.21011563241481782,
      "num_tokens": 57346843.0,
      "step": 25690
    },
    {
      "entropy": 5.339729499816895,
      "epoch": 2.8858314146122312,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041813807961600685,
      "loss": 5.0717,
      "mean_token_accuracy": 0.21233271360397338,
      "num_tokens": 57358253.0,
      "step": 25695
    },
    {
      "entropy": 5.350403833389282,
      "epoch": 2.8863929915201885,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004181071006116442,
      "loss": 5.1023,
      "mean_token_accuracy": 0.2038022667169571,
      "num_tokens": 57369204.0,
      "step": 25700
    },
    {
      "entropy": 5.312151765823364,
      "epoch": 2.886954568428146,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041807611705069593,
      "loss": 4.9643,
      "mean_token_accuracy": 0.2180697351694107,
      "num_tokens": 57380214.0,
      "step": 25705
    },
    {
      "entropy": 5.31151180267334,
      "epoch": 2.887516145336104,
      "grad_norm": 1.0,
      "learning_rate": 0.00041804512893414865,
      "loss": 5.0869,
      "mean_token_accuracy": 0.2099156379699707,
      "num_tokens": 57391848.0,
      "step": 25710
    },
    {
      "entropy": 5.40066876411438,
      "epoch": 2.8880777222440615,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041801413626298893,
      "loss": 5.1146,
      "mean_token_accuracy": 0.2090613827109337,
      "num_tokens": 57403026.0,
      "step": 25715
    },
    {
      "entropy": 5.308907747268677,
      "epoch": 2.888639299152019,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004179831390382037,
      "loss": 5.0414,
      "mean_token_accuracy": 0.20703014433383943,
      "num_tokens": 57414223.0,
      "step": 25720
    },
    {
      "entropy": 5.295155382156372,
      "epoch": 2.8892008760599763,
      "grad_norm": 1.0,
      "learning_rate": 0.0004179521372607798,
      "loss": 4.9829,
      "mean_token_accuracy": 0.21432085186243058,
      "num_tokens": 57424661.0,
      "step": 25725
    },
    {
      "entropy": 5.287267017364502,
      "epoch": 2.889762452967934,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041792113093170447,
      "loss": 4.9706,
      "mean_token_accuracy": 0.21422389894723892,
      "num_tokens": 57435859.0,
      "step": 25730
    },
    {
      "entropy": 5.299745798110962,
      "epoch": 2.8903240298758917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004178901200519649,
      "loss": 5.0091,
      "mean_token_accuracy": 0.20657630115747452,
      "num_tokens": 57447784.0,
      "step": 25735
    },
    {
      "entropy": 5.263573217391968,
      "epoch": 2.890885606783849,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004178591046225485,
      "loss": 5.0675,
      "mean_token_accuracy": 0.20629919022321702,
      "num_tokens": 57458881.0,
      "step": 25740
    },
    {
      "entropy": 5.312130784988403,
      "epoch": 2.8914471836918065,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041782808464444274,
      "loss": 5.0336,
      "mean_token_accuracy": 0.20933582484722138,
      "num_tokens": 57470287.0,
      "step": 25745
    },
    {
      "entropy": 5.384986543655396,
      "epoch": 2.892008760599764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041779706011863546,
      "loss": 5.0844,
      "mean_token_accuracy": 0.2067478895187378,
      "num_tokens": 57481723.0,
      "step": 25750
    },
    {
      "entropy": 5.381469488143921,
      "epoch": 2.8925703375077214,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041776603104611427,
      "loss": 5.0922,
      "mean_token_accuracy": 0.2049788475036621,
      "num_tokens": 57492804.0,
      "step": 25755
    },
    {
      "entropy": 5.273059368133545,
      "epoch": 2.893131914415679,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041773499742786736,
      "loss": 4.9683,
      "mean_token_accuracy": 0.214146426320076,
      "num_tokens": 57504148.0,
      "step": 25760
    },
    {
      "entropy": 5.344239854812622,
      "epoch": 2.8936934913236367,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041770395926488276,
      "loss": 5.0706,
      "mean_token_accuracy": 0.21323953717947006,
      "num_tokens": 57516026.0,
      "step": 25765
    },
    {
      "entropy": 5.338073492050171,
      "epoch": 2.8942550682315944,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004176729165581488,
      "loss": 5.0427,
      "mean_token_accuracy": 0.20185501575469972,
      "num_tokens": 57526882.0,
      "step": 25770
    },
    {
      "entropy": 5.259396171569824,
      "epoch": 2.894816645139552,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041764186930865377,
      "loss": 5.0436,
      "mean_token_accuracy": 0.2146101862192154,
      "num_tokens": 57537875.0,
      "step": 25775
    },
    {
      "entropy": 5.331795978546142,
      "epoch": 2.8953782220475093,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004176108175173863,
      "loss": 4.9984,
      "mean_token_accuracy": 0.21178067922592164,
      "num_tokens": 57548143.0,
      "step": 25780
    },
    {
      "entropy": 5.268396520614624,
      "epoch": 2.895939798955467,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004175797611853351,
      "loss": 4.9959,
      "mean_token_accuracy": 0.21596294194459914,
      "num_tokens": 57558127.0,
      "step": 25785
    },
    {
      "entropy": 5.361077976226807,
      "epoch": 2.8965013758634246,
      "grad_norm": 1.0,
      "learning_rate": 0.000417548700313489,
      "loss": 5.0933,
      "mean_token_accuracy": 0.21652275174856186,
      "num_tokens": 57570028.0,
      "step": 25790
    },
    {
      "entropy": 5.33809232711792,
      "epoch": 2.897062952771382,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000417517634902837,
      "loss": 4.9576,
      "mean_token_accuracy": 0.2158510357141495,
      "num_tokens": 57581089.0,
      "step": 25795
    },
    {
      "entropy": 5.377852439880371,
      "epoch": 2.8976245296793395,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00041748656495436834,
      "loss": 5.1192,
      "mean_token_accuracy": 0.20916281938552855,
      "num_tokens": 57592555.0,
      "step": 25800
    },
    {
      "entropy": 5.289712810516358,
      "epoch": 2.898186106587297,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041745549046907205,
      "loss": 4.9995,
      "mean_token_accuracy": 0.21070680767297745,
      "num_tokens": 57603492.0,
      "step": 25805
    },
    {
      "entropy": 5.257569932937622,
      "epoch": 2.898747683495255,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041742441144793774,
      "loss": 4.9198,
      "mean_token_accuracy": 0.20940887480974196,
      "num_tokens": 57613477.0,
      "step": 25810
    },
    {
      "entropy": 5.316675138473511,
      "epoch": 2.8993092604032125,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004173933278919549,
      "loss": 5.0113,
      "mean_token_accuracy": 0.20427666902542113,
      "num_tokens": 57623703.0,
      "step": 25815
    },
    {
      "entropy": 5.355856323242188,
      "epoch": 2.8998708373111697,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004173622398021133,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21193174570798873,
      "num_tokens": 57637172.0,
      "step": 25820
    },
    {
      "entropy": 5.347334718704223,
      "epoch": 2.9004324142191273,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041733114717940273,
      "loss": 5.046,
      "mean_token_accuracy": 0.20955029278993606,
      "num_tokens": 57648224.0,
      "step": 25825
    },
    {
      "entropy": 5.289870500564575,
      "epoch": 2.900993991127085,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004173000500248133,
      "loss": 5.0395,
      "mean_token_accuracy": 0.22108842581510543,
      "num_tokens": 57659093.0,
      "step": 25830
    },
    {
      "entropy": 5.374278879165649,
      "epoch": 2.901555568035042,
      "grad_norm": 0.88671875,
      "learning_rate": 0.0004172689483393351,
      "loss": 5.1288,
      "mean_token_accuracy": 0.20566474050283431,
      "num_tokens": 57671399.0,
      "step": 25835
    },
    {
      "entropy": 5.366710948944092,
      "epoch": 2.902117144943,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041723784212395834,
      "loss": 5.0391,
      "mean_token_accuracy": 0.2207150235772133,
      "num_tokens": 57682150.0,
      "step": 25840
    },
    {
      "entropy": 5.291201162338257,
      "epoch": 2.9026787218509575,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041720673137967356,
      "loss": 4.9985,
      "mean_token_accuracy": 0.21012357771396636,
      "num_tokens": 57693232.0,
      "step": 25845
    },
    {
      "entropy": 5.30739197731018,
      "epoch": 2.9032402987589148,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041717561610747124,
      "loss": 4.999,
      "mean_token_accuracy": 0.21349496096372605,
      "num_tokens": 57703422.0,
      "step": 25850
    },
    {
      "entropy": 5.327971696853638,
      "epoch": 2.9038018756668724,
      "grad_norm": 1.0,
      "learning_rate": 0.00041714449630834223,
      "loss": 5.0105,
      "mean_token_accuracy": 0.21135815978050232,
      "num_tokens": 57714454.0,
      "step": 25855
    },
    {
      "entropy": 5.249827337265015,
      "epoch": 2.90436345257483,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004171133719832774,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22364603877067565,
      "num_tokens": 57726243.0,
      "step": 25860
    },
    {
      "entropy": 5.263783693313599,
      "epoch": 2.9049250294827877,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041708224313326747,
      "loss": 4.9795,
      "mean_token_accuracy": 0.21339137405157088,
      "num_tokens": 57737882.0,
      "step": 25865
    },
    {
      "entropy": 5.380540466308593,
      "epoch": 2.9054866063907454,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000417051109759304,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2115017980337143,
      "num_tokens": 57750205.0,
      "step": 25870
    },
    {
      "entropy": 5.315791368484497,
      "epoch": 2.9060481832987026,
      "grad_norm": 1.09375,
      "learning_rate": 0.000417019971862378,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2172745645046234,
      "num_tokens": 57761762.0,
      "step": 25875
    },
    {
      "entropy": 5.315591478347779,
      "epoch": 2.9066097602066603,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041698882944348105,
      "loss": 5.0247,
      "mean_token_accuracy": 0.21109730899333953,
      "num_tokens": 57771280.0,
      "step": 25880
    },
    {
      "entropy": 5.307880401611328,
      "epoch": 2.907171337114618,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041695768250360466,
      "loss": 4.9985,
      "mean_token_accuracy": 0.20971003025770188,
      "num_tokens": 57784099.0,
      "step": 25885
    },
    {
      "entropy": 5.359955930709839,
      "epoch": 2.907732914022575,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0004169265310437407,
      "loss": 5.0267,
      "mean_token_accuracy": 0.20937896966934205,
      "num_tokens": 57795544.0,
      "step": 25890
    },
    {
      "entropy": 5.300588989257813,
      "epoch": 2.908294490930533,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004168953750648809,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22326743304729463,
      "num_tokens": 57805963.0,
      "step": 25895
    },
    {
      "entropy": 5.348090887069702,
      "epoch": 2.9088560678384905,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041686421456801724,
      "loss": 5.0773,
      "mean_token_accuracy": 0.20526752918958663,
      "num_tokens": 57815281.0,
      "step": 25900
    },
    {
      "entropy": 5.289106321334839,
      "epoch": 2.909417644746448,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0004168330495541421,
      "loss": 5.0893,
      "mean_token_accuracy": 0.20294949263334275,
      "num_tokens": 57826172.0,
      "step": 25905
    },
    {
      "entropy": 5.354646968841553,
      "epoch": 2.909979221654406,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041680188002424755,
      "loss": 4.94,
      "mean_token_accuracy": 0.21331207007169722,
      "num_tokens": 57836617.0,
      "step": 25910
    },
    {
      "entropy": 5.328772830963135,
      "epoch": 2.910540798562363,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004167707059793262,
      "loss": 4.9574,
      "mean_token_accuracy": 0.2124924972653389,
      "num_tokens": 57847636.0,
      "step": 25915
    },
    {
      "entropy": 5.392505884170532,
      "epoch": 2.9111023754703207,
      "grad_norm": 1.125,
      "learning_rate": 0.00041673952742037057,
      "loss": 5.1273,
      "mean_token_accuracy": 0.2029580369591713,
      "num_tokens": 57858816.0,
      "step": 25920
    },
    {
      "entropy": 5.347265100479126,
      "epoch": 2.9116639523782784,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004167083443483734,
      "loss": 5.0524,
      "mean_token_accuracy": 0.21296071857213975,
      "num_tokens": 57870749.0,
      "step": 25925
    },
    {
      "entropy": 5.33793363571167,
      "epoch": 2.9122255292862356,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004166771567643276,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2115020200610161,
      "num_tokens": 57883454.0,
      "step": 25930
    },
    {
      "entropy": 5.312660264968872,
      "epoch": 2.9127871061941932,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004166459646692262,
      "loss": 4.985,
      "mean_token_accuracy": 0.21133558452129364,
      "num_tokens": 57894498.0,
      "step": 25935
    },
    {
      "entropy": 5.212173986434936,
      "epoch": 2.913348683102151,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004166147680640623,
      "loss": 4.8338,
      "mean_token_accuracy": 0.22350470423698426,
      "num_tokens": 57905332.0,
      "step": 25940
    },
    {
      "entropy": 5.238060712814331,
      "epoch": 2.913910260010108,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041658356694982936,
      "loss": 4.956,
      "mean_token_accuracy": 0.21984383463859558,
      "num_tokens": 57916456.0,
      "step": 25945
    },
    {
      "entropy": 5.289606666564941,
      "epoch": 2.9144718369180658,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004165523613275207,
      "loss": 5.0198,
      "mean_token_accuracy": 0.2199738696217537,
      "num_tokens": 57927615.0,
      "step": 25950
    },
    {
      "entropy": 5.231566953659057,
      "epoch": 2.9150334138260234,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004165211511981299,
      "loss": 4.8644,
      "mean_token_accuracy": 0.22255100905895234,
      "num_tokens": 57937771.0,
      "step": 25955
    },
    {
      "entropy": 5.247180318832397,
      "epoch": 2.915594990733981,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004164899365626507,
      "loss": 4.8857,
      "mean_token_accuracy": 0.21227786540985108,
      "num_tokens": 57948564.0,
      "step": 25960
    },
    {
      "entropy": 5.238268613815308,
      "epoch": 2.9161565676419388,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004164587174220771,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2179258197546005,
      "num_tokens": 57959758.0,
      "step": 25965
    },
    {
      "entropy": 5.320282316207885,
      "epoch": 2.916718144549896,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041642749377740316,
      "loss": 5.0379,
      "mean_token_accuracy": 0.2148979276418686,
      "num_tokens": 57970988.0,
      "step": 25970
    },
    {
      "entropy": 5.250712537765503,
      "epoch": 2.9172797214578536,
      "grad_norm": 1.125,
      "learning_rate": 0.00041639626562962293,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22433241754770278,
      "num_tokens": 57981056.0,
      "step": 25975
    },
    {
      "entropy": 5.273747491836548,
      "epoch": 2.9178412983658113,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041636503297973074,
      "loss": 5.0225,
      "mean_token_accuracy": 0.21731463074684143,
      "num_tokens": 57992092.0,
      "step": 25980
    },
    {
      "entropy": 5.353913640975952,
      "epoch": 2.9184028752737685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000416333795828721,
      "loss": 5.006,
      "mean_token_accuracy": 0.22264466285705567,
      "num_tokens": 58003000.0,
      "step": 25985
    },
    {
      "entropy": 5.26265516281128,
      "epoch": 2.918964452181726,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004163025541775885,
      "loss": 4.9656,
      "mean_token_accuracy": 0.2135253742337227,
      "num_tokens": 58013932.0,
      "step": 25990
    },
    {
      "entropy": 5.368308401107788,
      "epoch": 2.919526029089684,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004162713080273278,
      "loss": 5.0645,
      "mean_token_accuracy": 0.21021942496299745,
      "num_tokens": 58024481.0,
      "step": 25995
    },
    {
      "entropy": 5.370873975753784,
      "epoch": 2.9200876059976415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041624005737893387,
      "loss": 5.0257,
      "mean_token_accuracy": 0.2125475987792015,
      "num_tokens": 58035753.0,
      "step": 26000
    },
    {
      "entropy": 5.42750244140625,
      "epoch": 2.920649182905599,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004162088022334018,
      "loss": 5.1804,
      "mean_token_accuracy": 0.2033931940793991,
      "num_tokens": 58046599.0,
      "step": 26005
    },
    {
      "entropy": 5.34784688949585,
      "epoch": 2.9212107598135564,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041617754259172665,
      "loss": 5.0696,
      "mean_token_accuracy": 0.2078058034181595,
      "num_tokens": 58058151.0,
      "step": 26010
    },
    {
      "entropy": 5.372073602676392,
      "epoch": 2.921772336721514,
      "grad_norm": 0.9375,
      "learning_rate": 0.00041614627845490377,
      "loss": 5.0665,
      "mean_token_accuracy": 0.20911595225334167,
      "num_tokens": 58070116.0,
      "step": 26015
    },
    {
      "entropy": 5.3132078647613525,
      "epoch": 2.9223339136294717,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041611500982392867,
      "loss": 5.0052,
      "mean_token_accuracy": 0.2131350129842758,
      "num_tokens": 58081089.0,
      "step": 26020
    },
    {
      "entropy": 5.228788089752197,
      "epoch": 2.922895490537429,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041608373669979696,
      "loss": 4.9344,
      "mean_token_accuracy": 0.21725170612335204,
      "num_tokens": 58092033.0,
      "step": 26025
    },
    {
      "entropy": 5.287259435653686,
      "epoch": 2.9234570674453866,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004160524590835042,
      "loss": 4.9354,
      "mean_token_accuracy": 0.21483835726976394,
      "num_tokens": 58102864.0,
      "step": 26030
    },
    {
      "entropy": 5.273019647598266,
      "epoch": 2.9240186443533442,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00041602117697604664,
      "loss": 5.0044,
      "mean_token_accuracy": 0.2051311358809471,
      "num_tokens": 58114607.0,
      "step": 26035
    },
    {
      "entropy": 5.311244487762451,
      "epoch": 2.9245802212613015,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041598989037842,
      "loss": 5.0327,
      "mean_token_accuracy": 0.21244224458932875,
      "num_tokens": 58125935.0,
      "step": 26040
    },
    {
      "entropy": 5.302919960021972,
      "epoch": 2.925141798169259,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004159585992916206,
      "loss": 4.971,
      "mean_token_accuracy": 0.22149900943040848,
      "num_tokens": 58136439.0,
      "step": 26045
    },
    {
      "entropy": 5.304346704483033,
      "epoch": 2.925703375077217,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0004159273037166447,
      "loss": 5.0178,
      "mean_token_accuracy": 0.2095719739794731,
      "num_tokens": 58147778.0,
      "step": 26050
    },
    {
      "entropy": 5.345755529403687,
      "epoch": 2.9262649519851744,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041589600365448884,
      "loss": 5.0261,
      "mean_token_accuracy": 0.2188476115465164,
      "num_tokens": 58159290.0,
      "step": 26055
    },
    {
      "entropy": 5.324849367141724,
      "epoch": 2.926826528893132,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004158646991061495,
      "loss": 5.0402,
      "mean_token_accuracy": 0.20837315917015076,
      "num_tokens": 58169917.0,
      "step": 26060
    },
    {
      "entropy": 5.276232671737671,
      "epoch": 2.9273881058010893,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004158333900726235,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2167723298072815,
      "num_tokens": 58180905.0,
      "step": 26065
    },
    {
      "entropy": 5.322358512878418,
      "epoch": 2.927949682709047,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00041580207655490776,
      "loss": 5.0132,
      "mean_token_accuracy": 0.21450288593769073,
      "num_tokens": 58192312.0,
      "step": 26070
    },
    {
      "entropy": 5.2651222229003904,
      "epoch": 2.9285112596170046,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004157707585539994,
      "loss": 4.9888,
      "mean_token_accuracy": 0.21682239919900895,
      "num_tokens": 58204739.0,
      "step": 26075
    },
    {
      "entropy": 5.256013345718384,
      "epoch": 2.929072836524962,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041573943607089536,
      "loss": 4.9726,
      "mean_token_accuracy": 0.21341193318367005,
      "num_tokens": 58215663.0,
      "step": 26080
    },
    {
      "entropy": 5.2971127986907955,
      "epoch": 2.9296344134329195,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004157081091065932,
      "loss": 4.9678,
      "mean_token_accuracy": 0.21081486940383912,
      "num_tokens": 58226480.0,
      "step": 26085
    },
    {
      "entropy": 5.508465242385864,
      "epoch": 2.930195990340877,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004156767776620902,
      "loss": 5.2959,
      "mean_token_accuracy": 0.20251937061548234,
      "num_tokens": 58238087.0,
      "step": 26090
    },
    {
      "entropy": 5.2879723072052,
      "epoch": 2.930757567248835,
      "grad_norm": 0.984375,
      "learning_rate": 0.000415645441738384,
      "loss": 5.0534,
      "mean_token_accuracy": 0.20413236021995546,
      "num_tokens": 58248738.0,
      "step": 26095
    },
    {
      "entropy": 5.275141906738281,
      "epoch": 2.9313191441567925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041561410133647244,
      "loss": 5.0202,
      "mean_token_accuracy": 0.2112732157111168,
      "num_tokens": 58260440.0,
      "step": 26100
    },
    {
      "entropy": 5.421850538253784,
      "epoch": 2.9318807210647497,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041558275645735336,
      "loss": 5.0594,
      "mean_token_accuracy": 0.21222403645515442,
      "num_tokens": 58272704.0,
      "step": 26105
    },
    {
      "entropy": 5.296739864349365,
      "epoch": 2.9324422979727074,
      "grad_norm": 1.0,
      "learning_rate": 0.00041555140710202484,
      "loss": 5.0496,
      "mean_token_accuracy": 0.20873338282108306,
      "num_tokens": 58284254.0,
      "step": 26110
    },
    {
      "entropy": 5.343094539642334,
      "epoch": 2.933003874880665,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00041552005327148494,
      "loss": 5.1213,
      "mean_token_accuracy": 0.21073902696371077,
      "num_tokens": 58296472.0,
      "step": 26115
    },
    {
      "entropy": 5.285445976257324,
      "epoch": 2.9335654517886223,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004154886949667321,
      "loss": 4.947,
      "mean_token_accuracy": 0.21836545765399934,
      "num_tokens": 58307734.0,
      "step": 26120
    },
    {
      "entropy": 5.2938987731933596,
      "epoch": 2.93412702869658,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041545733218876466,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21687083691358566,
      "num_tokens": 58318329.0,
      "step": 26125
    },
    {
      "entropy": 5.259852647781372,
      "epoch": 2.9346886056045376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004154259649385813,
      "loss": 4.9686,
      "mean_token_accuracy": 0.21235000491142272,
      "num_tokens": 58330036.0,
      "step": 26130
    },
    {
      "entropy": 5.277456760406494,
      "epoch": 2.935250182512495,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041539459321718076,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22042602598667144,
      "num_tokens": 58340303.0,
      "step": 26135
    },
    {
      "entropy": 5.2647754669189455,
      "epoch": 2.9358117594204525,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004153632170255619,
      "loss": 4.9586,
      "mean_token_accuracy": 0.22403233200311662,
      "num_tokens": 58352394.0,
      "step": 26140
    },
    {
      "entropy": 5.324893856048584,
      "epoch": 2.93637333632841,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041533183636472385,
      "loss": 5.1436,
      "mean_token_accuracy": 0.20431118011474608,
      "num_tokens": 58364605.0,
      "step": 26145
    },
    {
      "entropy": 5.339525175094605,
      "epoch": 2.936934913236368,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004153004512356656,
      "loss": 5.0114,
      "mean_token_accuracy": 0.22087572664022445,
      "num_tokens": 58375514.0,
      "step": 26150
    },
    {
      "entropy": 5.345520162582398,
      "epoch": 2.9374964901443255,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004152690616393866,
      "loss": 5.017,
      "mean_token_accuracy": 0.21503239423036574,
      "num_tokens": 58387135.0,
      "step": 26155
    },
    {
      "entropy": 5.270255613327026,
      "epoch": 2.9380580670522827,
      "grad_norm": 1.109375,
      "learning_rate": 0.00041523766757688634,
      "loss": 5.0102,
      "mean_token_accuracy": 0.2078711897134781,
      "num_tokens": 58399077.0,
      "step": 26160
    },
    {
      "entropy": 5.391632080078125,
      "epoch": 2.9386196439602403,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004152062690491643,
      "loss": 5.1295,
      "mean_token_accuracy": 0.21231186836957933,
      "num_tokens": 58409897.0,
      "step": 26165
    },
    {
      "entropy": 5.268762493133545,
      "epoch": 2.939181220868198,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004151748660572203,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22220344692468644,
      "num_tokens": 58421189.0,
      "step": 26170
    },
    {
      "entropy": 5.237098836898804,
      "epoch": 2.939742797776155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004151434586020541,
      "loss": 5.0334,
      "mean_token_accuracy": 0.21145639270544053,
      "num_tokens": 58432258.0,
      "step": 26175
    },
    {
      "entropy": 5.385358572006226,
      "epoch": 2.940304374684113,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041511204668466593,
      "loss": 5.0701,
      "mean_token_accuracy": 0.2096214085817337,
      "num_tokens": 58444288.0,
      "step": 26180
    },
    {
      "entropy": 5.355762577056884,
      "epoch": 2.9408659515920705,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041508063030605585,
      "loss": 5.0197,
      "mean_token_accuracy": 0.21616208851337432,
      "num_tokens": 58453367.0,
      "step": 26185
    },
    {
      "entropy": 5.322093200683594,
      "epoch": 2.941427528500028,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00041504920946722413,
      "loss": 5.154,
      "mean_token_accuracy": 0.20784422606229783,
      "num_tokens": 58464862.0,
      "step": 26190
    },
    {
      "entropy": 5.370402097702026,
      "epoch": 2.941989105407986,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004150177841691713,
      "loss": 5.0192,
      "mean_token_accuracy": 0.2237808436155319,
      "num_tokens": 58476353.0,
      "step": 26195
    },
    {
      "entropy": 5.3549644947052,
      "epoch": 2.942550682315943,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004149863544128979,
      "loss": 5.0752,
      "mean_token_accuracy": 0.2079213798046112,
      "num_tokens": 58488330.0,
      "step": 26200
    },
    {
      "entropy": 5.303336095809937,
      "epoch": 2.9431122592239007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041495492019940467,
      "loss": 5.0173,
      "mean_token_accuracy": 0.2191261664032936,
      "num_tokens": 58500004.0,
      "step": 26205
    },
    {
      "entropy": 5.289502954483032,
      "epoch": 2.9436738361318584,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004149234815296925,
      "loss": 5.0076,
      "mean_token_accuracy": 0.20923326909542084,
      "num_tokens": 58511433.0,
      "step": 26210
    },
    {
      "entropy": 5.244606637954712,
      "epoch": 2.9442354130398156,
      "grad_norm": 0.953125,
      "learning_rate": 0.00041489203840476243,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2140759974718094,
      "num_tokens": 58522234.0,
      "step": 26215
    },
    {
      "entropy": 5.333122587203979,
      "epoch": 2.9447969899477733,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041486059082561556,
      "loss": 5.0321,
      "mean_token_accuracy": 0.21421577036380768,
      "num_tokens": 58534370.0,
      "step": 26220
    },
    {
      "entropy": 5.354115772247314,
      "epoch": 2.945358566855731,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004148291387932533,
      "loss": 4.9825,
      "mean_token_accuracy": 0.21683958172798157,
      "num_tokens": 58545859.0,
      "step": 26225
    },
    {
      "entropy": 5.238599395751953,
      "epoch": 2.9459201437636886,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00041479768230867696,
      "loss": 4.928,
      "mean_token_accuracy": 0.2230888932943344,
      "num_tokens": 58557159.0,
      "step": 26230
    },
    {
      "entropy": 5.299037027359009,
      "epoch": 2.946481720671646,
      "grad_norm": 1.0,
      "learning_rate": 0.00041476622137288823,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20963150262832642,
      "num_tokens": 58568985.0,
      "step": 26235
    },
    {
      "entropy": 5.370597314834595,
      "epoch": 2.9470432975796035,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041473475598688886,
      "loss": 5.0243,
      "mean_token_accuracy": 0.21380632519721984,
      "num_tokens": 58579897.0,
      "step": 26240
    },
    {
      "entropy": 5.282123136520386,
      "epoch": 2.947604874487561,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004147032861516805,
      "loss": 4.9142,
      "mean_token_accuracy": 0.2238635465502739,
      "num_tokens": 58590042.0,
      "step": 26245
    },
    {
      "entropy": 5.2872882843017575,
      "epoch": 2.948166451395519,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041467181186826546,
      "loss": 5.0749,
      "mean_token_accuracy": 0.21359614580869674,
      "num_tokens": 58601645.0,
      "step": 26250
    },
    {
      "entropy": 5.338930654525757,
      "epoch": 2.948728028303476,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041464033313764575,
      "loss": 5.0609,
      "mean_token_accuracy": 0.2101604864001274,
      "num_tokens": 58613881.0,
      "step": 26255
    },
    {
      "entropy": 5.278785753250122,
      "epoch": 2.9492896052114337,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041460884996082363,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21748945564031602,
      "num_tokens": 58623430.0,
      "step": 26260
    },
    {
      "entropy": 5.2728065013885494,
      "epoch": 2.9498511821193913,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041457736233880166,
      "loss": 4.9422,
      "mean_token_accuracy": 0.20996131896972656,
      "num_tokens": 58635289.0,
      "step": 26265
    },
    {
      "entropy": 5.2712647914886475,
      "epoch": 2.9504127590273486,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004145458702725823,
      "loss": 4.9997,
      "mean_token_accuracy": 0.21318752467632293,
      "num_tokens": 58646086.0,
      "step": 26270
    },
    {
      "entropy": 5.3432587623596195,
      "epoch": 2.9509743359353062,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041451437376316825,
      "loss": 4.9819,
      "mean_token_accuracy": 0.20954934060573577,
      "num_tokens": 58657662.0,
      "step": 26275
    },
    {
      "entropy": 5.344134712219239,
      "epoch": 2.951535912843264,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004144828728115624,
      "loss": 4.9757,
      "mean_token_accuracy": 0.21716747283935547,
      "num_tokens": 58668054.0,
      "step": 26280
    },
    {
      "entropy": 5.336728906631469,
      "epoch": 2.9520974897512215,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004144513674187679,
      "loss": 5.1147,
      "mean_token_accuracy": 0.20785349905490874,
      "num_tokens": 58679285.0,
      "step": 26285
    },
    {
      "entropy": 5.363505458831787,
      "epoch": 2.952659066659179,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004144198575857878,
      "loss": 5.0895,
      "mean_token_accuracy": 0.21629926413297654,
      "num_tokens": 58691878.0,
      "step": 26290
    },
    {
      "entropy": 5.244097948074341,
      "epoch": 2.9532206435671364,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004143883433136252,
      "loss": 4.9353,
      "mean_token_accuracy": 0.22017164528369904,
      "num_tokens": 58702130.0,
      "step": 26295
    },
    {
      "entropy": 5.304179763793945,
      "epoch": 2.953782220475094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004143568246032838,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21816360652446748,
      "num_tokens": 58712872.0,
      "step": 26300
    },
    {
      "entropy": 5.296761417388916,
      "epoch": 2.9543437973830518,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004143253014557671,
      "loss": 5.0554,
      "mean_token_accuracy": 0.20917683839797974,
      "num_tokens": 58724783.0,
      "step": 26305
    },
    {
      "entropy": 5.352311372756958,
      "epoch": 2.954905374291009,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041429377387207865,
      "loss": 5.0632,
      "mean_token_accuracy": 0.21134239286184311,
      "num_tokens": 58736473.0,
      "step": 26310
    },
    {
      "entropy": 5.376512241363526,
      "epoch": 2.9554669511989666,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00041426224185322255,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20936593264341355,
      "num_tokens": 58748223.0,
      "step": 26315
    },
    {
      "entropy": 5.229674577713013,
      "epoch": 2.9560285281069243,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004142307054002025,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21993909627199174,
      "num_tokens": 58759876.0,
      "step": 26320
    },
    {
      "entropy": 5.317415809631347,
      "epoch": 2.956590105014882,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004141991645140228,
      "loss": 5.0642,
      "mean_token_accuracy": 0.21112691462039948,
      "num_tokens": 58771288.0,
      "step": 26325
    },
    {
      "entropy": 5.3860115051269535,
      "epoch": 2.957151681922839,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041416761919568784,
      "loss": 5.0929,
      "mean_token_accuracy": 0.2081226885318756,
      "num_tokens": 58781664.0,
      "step": 26330
    },
    {
      "entropy": 5.321681070327759,
      "epoch": 2.957713258830797,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00041413606944620183,
      "loss": 5.0249,
      "mean_token_accuracy": 0.2060571238398552,
      "num_tokens": 58792808.0,
      "step": 26335
    },
    {
      "entropy": 5.257224178314209,
      "epoch": 2.9582748357387545,
      "grad_norm": 1.0,
      "learning_rate": 0.0004141045152665695,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2169914647936821,
      "num_tokens": 58804455.0,
      "step": 26340
    },
    {
      "entropy": 5.263885068893432,
      "epoch": 2.958836412646712,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041407295665779533,
      "loss": 4.9446,
      "mean_token_accuracy": 0.21996526420116425,
      "num_tokens": 58814718.0,
      "step": 26345
    },
    {
      "entropy": 5.369960927963257,
      "epoch": 2.9593979895546694,
      "grad_norm": 1.0,
      "learning_rate": 0.00041404139362088434,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20863032042980195,
      "num_tokens": 58825649.0,
      "step": 26350
    },
    {
      "entropy": 5.336319446563721,
      "epoch": 2.959959566462627,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004140098261568414,
      "loss": 5.0802,
      "mean_token_accuracy": 0.2050530344247818,
      "num_tokens": 58836529.0,
      "step": 26355
    },
    {
      "entropy": 5.318717193603516,
      "epoch": 2.9605211433705847,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041397825426667175,
      "loss": 5.0223,
      "mean_token_accuracy": 0.20927380621433259,
      "num_tokens": 58847879.0,
      "step": 26360
    },
    {
      "entropy": 5.400581693649292,
      "epoch": 2.961082720278542,
      "grad_norm": 1.125,
      "learning_rate": 0.0004139466779513805,
      "loss": 5.1702,
      "mean_token_accuracy": 0.20126103311777116,
      "num_tokens": 58859283.0,
      "step": 26365
    },
    {
      "entropy": 5.36825795173645,
      "epoch": 2.9616442971864996,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004139150972119732,
      "loss": 5.0489,
      "mean_token_accuracy": 0.2114264965057373,
      "num_tokens": 58870529.0,
      "step": 26370
    },
    {
      "entropy": 5.340286493301392,
      "epoch": 2.9622058740944572,
      "grad_norm": 1.125,
      "learning_rate": 0.00041388351204945534,
      "loss": 4.973,
      "mean_token_accuracy": 0.21533016562461854,
      "num_tokens": 58880828.0,
      "step": 26375
    },
    {
      "entropy": 5.405907344818115,
      "epoch": 2.962767451002415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041385192246483257,
      "loss": 5.1063,
      "mean_token_accuracy": 0.21363508552312852,
      "num_tokens": 58892121.0,
      "step": 26380
    },
    {
      "entropy": 5.341250848770142,
      "epoch": 2.9633290279103726,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041382032845911073,
      "loss": 5.0872,
      "mean_token_accuracy": 0.2105294331908226,
      "num_tokens": 58902768.0,
      "step": 26385
    },
    {
      "entropy": 5.353049898147583,
      "epoch": 2.9638906048183298,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041378873003329575,
      "loss": 5.0483,
      "mean_token_accuracy": 0.2150912269949913,
      "num_tokens": 58913404.0,
      "step": 26390
    },
    {
      "entropy": 5.436030578613281,
      "epoch": 2.9644521817262874,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004137571271883938,
      "loss": 5.135,
      "mean_token_accuracy": 0.21375291645526887,
      "num_tokens": 58925086.0,
      "step": 26395
    },
    {
      "entropy": 5.325554466247558,
      "epoch": 2.965013758634245,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004137255199254112,
      "loss": 5.0062,
      "mean_token_accuracy": 0.21172477006912233,
      "num_tokens": 58935425.0,
      "step": 26400
    },
    {
      "entropy": 5.34139404296875,
      "epoch": 2.9655753355422023,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041369390824535417,
      "loss": 5.0622,
      "mean_token_accuracy": 0.20591685622930528,
      "num_tokens": 58945528.0,
      "step": 26405
    },
    {
      "entropy": 5.348260450363159,
      "epoch": 2.96613691245016,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041366229214922926,
      "loss": 5.0026,
      "mean_token_accuracy": 0.21643766313791274,
      "num_tokens": 58957008.0,
      "step": 26410
    },
    {
      "entropy": 5.3640235424041744,
      "epoch": 2.9666984893581176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004136306716380432,
      "loss": 5.0436,
      "mean_token_accuracy": 0.20437567830085754,
      "num_tokens": 58968648.0,
      "step": 26415
    },
    {
      "entropy": 5.315442228317261,
      "epoch": 2.9672600662660753,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041359904671280285,
      "loss": 5.0331,
      "mean_token_accuracy": 0.20708576142787932,
      "num_tokens": 58979794.0,
      "step": 26420
    },
    {
      "entropy": 5.302256870269775,
      "epoch": 2.9678216431740325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004135674173745151,
      "loss": 4.8975,
      "mean_token_accuracy": 0.21830620020627975,
      "num_tokens": 58990534.0,
      "step": 26425
    },
    {
      "entropy": 5.2912403583526615,
      "epoch": 2.96838322008199,
      "grad_norm": 1.21875,
      "learning_rate": 0.00041353578362418695,
      "loss": 4.9925,
      "mean_token_accuracy": 0.21098758280277252,
      "num_tokens": 59001717.0,
      "step": 26430
    },
    {
      "entropy": 5.26113486289978,
      "epoch": 2.968944796989948,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004135041454628258,
      "loss": 4.9669,
      "mean_token_accuracy": 0.21057455092668534,
      "num_tokens": 59012118.0,
      "step": 26435
    },
    {
      "entropy": 5.372827577590942,
      "epoch": 2.9695063738979055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004134725028914389,
      "loss": 5.076,
      "mean_token_accuracy": 0.2087334305047989,
      "num_tokens": 59023940.0,
      "step": 26440
    },
    {
      "entropy": 5.3393457412719725,
      "epoch": 2.9700679508058627,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041344085591103383,
      "loss": 4.9736,
      "mean_token_accuracy": 0.2161242425441742,
      "num_tokens": 59035147.0,
      "step": 26445
    },
    {
      "entropy": 5.22050895690918,
      "epoch": 2.9706295277138204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004134092045226182,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22392189651727676,
      "num_tokens": 59045531.0,
      "step": 26450
    },
    {
      "entropy": 5.337694549560547,
      "epoch": 2.971191104621778,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041337754872719984,
      "loss": 5.1356,
      "mean_token_accuracy": 0.2142123982310295,
      "num_tokens": 59056364.0,
      "step": 26455
    },
    {
      "entropy": 5.323794507980347,
      "epoch": 2.9717526815297353,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004133458885257866,
      "loss": 4.9352,
      "mean_token_accuracy": 0.21617476344108583,
      "num_tokens": 59066504.0,
      "step": 26460
    },
    {
      "entropy": 5.25397834777832,
      "epoch": 2.972314258437693,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041331422391938663,
      "loss": 4.9996,
      "mean_token_accuracy": 0.21824904084205626,
      "num_tokens": 59077888.0,
      "step": 26465
    },
    {
      "entropy": 5.348627662658691,
      "epoch": 2.9728758353456506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004132825549090082,
      "loss": 4.9832,
      "mean_token_accuracy": 0.21695564985275267,
      "num_tokens": 59088222.0,
      "step": 26470
    },
    {
      "entropy": 5.377591800689697,
      "epoch": 2.9734374122536082,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004132508814956595,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20933226495981216,
      "num_tokens": 59099302.0,
      "step": 26475
    },
    {
      "entropy": 5.3104654312133786,
      "epoch": 2.973998989161566,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004132192036803491,
      "loss": 4.9919,
      "mean_token_accuracy": 0.21135504096746444,
      "num_tokens": 59109238.0,
      "step": 26480
    },
    {
      "entropy": 5.2187415599823,
      "epoch": 2.974560566069523,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041318752146408575,
      "loss": 4.9351,
      "mean_token_accuracy": 0.21749639213085176,
      "num_tokens": 59119795.0,
      "step": 26485
    },
    {
      "entropy": 5.225346469879151,
      "epoch": 2.975122142977481,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004131558348478781,
      "loss": 4.997,
      "mean_token_accuracy": 0.20957348346710206,
      "num_tokens": 59129649.0,
      "step": 26490
    },
    {
      "entropy": 5.356751203536987,
      "epoch": 2.9756837198854384,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041312414383273507,
      "loss": 5.0707,
      "mean_token_accuracy": 0.20923053473234177,
      "num_tokens": 59140802.0,
      "step": 26495
    },
    {
      "entropy": 5.352287769317627,
      "epoch": 2.9762452967933957,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041309244841966565,
      "loss": 4.9543,
      "mean_token_accuracy": 0.21243455410003662,
      "num_tokens": 59151333.0,
      "step": 26500
    },
    {
      "entropy": 5.342854022979736,
      "epoch": 2.9768068737013533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041306074860967916,
      "loss": 5.001,
      "mean_token_accuracy": 0.20901533216238022,
      "num_tokens": 59163414.0,
      "step": 26505
    },
    {
      "entropy": 5.317862892150879,
      "epoch": 2.977368450609311,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00041302904440378494,
      "loss": 5.0267,
      "mean_token_accuracy": 0.2151505559682846,
      "num_tokens": 59173884.0,
      "step": 26510
    },
    {
      "entropy": 5.3434820652008055,
      "epoch": 2.9779300275172687,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041299733580299236,
      "loss": 5.0709,
      "mean_token_accuracy": 0.20641091465950012,
      "num_tokens": 59185263.0,
      "step": 26515
    },
    {
      "entropy": 5.409268569946289,
      "epoch": 2.978491604425226,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041296562280831106,
      "loss": 5.1377,
      "mean_token_accuracy": 0.20286353230476378,
      "num_tokens": 59196279.0,
      "step": 26520
    },
    {
      "entropy": 5.19509744644165,
      "epoch": 2.9790531813331835,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041293390542075095,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22210338711738586,
      "num_tokens": 59207537.0,
      "step": 26525
    },
    {
      "entropy": 5.236320734024048,
      "epoch": 2.979614758241141,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004129021836413217,
      "loss": 4.9622,
      "mean_token_accuracy": 0.21667500734329223,
      "num_tokens": 59218583.0,
      "step": 26530
    },
    {
      "entropy": 5.2854663848876955,
      "epoch": 2.980176335149099,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004128704574710334,
      "loss": 4.9085,
      "mean_token_accuracy": 0.21540388315916062,
      "num_tokens": 59228638.0,
      "step": 26535
    },
    {
      "entropy": 5.3975907325744625,
      "epoch": 2.980737912057056,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004128387269108963,
      "loss": 5.1294,
      "mean_token_accuracy": 0.2089935675263405,
      "num_tokens": 59240574.0,
      "step": 26540
    },
    {
      "entropy": 5.27879695892334,
      "epoch": 2.9812994889650137,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041280699196192074,
      "loss": 4.931,
      "mean_token_accuracy": 0.21540718227624894,
      "num_tokens": 59252986.0,
      "step": 26545
    },
    {
      "entropy": 5.304366636276245,
      "epoch": 2.9818610658729714,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00041277525262511705,
      "loss": 5.074,
      "mean_token_accuracy": 0.20503122508525848,
      "num_tokens": 59263034.0,
      "step": 26550
    },
    {
      "entropy": 5.37549204826355,
      "epoch": 2.9824226427809286,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041274350890149584,
      "loss": 5.0816,
      "mean_token_accuracy": 0.20575810819864274,
      "num_tokens": 59275170.0,
      "step": 26555
    },
    {
      "entropy": 5.369757986068725,
      "epoch": 2.9829842196888863,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000412711760792068,
      "loss": 5.0694,
      "mean_token_accuracy": 0.21420719772577285,
      "num_tokens": 59287168.0,
      "step": 26560
    },
    {
      "entropy": 5.3694658279418945,
      "epoch": 2.983545796596844,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004126800082978442,
      "loss": 5.0876,
      "mean_token_accuracy": 0.2161261186003685,
      "num_tokens": 59298043.0,
      "step": 26565
    },
    {
      "entropy": 5.246673822402954,
      "epoch": 2.9841073735048016,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041264825141983547,
      "loss": 4.8715,
      "mean_token_accuracy": 0.2186397910118103,
      "num_tokens": 59310454.0,
      "step": 26570
    },
    {
      "entropy": 5.39294023513794,
      "epoch": 2.9846689504127593,
      "grad_norm": 0.96875,
      "learning_rate": 0.00041261649015905314,
      "loss": 5.0688,
      "mean_token_accuracy": 0.2047386422753334,
      "num_tokens": 59323275.0,
      "step": 26575
    },
    {
      "entropy": 5.225599575042724,
      "epoch": 2.9852305273207165,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004125847245165084,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21350884139537812,
      "num_tokens": 59334003.0,
      "step": 26580
    },
    {
      "entropy": 5.2593896865844725,
      "epoch": 2.985792104228674,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004125529544932126,
      "loss": 4.9377,
      "mean_token_accuracy": 0.2174226626753807,
      "num_tokens": 59345403.0,
      "step": 26585
    },
    {
      "entropy": 5.332485103607178,
      "epoch": 2.986353681136632,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004125211800901774,
      "loss": 5.0168,
      "mean_token_accuracy": 0.21049659997224807,
      "num_tokens": 59357457.0,
      "step": 26590
    },
    {
      "entropy": 5.283299922943115,
      "epoch": 2.986915258044589,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004124894013084145,
      "loss": 4.9122,
      "mean_token_accuracy": 0.22144443392753602,
      "num_tokens": 59369059.0,
      "step": 26595
    },
    {
      "entropy": 5.287511730194092,
      "epoch": 2.9874768349525467,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00041245761814893576,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21874107122421266,
      "num_tokens": 59380700.0,
      "step": 26600
    },
    {
      "entropy": 5.2886768817901615,
      "epoch": 2.9880384118605043,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004124258306127531,
      "loss": 4.9588,
      "mean_token_accuracy": 0.20679187327623366,
      "num_tokens": 59391504.0,
      "step": 26605
    },
    {
      "entropy": 5.344568109512329,
      "epoch": 2.988599988768462,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004123940387008787,
      "loss": 5.0431,
      "mean_token_accuracy": 0.2080456167459488,
      "num_tokens": 59402648.0,
      "step": 26610
    },
    {
      "entropy": 5.329212331771851,
      "epoch": 2.9891615656764197,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004123622424143248,
      "loss": 5.0923,
      "mean_token_accuracy": 0.2063878133893013,
      "num_tokens": 59413842.0,
      "step": 26615
    },
    {
      "entropy": 5.3650284767150875,
      "epoch": 2.989723142584377,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00041233044175410387,
      "loss": 5.1099,
      "mean_token_accuracy": 0.21043772995471954,
      "num_tokens": 59425630.0,
      "step": 26620
    },
    {
      "entropy": 5.313165330886841,
      "epoch": 2.9902847194923345,
      "grad_norm": 1.0625,
      "learning_rate": 0.00041229863672122844,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21838281601667403,
      "num_tokens": 59437140.0,
      "step": 26625
    },
    {
      "entropy": 5.281536626815796,
      "epoch": 2.990846296400292,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004122668273167111,
      "loss": 4.8709,
      "mean_token_accuracy": 0.21920794993638992,
      "num_tokens": 59447629.0,
      "step": 26630
    },
    {
      "entropy": 5.259185838699341,
      "epoch": 2.9914078733082494,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004122350135415648,
      "loss": 5.0379,
      "mean_token_accuracy": 0.20771881192922592,
      "num_tokens": 59458469.0,
      "step": 26635
    },
    {
      "entropy": 5.302764081954956,
      "epoch": 2.991969450216207,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004122031953968024,
      "loss": 4.9647,
      "mean_token_accuracy": 0.2197152316570282,
      "num_tokens": 59468864.0,
      "step": 26640
    },
    {
      "entropy": 5.238961219787598,
      "epoch": 2.9925310271241647,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041217137288343704,
      "loss": 4.9759,
      "mean_token_accuracy": 0.21167472153902053,
      "num_tokens": 59480689.0,
      "step": 26645
    },
    {
      "entropy": 5.34367036819458,
      "epoch": 2.993092604032122,
      "grad_norm": 1.078125,
      "learning_rate": 0.000412139546002482,
      "loss": 5.0907,
      "mean_token_accuracy": 0.2099883198738098,
      "num_tokens": 59491510.0,
      "step": 26650
    },
    {
      "entropy": 5.3577516078948975,
      "epoch": 2.9936541809400796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041210771475495064,
      "loss": 5.0403,
      "mean_token_accuracy": 0.20663178116083145,
      "num_tokens": 59502838.0,
      "step": 26655
    },
    {
      "entropy": 5.348137140274048,
      "epoch": 2.9942157578480373,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004120758791418564,
      "loss": 5.0154,
      "mean_token_accuracy": 0.21593865156173705,
      "num_tokens": 59514356.0,
      "step": 26660
    },
    {
      "entropy": 5.272879552841187,
      "epoch": 2.994777334755995,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004120440391642131,
      "loss": 4.9924,
      "mean_token_accuracy": 0.2091165915131569,
      "num_tokens": 59525933.0,
      "step": 26665
    },
    {
      "entropy": 5.364923524856567,
      "epoch": 2.9953389116639526,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004120121948230343,
      "loss": 5.0901,
      "mean_token_accuracy": 0.21137455552816392,
      "num_tokens": 59536856.0,
      "step": 26670
    },
    {
      "entropy": 5.337873649597168,
      "epoch": 2.99590048857191,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004119803461193342,
      "loss": 5.0242,
      "mean_token_accuracy": 0.21671769320964812,
      "num_tokens": 59546023.0,
      "step": 26675
    },
    {
      "entropy": 5.291718053817749,
      "epoch": 2.9964620654798675,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004119484930541267,
      "loss": 4.9864,
      "mean_token_accuracy": 0.22094185948371886,
      "num_tokens": 59557515.0,
      "step": 26680
    },
    {
      "entropy": 5.249095249176025,
      "epoch": 2.997023642387825,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00041191663562842605,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20362828224897384,
      "num_tokens": 59570198.0,
      "step": 26685
    },
    {
      "entropy": 5.393795108795166,
      "epoch": 2.9975852192957824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004118847738432467,
      "loss": 5.1107,
      "mean_token_accuracy": 0.20814573615789414,
      "num_tokens": 59580365.0,
      "step": 26690
    },
    {
      "entropy": 5.2620924949646,
      "epoch": 2.99814679620374,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004118529076996029,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2151049017906189,
      "num_tokens": 59591304.0,
      "step": 26695
    },
    {
      "entropy": 5.331595468521118,
      "epoch": 2.9987083731116977,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041182103719850967,
      "loss": 5.0933,
      "mean_token_accuracy": 0.20791874676942826,
      "num_tokens": 59603005.0,
      "step": 26700
    },
    {
      "entropy": 5.245297002792358,
      "epoch": 2.9992699500196554,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041178916234098134,
      "loss": 4.9141,
      "mean_token_accuracy": 0.22075089663267136,
      "num_tokens": 59614211.0,
      "step": 26705
    },
    {
      "entropy": 5.270608520507812,
      "epoch": 2.999831526927613,
      "grad_norm": 1.0,
      "learning_rate": 0.0004117572831280332,
      "loss": 4.9779,
      "mean_token_accuracy": 0.20935892462730407,
      "num_tokens": 59624162.0,
      "step": 26710
    },
    {
      "entropy": 5.299653212229411,
      "epoch": 3.0003369461447744,
      "grad_norm": 1.078125,
      "learning_rate": 0.00041172539956068,
      "loss": 4.9408,
      "mean_token_accuracy": 0.20549897849559784,
      "num_tokens": 59633165.0,
      "step": 26715
    },
    {
      "entropy": 5.4137468338012695,
      "epoch": 3.000898523052732,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041169351163993705,
      "loss": 5.0099,
      "mean_token_accuracy": 0.20984359979629516,
      "num_tokens": 59645130.0,
      "step": 26720
    },
    {
      "entropy": 5.24041862487793,
      "epoch": 3.0014600999606897,
      "grad_norm": 0.984375,
      "learning_rate": 0.00041166161936681973,
      "loss": 4.873,
      "mean_token_accuracy": 0.2286424830555916,
      "num_tokens": 59656826.0,
      "step": 26725
    },
    {
      "entropy": 5.3089280128479,
      "epoch": 3.002021676868647,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004116297227423435,
      "loss": 5.03,
      "mean_token_accuracy": 0.2126267060637474,
      "num_tokens": 59667743.0,
      "step": 26730
    },
    {
      "entropy": 5.197865676879883,
      "epoch": 3.0025832537766046,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00041159782176752375,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22670254111289978,
      "num_tokens": 59679750.0,
      "step": 26735
    },
    {
      "entropy": 5.221135425567627,
      "epoch": 3.0031448306845623,
      "grad_norm": 1.171875,
      "learning_rate": 0.0004115659164433765,
      "loss": 4.885,
      "mean_token_accuracy": 0.21906069964170455,
      "num_tokens": 59689113.0,
      "step": 26740
    },
    {
      "entropy": 5.227197933197021,
      "epoch": 3.00370640759252,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041153400677091745,
      "loss": 4.8388,
      "mean_token_accuracy": 0.2213949218392372,
      "num_tokens": 59701282.0,
      "step": 26745
    },
    {
      "entropy": 5.291519260406494,
      "epoch": 3.004267984500477,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041150209275116266,
      "loss": 4.9404,
      "mean_token_accuracy": 0.21599286198616027,
      "num_tokens": 59713536.0,
      "step": 26750
    },
    {
      "entropy": 5.391168737411499,
      "epoch": 3.004829561408435,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041147017438512834,
      "loss": 5.0447,
      "mean_token_accuracy": 0.20811798572540283,
      "num_tokens": 59723818.0,
      "step": 26755
    },
    {
      "entropy": 5.393056106567383,
      "epoch": 3.0053911383163925,
      "grad_norm": 1.046875,
      "learning_rate": 0.00041143825167383073,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21167832762002944,
      "num_tokens": 59735486.0,
      "step": 26760
    },
    {
      "entropy": 5.3545066833496096,
      "epoch": 3.00595271522435,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004114063246182863,
      "loss": 4.9148,
      "mean_token_accuracy": 0.21444075405597687,
      "num_tokens": 59746006.0,
      "step": 26765
    },
    {
      "entropy": 5.32329683303833,
      "epoch": 3.0065142921323074,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041137439321951147,
      "loss": 4.985,
      "mean_token_accuracy": 0.21750040650367736,
      "num_tokens": 59756346.0,
      "step": 26770
    },
    {
      "entropy": 5.2770946502685545,
      "epoch": 3.007075869040265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00041134245747852314,
      "loss": 4.9126,
      "mean_token_accuracy": 0.22261320501565934,
      "num_tokens": 59768038.0,
      "step": 26775
    },
    {
      "entropy": 5.320449113845825,
      "epoch": 3.0076374459482227,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004113105173963382,
      "loss": 4.9698,
      "mean_token_accuracy": 0.21122576296329498,
      "num_tokens": 59779211.0,
      "step": 26780
    },
    {
      "entropy": 5.257280111312866,
      "epoch": 3.0081990228561803,
      "grad_norm": 1.0,
      "learning_rate": 0.0004112785729739733,
      "loss": 4.8959,
      "mean_token_accuracy": 0.2179553896188736,
      "num_tokens": 59790777.0,
      "step": 26785
    },
    {
      "entropy": 5.310897731781006,
      "epoch": 3.0087605997641376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041124662421244593,
      "loss": 4.9507,
      "mean_token_accuracy": 0.209494549036026,
      "num_tokens": 59802030.0,
      "step": 26790
    },
    {
      "entropy": 5.329885005950928,
      "epoch": 3.0093221766720952,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004112146711127732,
      "loss": 5.0179,
      "mean_token_accuracy": 0.21289959102869033,
      "num_tokens": 59814616.0,
      "step": 26795
    },
    {
      "entropy": 5.237038421630859,
      "epoch": 3.009883753580053,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004111827136759725,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21981438398361205,
      "num_tokens": 59825797.0,
      "step": 26800
    },
    {
      "entropy": 5.263896942138672,
      "epoch": 3.0104453304880106,
      "grad_norm": 1.0,
      "learning_rate": 0.0004111507519030613,
      "loss": 4.9533,
      "mean_token_accuracy": 0.220047265291214,
      "num_tokens": 59836699.0,
      "step": 26805
    },
    {
      "entropy": 5.330453586578369,
      "epoch": 3.0110069073959678,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004111187857950574,
      "loss": 5.0039,
      "mean_token_accuracy": 0.21611579656600952,
      "num_tokens": 59848575.0,
      "step": 26810
    },
    {
      "entropy": 5.272697162628174,
      "epoch": 3.0115684843039254,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004110868153529786,
      "loss": 4.8621,
      "mean_token_accuracy": 0.22565509080886842,
      "num_tokens": 59860471.0,
      "step": 26815
    },
    {
      "entropy": 5.289985656738281,
      "epoch": 3.012130061211883,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004110548405778428,
      "loss": 4.9576,
      "mean_token_accuracy": 0.22019245773553847,
      "num_tokens": 59871759.0,
      "step": 26820
    },
    {
      "entropy": 5.37222466468811,
      "epoch": 3.0126916381198403,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004110228614706681,
      "loss": 4.911,
      "mean_token_accuracy": 0.21389778703451157,
      "num_tokens": 59884375.0,
      "step": 26825
    },
    {
      "entropy": 5.294230318069458,
      "epoch": 3.013253215027798,
      "grad_norm": 1.0,
      "learning_rate": 0.00041099087803247276,
      "loss": 4.9569,
      "mean_token_accuracy": 0.21779309809207917,
      "num_tokens": 59895606.0,
      "step": 26830
    },
    {
      "entropy": 5.2731691837310795,
      "epoch": 3.0138147919357556,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004109588902642751,
      "loss": 4.9931,
      "mean_token_accuracy": 0.21219695061445237,
      "num_tokens": 59908090.0,
      "step": 26835
    },
    {
      "entropy": 5.298471879959107,
      "epoch": 3.0143763688437133,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004109268981670936,
      "loss": 4.9647,
      "mean_token_accuracy": 0.21612309962511062,
      "num_tokens": 59919218.0,
      "step": 26840
    },
    {
      "entropy": 5.347947835922241,
      "epoch": 3.0149379457516705,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000410894901741947,
      "loss": 5.0648,
      "mean_token_accuracy": 0.2189454510807991,
      "num_tokens": 59931195.0,
      "step": 26845
    },
    {
      "entropy": 5.33633975982666,
      "epoch": 3.015499522659628,
      "grad_norm": 1.0,
      "learning_rate": 0.000410862900989854,
      "loss": 4.9844,
      "mean_token_accuracy": 0.21480778604745865,
      "num_tokens": 59943653.0,
      "step": 26850
    },
    {
      "entropy": 5.3032448291778564,
      "epoch": 3.016061099567586,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004108308959118336,
      "loss": 4.9076,
      "mean_token_accuracy": 0.21558827459812163,
      "num_tokens": 59954677.0,
      "step": 26855
    },
    {
      "entropy": 5.27733359336853,
      "epoch": 3.0166226764755435,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004107988865089047,
      "loss": 5.0173,
      "mean_token_accuracy": 0.21431252658367156,
      "num_tokens": 59966135.0,
      "step": 26860
    },
    {
      "entropy": 5.264432382583618,
      "epoch": 3.0171842533835007,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00041076687278208664,
      "loss": 4.8151,
      "mean_token_accuracy": 0.22795940041542054,
      "num_tokens": 59977828.0,
      "step": 26865
    },
    {
      "entropy": 5.232567882537841,
      "epoch": 3.0177458302914584,
      "grad_norm": 1.0,
      "learning_rate": 0.0004107348547323987,
      "loss": 4.9189,
      "mean_token_accuracy": 0.22230804562568665,
      "num_tokens": 59989226.0,
      "step": 26870
    },
    {
      "entropy": 5.304464864730835,
      "epoch": 3.018307407199416,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00041070283236086024,
      "loss": 4.9161,
      "mean_token_accuracy": 0.2214004173874855,
      "num_tokens": 60000461.0,
      "step": 26875
    },
    {
      "entropy": 5.308046674728393,
      "epoch": 3.0188689841073737,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004106708056684911,
      "loss": 5.0652,
      "mean_token_accuracy": 0.21466798931360245,
      "num_tokens": 60012364.0,
      "step": 26880
    },
    {
      "entropy": 5.291409540176391,
      "epoch": 3.019430561015331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00041063877465631084,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21750674843788148,
      "num_tokens": 60022913.0,
      "step": 26885
    },
    {
      "entropy": 5.283042573928833,
      "epoch": 3.0199921379232886,
      "grad_norm": 1.0,
      "learning_rate": 0.0004106067393253394,
      "loss": 4.9268,
      "mean_token_accuracy": 0.2227788358926773,
      "num_tokens": 60034407.0,
      "step": 26890
    },
    {
      "entropy": 5.269110012054443,
      "epoch": 3.0205537148312462,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004105746996765968,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21141313165426254,
      "num_tokens": 60044063.0,
      "step": 26895
    },
    {
      "entropy": 5.287623023986816,
      "epoch": 3.021115291739204,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004105426557111031,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22343397736549378,
      "num_tokens": 60054432.0,
      "step": 26900
    },
    {
      "entropy": 5.3091024398803714,
      "epoch": 3.021676868647161,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004105106074298787,
      "loss": 4.9293,
      "mean_token_accuracy": 0.21464851945638658,
      "num_tokens": 60065876.0,
      "step": 26905
    },
    {
      "entropy": 5.294310855865478,
      "epoch": 3.022238445555119,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000410478554833944,
      "loss": 4.8977,
      "mean_token_accuracy": 0.21261270791292192,
      "num_tokens": 60076108.0,
      "step": 26910
    },
    {
      "entropy": 5.2359521865844725,
      "epoch": 3.0228000224630764,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00041044649792431963,
      "loss": 4.9042,
      "mean_token_accuracy": 0.21886456310749053,
      "num_tokens": 60087386.0,
      "step": 26915
    },
    {
      "entropy": 5.2752526760101315,
      "epoch": 3.0233615993710337,
      "grad_norm": 1.015625,
      "learning_rate": 0.00041041443670202616,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22720755338668824,
      "num_tokens": 60099017.0,
      "step": 26920
    },
    {
      "entropy": 5.266715860366821,
      "epoch": 3.0239231762789913,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004103823711680846,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22143120616674422,
      "num_tokens": 60110676.0,
      "step": 26925
    },
    {
      "entropy": 5.206411361694336,
      "epoch": 3.024484753186949,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00041035030132351575,
      "loss": 4.8918,
      "mean_token_accuracy": 0.2259598806500435,
      "num_tokens": 60120807.0,
      "step": 26930
    },
    {
      "entropy": 5.302015972137451,
      "epoch": 3.0250463300949066,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00041031822716934086,
      "loss": 4.8874,
      "mean_token_accuracy": 0.21306771636009217,
      "num_tokens": 60132174.0,
      "step": 26935
    },
    {
      "entropy": 5.364853096008301,
      "epoch": 3.025607907002864,
      "grad_norm": 1.03125,
      "learning_rate": 0.00041028614870658106,
      "loss": 5.0743,
      "mean_token_accuracy": 0.20861130058765412,
      "num_tokens": 60144311.0,
      "step": 26940
    },
    {
      "entropy": 5.266226291656494,
      "epoch": 3.0261694839108215,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004102540659362579,
      "loss": 4.9928,
      "mean_token_accuracy": 0.21295495927333832,
      "num_tokens": 60155368.0,
      "step": 26945
    },
    {
      "entropy": 5.353194952011108,
      "epoch": 3.026731060818779,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00041022197885939285,
      "loss": 5.026,
      "mean_token_accuracy": 0.2099367454648018,
      "num_tokens": 60167633.0,
      "step": 26950
    },
    {
      "entropy": 5.273362445831299,
      "epoch": 3.027292637726737,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00041018988747700745,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22138795852661133,
      "num_tokens": 60180596.0,
      "step": 26955
    },
    {
      "entropy": 5.353603076934815,
      "epoch": 3.027854214634694,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004101577917901236,
      "loss": 5.0078,
      "mean_token_accuracy": 0.2094623401761055,
      "num_tokens": 60191811.0,
      "step": 26960
    },
    {
      "entropy": 5.382225847244262,
      "epoch": 3.0284157915426517,
      "grad_norm": 1.125,
      "learning_rate": 0.00041012569179976326,
      "loss": 4.9267,
      "mean_token_accuracy": 0.21784452050924302,
      "num_tokens": 60202575.0,
      "step": 26965
    },
    {
      "entropy": 5.2621244430542,
      "epoch": 3.0289773684506094,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004100935875069485,
      "loss": 4.882,
      "mean_token_accuracy": 0.2206817552447319,
      "num_tokens": 60212541.0,
      "step": 26970
    },
    {
      "entropy": 5.278296327590942,
      "epoch": 3.029538945358567,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004100614789127015,
      "loss": 4.9736,
      "mean_token_accuracy": 0.21695628464221955,
      "num_tokens": 60223018.0,
      "step": 26975
    },
    {
      "entropy": 5.347899675369263,
      "epoch": 3.0301005222665243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004100293660180447,
      "loss": 4.9866,
      "mean_token_accuracy": 0.2161340370774269,
      "num_tokens": 60234104.0,
      "step": 26980
    },
    {
      "entropy": 5.2189099311828615,
      "epoch": 3.030662099174482,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040999724882400046,
      "loss": 4.941,
      "mean_token_accuracy": 0.2183496668934822,
      "num_tokens": 60245119.0,
      "step": 26985
    },
    {
      "entropy": 5.296947193145752,
      "epoch": 3.0312236760824396,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004099651273315915,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22462039142847062,
      "num_tokens": 60255984.0,
      "step": 26990
    },
    {
      "entropy": 5.333605861663818,
      "epoch": 3.0317852529903973,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004099330015418405,
      "loss": 4.9207,
      "mean_token_accuracy": 0.2245437279343605,
      "num_tokens": 60266764.0,
      "step": 26995
    },
    {
      "entropy": 5.221853542327881,
      "epoch": 3.0323468298983545,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040990087145577044,
      "loss": 4.904,
      "mean_token_accuracy": 0.2231446996331215,
      "num_tokens": 60278284.0,
      "step": 27000
    },
    {
      "epoch": 3.0323468298983545,
      "eval_entropy": 5.070939621176337,
      "eval_loss": 5.1366729736328125,
      "eval_mean_token_accuracy": 0.2150181639216012,
      "eval_num_tokens": 60278284.0,
      "eval_runtime": 23.9013,
      "eval_samples_per_second": 1297.544,
      "eval_steps_per_second": 162.209,
      "step": 27000
    },
    {
      "entropy": 5.242385005950927,
      "epoch": 3.032908406806312,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004098687370744042,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22299651950597763,
      "num_tokens": 60287993.0,
      "step": 27005
    },
    {
      "entropy": 5.332368278503418,
      "epoch": 3.03346998371427,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004098365983987652,
      "loss": 5.0432,
      "mean_token_accuracy": 0.2113705739378929,
      "num_tokens": 60300040.0,
      "step": 27010
    },
    {
      "entropy": 5.314341878890991,
      "epoch": 3.034031560622227,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040980445542987653,
      "loss": 4.9496,
      "mean_token_accuracy": 0.21754564344882965,
      "num_tokens": 60311008.0,
      "step": 27015
    },
    {
      "entropy": 5.293119096755982,
      "epoch": 3.0345931375301847,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040977230816876176,
      "loss": 4.8767,
      "mean_token_accuracy": 0.21696203649044038,
      "num_tokens": 60321071.0,
      "step": 27020
    },
    {
      "entropy": 5.27727108001709,
      "epoch": 3.0351547144381423,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040974015661644433,
      "loss": 4.9255,
      "mean_token_accuracy": 0.21422241777181625,
      "num_tokens": 60333075.0,
      "step": 27025
    },
    {
      "entropy": 5.318404483795166,
      "epoch": 3.0357162913461,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004097080007739481,
      "loss": 5.0029,
      "mean_token_accuracy": 0.21117625683546065,
      "num_tokens": 60343393.0,
      "step": 27030
    },
    {
      "entropy": 5.283148574829101,
      "epoch": 3.036277868254057,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040967584064229695,
      "loss": 5.0106,
      "mean_token_accuracy": 0.20701437294483185,
      "num_tokens": 60353981.0,
      "step": 27035
    },
    {
      "entropy": 5.318243741989136,
      "epoch": 3.036839445162015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040964367622251473,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22187511175870894,
      "num_tokens": 60364544.0,
      "step": 27040
    },
    {
      "entropy": 5.3469500064849855,
      "epoch": 3.0374010220699725,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004096115075156257,
      "loss": 4.994,
      "mean_token_accuracy": 0.22056198120117188,
      "num_tokens": 60376031.0,
      "step": 27045
    },
    {
      "entropy": 5.313948011398315,
      "epoch": 3.03796259897793,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004095793345226541,
      "loss": 5.0151,
      "mean_token_accuracy": 0.20754031836986542,
      "num_tokens": 60387320.0,
      "step": 27050
    },
    {
      "entropy": 5.280464172363281,
      "epoch": 3.0385241758858874,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004095471572446242,
      "loss": 4.9279,
      "mean_token_accuracy": 0.22071481943130494,
      "num_tokens": 60397345.0,
      "step": 27055
    },
    {
      "entropy": 5.259451389312744,
      "epoch": 3.039085752793845,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004095149756825606,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21228049844503402,
      "num_tokens": 60408194.0,
      "step": 27060
    },
    {
      "entropy": 5.3609662532806395,
      "epoch": 3.0396473297018027,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004094827898374881,
      "loss": 4.9978,
      "mean_token_accuracy": 0.21365696489810942,
      "num_tokens": 60418928.0,
      "step": 27065
    },
    {
      "entropy": 5.290432453155518,
      "epoch": 3.0402089066097604,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004094505997104314,
      "loss": 4.834,
      "mean_token_accuracy": 0.22382923811674119,
      "num_tokens": 60428819.0,
      "step": 27070
    },
    {
      "entropy": 5.1829570770263675,
      "epoch": 3.0407704835177176,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040941840530241537,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2221047982573509,
      "num_tokens": 60440575.0,
      "step": 27075
    },
    {
      "entropy": 5.287409496307373,
      "epoch": 3.0413320604256753,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040938620661446526,
      "loss": 4.9676,
      "mean_token_accuracy": 0.21258632987737655,
      "num_tokens": 60452977.0,
      "step": 27080
    },
    {
      "entropy": 5.340865802764893,
      "epoch": 3.041893637333633,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004093540036476062,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2141840547323227,
      "num_tokens": 60463552.0,
      "step": 27085
    },
    {
      "entropy": 5.342369842529297,
      "epoch": 3.0424552142415906,
      "grad_norm": 1.0,
      "learning_rate": 0.00040932179640286356,
      "loss": 4.9833,
      "mean_token_accuracy": 0.2157636269927025,
      "num_tokens": 60474674.0,
      "step": 27090
    },
    {
      "entropy": 5.267340230941772,
      "epoch": 3.043016791149548,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004092895848812628,
      "loss": 4.93,
      "mean_token_accuracy": 0.2233830139040947,
      "num_tokens": 60486140.0,
      "step": 27095
    },
    {
      "entropy": 5.30108380317688,
      "epoch": 3.0435783680575055,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040925736908382953,
      "loss": 4.9146,
      "mean_token_accuracy": 0.22269794344902039,
      "num_tokens": 60497386.0,
      "step": 27100
    },
    {
      "entropy": 5.351088380813598,
      "epoch": 3.044139944965463,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004092251490115896,
      "loss": 5.0451,
      "mean_token_accuracy": 0.20471261888742448,
      "num_tokens": 60508535.0,
      "step": 27105
    },
    {
      "entropy": 5.30300817489624,
      "epoch": 3.0447015218734204,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040919292466556895,
      "loss": 4.9959,
      "mean_token_accuracy": 0.20773118138313293,
      "num_tokens": 60520791.0,
      "step": 27110
    },
    {
      "entropy": 5.285543632507324,
      "epoch": 3.045263098781378,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004091606960467934,
      "loss": 4.9019,
      "mean_token_accuracy": 0.2160048007965088,
      "num_tokens": 60533676.0,
      "step": 27115
    },
    {
      "entropy": 5.310209512710571,
      "epoch": 3.0458246756893357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040912846315628934,
      "loss": 4.972,
      "mean_token_accuracy": 0.21514423936605453,
      "num_tokens": 60545216.0,
      "step": 27120
    },
    {
      "entropy": 5.1877827644348145,
      "epoch": 3.0463862525972933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000409096225995083,
      "loss": 4.8753,
      "mean_token_accuracy": 0.22178203612565994,
      "num_tokens": 60554730.0,
      "step": 27125
    },
    {
      "entropy": 5.1814384937286375,
      "epoch": 3.0469478295052506,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040906398456420077,
      "loss": 4.8052,
      "mean_token_accuracy": 0.21709456741809846,
      "num_tokens": 60565494.0,
      "step": 27130
    },
    {
      "entropy": 5.263485240936279,
      "epoch": 3.047509406413208,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040903173886466934,
      "loss": 4.9374,
      "mean_token_accuracy": 0.21262044906616212,
      "num_tokens": 60576249.0,
      "step": 27135
    },
    {
      "entropy": 5.2372777462005615,
      "epoch": 3.048070983321166,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040899948889751525,
      "loss": 4.8753,
      "mean_token_accuracy": 0.2222362369298935,
      "num_tokens": 60587253.0,
      "step": 27140
    },
    {
      "entropy": 5.342047691345215,
      "epoch": 3.0486325602291235,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040896723466376557,
      "loss": 4.9969,
      "mean_token_accuracy": 0.20758214145898818,
      "num_tokens": 60598158.0,
      "step": 27145
    },
    {
      "entropy": 5.324070596694947,
      "epoch": 3.0491941371370808,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00040893497616444714,
      "loss": 4.8654,
      "mean_token_accuracy": 0.22577945590019227,
      "num_tokens": 60608812.0,
      "step": 27150
    },
    {
      "entropy": 5.2503194332122805,
      "epoch": 3.0497557140450384,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004089027134005871,
      "loss": 4.8924,
      "mean_token_accuracy": 0.2255984663963318,
      "num_tokens": 60619264.0,
      "step": 27155
    },
    {
      "entropy": 5.168814611434937,
      "epoch": 3.050317290952996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004088704463732127,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22378087788820267,
      "num_tokens": 60629688.0,
      "step": 27160
    },
    {
      "entropy": 5.2853881359100345,
      "epoch": 3.0508788678609537,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004088381750833514,
      "loss": 4.8892,
      "mean_token_accuracy": 0.2194615423679352,
      "num_tokens": 60640790.0,
      "step": 27165
    },
    {
      "entropy": 5.309791469573975,
      "epoch": 3.051440444768911,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004088058995320307,
      "loss": 4.8777,
      "mean_token_accuracy": 0.21943918764591216,
      "num_tokens": 60651952.0,
      "step": 27170
    },
    {
      "entropy": 5.3356804847717285,
      "epoch": 3.0520020216768686,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004087736197202784,
      "loss": 5.0218,
      "mean_token_accuracy": 0.21208076924085617,
      "num_tokens": 60665388.0,
      "step": 27175
    },
    {
      "entropy": 5.224979734420776,
      "epoch": 3.0525635985848263,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040874133564912195,
      "loss": 4.8413,
      "mean_token_accuracy": 0.22835836410522461,
      "num_tokens": 60676407.0,
      "step": 27180
    },
    {
      "entropy": 5.2648204326629635,
      "epoch": 3.053125175492784,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040870904731958965,
      "loss": 4.879,
      "mean_token_accuracy": 0.21556126028299333,
      "num_tokens": 60686796.0,
      "step": 27185
    },
    {
      "entropy": 5.316006708145141,
      "epoch": 3.053686752400741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004086767547327094,
      "loss": 4.994,
      "mean_token_accuracy": 0.21388425976037978,
      "num_tokens": 60697770.0,
      "step": 27190
    },
    {
      "entropy": 5.283238554000855,
      "epoch": 3.054248329308699,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00040864445788950935,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2165680095553398,
      "num_tokens": 60709059.0,
      "step": 27195
    },
    {
      "entropy": 5.258066177368164,
      "epoch": 3.0548099062166565,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000408612156791018,
      "loss": 4.9113,
      "mean_token_accuracy": 0.21936270445585251,
      "num_tokens": 60720228.0,
      "step": 27200
    },
    {
      "entropy": 5.3714494705200195,
      "epoch": 3.0553714831246137,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004085798514382637,
      "loss": 5.0253,
      "mean_token_accuracy": 0.21310136020183562,
      "num_tokens": 60730658.0,
      "step": 27205
    },
    {
      "entropy": 5.325200080871582,
      "epoch": 3.0559330600325714,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040854754183227516,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2203901931643486,
      "num_tokens": 60741757.0,
      "step": 27210
    },
    {
      "entropy": 5.258681440353394,
      "epoch": 3.056494636940529,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004085152279740811,
      "loss": 4.9432,
      "mean_token_accuracy": 0.21728675216436386,
      "num_tokens": 60755452.0,
      "step": 27215
    },
    {
      "entropy": 5.293448114395142,
      "epoch": 3.0570562138484867,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040848290986471037,
      "loss": 4.9554,
      "mean_token_accuracy": 0.21653492599725724,
      "num_tokens": 60765715.0,
      "step": 27220
    },
    {
      "entropy": 5.316315460205078,
      "epoch": 3.057617790756444,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040845058750519205,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21597348600625993,
      "num_tokens": 60777021.0,
      "step": 27225
    },
    {
      "entropy": 5.2274558544158936,
      "epoch": 3.0581793676644016,
      "grad_norm": 1.0,
      "learning_rate": 0.00040841826089655525,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2181097149848938,
      "num_tokens": 60787309.0,
      "step": 27230
    },
    {
      "entropy": 5.305422830581665,
      "epoch": 3.0587409445723592,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040838593003982926,
      "loss": 4.8694,
      "mean_token_accuracy": 0.21525210440158843,
      "num_tokens": 60797566.0,
      "step": 27235
    },
    {
      "entropy": 5.348981857299805,
      "epoch": 3.059302521480317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004083535949360436,
      "loss": 4.9466,
      "mean_token_accuracy": 0.22147813737392424,
      "num_tokens": 60808213.0,
      "step": 27240
    },
    {
      "entropy": 5.284787797927857,
      "epoch": 3.059864098388274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004083212555862277,
      "loss": 4.9808,
      "mean_token_accuracy": 0.21393701285123826,
      "num_tokens": 60820656.0,
      "step": 27245
    },
    {
      "entropy": 5.2887632846832275,
      "epoch": 3.0604256752962318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004082889119914113,
      "loss": 4.998,
      "mean_token_accuracy": 0.22194674760103225,
      "num_tokens": 60831572.0,
      "step": 27250
    },
    {
      "entropy": 5.2301562309265135,
      "epoch": 3.0609872522041894,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004082565641526242,
      "loss": 4.8323,
      "mean_token_accuracy": 0.22599599063396453,
      "num_tokens": 60842443.0,
      "step": 27255
    },
    {
      "entropy": 5.225492000579834,
      "epoch": 3.061548829112147,
      "grad_norm": 1.0,
      "learning_rate": 0.00040822421207089654,
      "loss": 4.8795,
      "mean_token_accuracy": 0.2225996047258377,
      "num_tokens": 60853427.0,
      "step": 27260
    },
    {
      "entropy": 5.323784828186035,
      "epoch": 3.0621104060201043,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004081918557472582,
      "loss": 5.0326,
      "mean_token_accuracy": 0.2120475262403488,
      "num_tokens": 60864517.0,
      "step": 27265
    },
    {
      "entropy": 5.356295013427735,
      "epoch": 3.062671982928062,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004081594951827396,
      "loss": 4.9834,
      "mean_token_accuracy": 0.21667075008153916,
      "num_tokens": 60874280.0,
      "step": 27270
    },
    {
      "entropy": 5.326799011230468,
      "epoch": 3.0632335598360196,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040812713037837096,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21464948803186418,
      "num_tokens": 60885656.0,
      "step": 27275
    },
    {
      "entropy": 5.321029138565064,
      "epoch": 3.0637951367439773,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040809476133518284,
      "loss": 4.9476,
      "mean_token_accuracy": 0.21719769835472108,
      "num_tokens": 60896856.0,
      "step": 27280
    },
    {
      "entropy": 5.209444952011109,
      "epoch": 3.0643567136519345,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00040806238805420595,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22314758598804474,
      "num_tokens": 60907464.0,
      "step": 27285
    },
    {
      "entropy": 5.3097363948822025,
      "epoch": 3.064918290559892,
      "grad_norm": 1.015625,
      "learning_rate": 0.000408030010536471,
      "loss": 4.9638,
      "mean_token_accuracy": 0.21182450354099275,
      "num_tokens": 60918321.0,
      "step": 27290
    },
    {
      "entropy": 5.325767660140992,
      "epoch": 3.06547986746785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040799762878300895,
      "loss": 4.984,
      "mean_token_accuracy": 0.20377658158540726,
      "num_tokens": 60928823.0,
      "step": 27295
    },
    {
      "entropy": 5.375059127807617,
      "epoch": 3.066041444375807,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040796524279485084,
      "loss": 5.0615,
      "mean_token_accuracy": 0.21281849294900895,
      "num_tokens": 60940161.0,
      "step": 27300
    },
    {
      "entropy": 5.324769306182861,
      "epoch": 3.0666030212837647,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040793285257302776,
      "loss": 5.0173,
      "mean_token_accuracy": 0.2035442739725113,
      "num_tokens": 60951968.0,
      "step": 27305
    },
    {
      "entropy": 5.394548749923706,
      "epoch": 3.0671645981917224,
      "grad_norm": 0.96875,
      "learning_rate": 0.000407900458118571,
      "loss": 5.0066,
      "mean_token_accuracy": 0.20765697956085205,
      "num_tokens": 60963619.0,
      "step": 27310
    },
    {
      "entropy": 5.327590847015381,
      "epoch": 3.06772617509968,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004078680594325123,
      "loss": 4.9269,
      "mean_token_accuracy": 0.21886589080095292,
      "num_tokens": 60974615.0,
      "step": 27315
    },
    {
      "entropy": 5.299466562271118,
      "epoch": 3.0682877520076373,
      "grad_norm": 1.0625,
      "learning_rate": 0.000407835656515883,
      "loss": 5.024,
      "mean_token_accuracy": 0.2087086856365204,
      "num_tokens": 60986943.0,
      "step": 27320
    },
    {
      "entropy": 5.308781147003174,
      "epoch": 3.068849328915595,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040780324936971486,
      "loss": 5.0544,
      "mean_token_accuracy": 0.21558828949928283,
      "num_tokens": 60998306.0,
      "step": 27325
    },
    {
      "entropy": 5.367514896392822,
      "epoch": 3.0694109058235526,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004077708379950397,
      "loss": 4.9722,
      "mean_token_accuracy": 0.21178533136844635,
      "num_tokens": 61010102.0,
      "step": 27330
    },
    {
      "entropy": 5.368476533889771,
      "epoch": 3.0699724827315102,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004077384223928896,
      "loss": 4.9413,
      "mean_token_accuracy": 0.21096652299165725,
      "num_tokens": 61019474.0,
      "step": 27335
    },
    {
      "entropy": 5.239809560775757,
      "epoch": 3.0705340596394675,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004077060025642967,
      "loss": 4.9395,
      "mean_token_accuracy": 0.21958870440721512,
      "num_tokens": 61030401.0,
      "step": 27340
    },
    {
      "entropy": 5.397114896774292,
      "epoch": 3.071095636547425,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004076735785102931,
      "loss": 5.0219,
      "mean_token_accuracy": 0.2120201274752617,
      "num_tokens": 61042139.0,
      "step": 27345
    },
    {
      "entropy": 5.274952173233032,
      "epoch": 3.071657213455383,
      "grad_norm": 1.0,
      "learning_rate": 0.0004076411502319114,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2223646327853203,
      "num_tokens": 61052468.0,
      "step": 27350
    },
    {
      "entropy": 5.309836196899414,
      "epoch": 3.0722187903633404,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040760871773018405,
      "loss": 4.9225,
      "mean_token_accuracy": 0.21765222102403642,
      "num_tokens": 61063625.0,
      "step": 27355
    },
    {
      "entropy": 5.2137542247772215,
      "epoch": 3.0727803672712977,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004075762810061437,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22322998344898223,
      "num_tokens": 61074246.0,
      "step": 27360
    },
    {
      "entropy": 5.317661190032959,
      "epoch": 3.0733419441792553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000407543840060823,
      "loss": 5.0129,
      "mean_token_accuracy": 0.20634758323431016,
      "num_tokens": 61086096.0,
      "step": 27365
    },
    {
      "entropy": 5.304735279083252,
      "epoch": 3.073903521087213,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0004075113948952551,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22013984024524688,
      "num_tokens": 61097176.0,
      "step": 27370
    },
    {
      "entropy": 5.3826415061950685,
      "epoch": 3.0744650979951706,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004074789455104731,
      "loss": 5.0776,
      "mean_token_accuracy": 0.21531172841787338,
      "num_tokens": 61107865.0,
      "step": 27375
    },
    {
      "entropy": 5.341491842269898,
      "epoch": 3.075026674903128,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040744649190750995,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21859110891819,
      "num_tokens": 61118682.0,
      "step": 27380
    },
    {
      "entropy": 5.262582206726075,
      "epoch": 3.0755882518110855,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004074140340873992,
      "loss": 4.8902,
      "mean_token_accuracy": 0.2291736990213394,
      "num_tokens": 61130210.0,
      "step": 27385
    },
    {
      "entropy": 5.218521928787231,
      "epoch": 3.076149828719043,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004073815720511743,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21491551846265794,
      "num_tokens": 61141216.0,
      "step": 27390
    },
    {
      "entropy": 5.31801815032959,
      "epoch": 3.0767114056270004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040734910579986875,
      "loss": 4.9962,
      "mean_token_accuracy": 0.21597783267498016,
      "num_tokens": 61153233.0,
      "step": 27395
    },
    {
      "entropy": 5.33681435585022,
      "epoch": 3.077272982534958,
      "grad_norm": 1.0,
      "learning_rate": 0.00040731663533451624,
      "loss": 4.8994,
      "mean_token_accuracy": 0.2243298262357712,
      "num_tokens": 61163044.0,
      "step": 27400
    },
    {
      "entropy": 5.227513694763184,
      "epoch": 3.0778345594429157,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00040728416065615087,
      "loss": 4.978,
      "mean_token_accuracy": 0.21316144615411758,
      "num_tokens": 61174595.0,
      "step": 27405
    },
    {
      "entropy": 5.252131557464599,
      "epoch": 3.0783961363508734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004072516817658065,
      "loss": 4.9185,
      "mean_token_accuracy": 0.2160241574048996,
      "num_tokens": 61185215.0,
      "step": 27410
    },
    {
      "entropy": 5.2458281993865965,
      "epoch": 3.0789577132588306,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040721919866451724,
      "loss": 4.7768,
      "mean_token_accuracy": 0.22914662659168245,
      "num_tokens": 61195504.0,
      "step": 27415
    },
    {
      "entropy": 5.206886291503906,
      "epoch": 3.0795192901667883,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004071867113533174,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22024173885583878,
      "num_tokens": 61206040.0,
      "step": 27420
    },
    {
      "entropy": 5.321932458877564,
      "epoch": 3.080080867074746,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040715421983324147,
      "loss": 5.0021,
      "mean_token_accuracy": 0.21153566986322403,
      "num_tokens": 61217797.0,
      "step": 27425
    },
    {
      "entropy": 5.363031005859375,
      "epoch": 3.0806424439827036,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00040712172410532397,
      "loss": 4.9918,
      "mean_token_accuracy": 0.2148149162530899,
      "num_tokens": 61230158.0,
      "step": 27430
    },
    {
      "entropy": 5.3415509223937985,
      "epoch": 3.081204020890661,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004070892241705994,
      "loss": 4.9441,
      "mean_token_accuracy": 0.22021480947732924,
      "num_tokens": 61241072.0,
      "step": 27435
    },
    {
      "entropy": 5.153861474990845,
      "epoch": 3.0817655977986185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040705672003010275,
      "loss": 4.815,
      "mean_token_accuracy": 0.2240742489695549,
      "num_tokens": 61251057.0,
      "step": 27440
    },
    {
      "entropy": 5.180303573608398,
      "epoch": 3.082327174706576,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000407024211684869,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2197699263691902,
      "num_tokens": 61262729.0,
      "step": 27445
    },
    {
      "entropy": 5.303693437576294,
      "epoch": 3.082888751614534,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040699169913593306,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2262299418449402,
      "num_tokens": 61274665.0,
      "step": 27450
    },
    {
      "entropy": 5.329204559326172,
      "epoch": 3.083450328522491,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004069591823843302,
      "loss": 4.9332,
      "mean_token_accuracy": 0.2164851114153862,
      "num_tokens": 61285244.0,
      "step": 27455
    },
    {
      "entropy": 5.269337558746338,
      "epoch": 3.0840119054304487,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004069266614310959,
      "loss": 4.9592,
      "mean_token_accuracy": 0.21682755500078202,
      "num_tokens": 61295739.0,
      "step": 27460
    },
    {
      "entropy": 5.272413301467895,
      "epoch": 3.0845734823384063,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040689413627726544,
      "loss": 4.8838,
      "mean_token_accuracy": 0.22113099545240403,
      "num_tokens": 61305964.0,
      "step": 27465
    },
    {
      "entropy": 5.331844806671143,
      "epoch": 3.085135059246364,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040686160692387455,
      "loss": 5.0208,
      "mean_token_accuracy": 0.20951367765665055,
      "num_tokens": 61317004.0,
      "step": 27470
    },
    {
      "entropy": 5.363018560409546,
      "epoch": 3.085696636154321,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00040682907337195894,
      "loss": 5.0584,
      "mean_token_accuracy": 0.2052936941385269,
      "num_tokens": 61329014.0,
      "step": 27475
    },
    {
      "entropy": 5.252956914901733,
      "epoch": 3.086258213062279,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004067965356225545,
      "loss": 4.9052,
      "mean_token_accuracy": 0.2180042877793312,
      "num_tokens": 61340522.0,
      "step": 27480
    },
    {
      "entropy": 5.327788066864014,
      "epoch": 3.0868197899702365,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040676399367669733,
      "loss": 4.9572,
      "mean_token_accuracy": 0.2168538600206375,
      "num_tokens": 61351595.0,
      "step": 27485
    },
    {
      "entropy": 5.352351665496826,
      "epoch": 3.0873813668781938,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004067314475354234,
      "loss": 4.983,
      "mean_token_accuracy": 0.2155556008219719,
      "num_tokens": 61362432.0,
      "step": 27490
    },
    {
      "entropy": 5.283454847335816,
      "epoch": 3.0879429437861514,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040669889719976916,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22042381316423415,
      "num_tokens": 61372717.0,
      "step": 27495
    },
    {
      "entropy": 5.285530471801758,
      "epoch": 3.088504520694109,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004066663426707709,
      "loss": 4.916,
      "mean_token_accuracy": 0.21332804560661317,
      "num_tokens": 61383150.0,
      "step": 27500
    },
    {
      "entropy": 5.314427232742309,
      "epoch": 3.0890660976020667,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040663378394946534,
      "loss": 5.0645,
      "mean_token_accuracy": 0.2053431898355484,
      "num_tokens": 61395538.0,
      "step": 27505
    },
    {
      "entropy": 5.345777416229248,
      "epoch": 3.089627674510024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004066012210368889,
      "loss": 4.9402,
      "mean_token_accuracy": 0.2172325313091278,
      "num_tokens": 61407523.0,
      "step": 27510
    },
    {
      "entropy": 5.325271511077881,
      "epoch": 3.0901892514179816,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004065686539340786,
      "loss": 4.9249,
      "mean_token_accuracy": 0.20889340043067933,
      "num_tokens": 61418680.0,
      "step": 27515
    },
    {
      "entropy": 5.24090609550476,
      "epoch": 3.0907508283259393,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040653608264207134,
      "loss": 4.9291,
      "mean_token_accuracy": 0.21800231635570527,
      "num_tokens": 61430006.0,
      "step": 27520
    },
    {
      "entropy": 5.1727875709533695,
      "epoch": 3.091312405233897,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040650350716190417,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23213435411453248,
      "num_tokens": 61440564.0,
      "step": 27525
    },
    {
      "entropy": 5.1908698081970215,
      "epoch": 3.091873982141854,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004064709274946143,
      "loss": 4.7843,
      "mean_token_accuracy": 0.21905486136674882,
      "num_tokens": 61451589.0,
      "step": 27530
    },
    {
      "entropy": 5.182928657531738,
      "epoch": 3.092435559049812,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004064383436412392,
      "loss": 4.7894,
      "mean_token_accuracy": 0.23526723086833953,
      "num_tokens": 61462177.0,
      "step": 27535
    },
    {
      "entropy": 5.330151844024658,
      "epoch": 3.0929971359577695,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004064057556028162,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21723366975784303,
      "num_tokens": 61472626.0,
      "step": 27540
    },
    {
      "entropy": 5.232186269760132,
      "epoch": 3.093558712865727,
      "grad_norm": 1.015625,
      "learning_rate": 0.000406373163380383,
      "loss": 4.8507,
      "mean_token_accuracy": 0.22203682363033295,
      "num_tokens": 61483839.0,
      "step": 27545
    },
    {
      "entropy": 5.2619194984436035,
      "epoch": 3.0941202897736844,
      "grad_norm": 0.90625,
      "learning_rate": 0.00040634056697497726,
      "loss": 4.9565,
      "mean_token_accuracy": 0.22272822260856628,
      "num_tokens": 61495099.0,
      "step": 27550
    },
    {
      "entropy": 5.296092557907104,
      "epoch": 3.094681866681642,
      "grad_norm": 1.015625,
      "learning_rate": 0.000406307966387637,
      "loss": 4.9368,
      "mean_token_accuracy": 0.22227703034877777,
      "num_tokens": 61506902.0,
      "step": 27555
    },
    {
      "entropy": 5.3051536083221436,
      "epoch": 3.0952434435895997,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040627536161940016,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21018146723508835,
      "num_tokens": 61519357.0,
      "step": 27560
    },
    {
      "entropy": 5.318592357635498,
      "epoch": 3.0958050204975573,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004062427526713049,
      "loss": 4.8328,
      "mean_token_accuracy": 0.22799131274223328,
      "num_tokens": 61529510.0,
      "step": 27565
    },
    {
      "entropy": 5.193221807479858,
      "epoch": 3.0963665974055146,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004062101395443894,
      "loss": 4.793,
      "mean_token_accuracy": 0.223403200507164,
      "num_tokens": 61539679.0,
      "step": 27570
    },
    {
      "entropy": 5.192599058151245,
      "epoch": 3.0969281743134722,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004061775222396922,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22579479217529297,
      "num_tokens": 61549488.0,
      "step": 27575
    },
    {
      "entropy": 5.2225258350372314,
      "epoch": 3.09748975122143,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004061449007582519,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21230133324861528,
      "num_tokens": 61561167.0,
      "step": 27580
    },
    {
      "entropy": 5.3065143585205075,
      "epoch": 3.098051328129387,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000406112275101107,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22390414029359818,
      "num_tokens": 61572730.0,
      "step": 27585
    },
    {
      "entropy": 5.237852716445923,
      "epoch": 3.0986129050373448,
      "grad_norm": 1.0,
      "learning_rate": 0.00040607964526929655,
      "loss": 4.8695,
      "mean_token_accuracy": 0.22113245278596877,
      "num_tokens": 61583846.0,
      "step": 27590
    },
    {
      "entropy": 5.289346313476562,
      "epoch": 3.0991744819453024,
      "grad_norm": 1.09375,
      "learning_rate": 0.00040604701126385924,
      "loss": 4.9052,
      "mean_token_accuracy": 0.21580614894628525,
      "num_tokens": 61594360.0,
      "step": 27595
    },
    {
      "entropy": 5.29053750038147,
      "epoch": 3.09973605885326,
      "grad_norm": 1.0,
      "learning_rate": 0.0004060143730858342,
      "loss": 4.9653,
      "mean_token_accuracy": 0.22019322514533995,
      "num_tokens": 61606750.0,
      "step": 27600
    },
    {
      "entropy": 5.248234462738037,
      "epoch": 3.1002976357612173,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040598173073626086,
      "loss": 4.9498,
      "mean_token_accuracy": 0.209003746509552,
      "num_tokens": 61618238.0,
      "step": 27605
    },
    {
      "entropy": 5.297725248336792,
      "epoch": 3.100859212669175,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004059490842161783,
      "loss": 4.9737,
      "mean_token_accuracy": 0.2115412473678589,
      "num_tokens": 61629568.0,
      "step": 27610
    },
    {
      "entropy": 5.345290422439575,
      "epoch": 3.1014207895771326,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040591643352662615,
      "loss": 5.015,
      "mean_token_accuracy": 0.21978553384542465,
      "num_tokens": 61640103.0,
      "step": 27615
    },
    {
      "entropy": 5.288796854019165,
      "epoch": 3.1019823664850903,
      "grad_norm": 1.125,
      "learning_rate": 0.000405883778668644,
      "loss": 4.8491,
      "mean_token_accuracy": 0.23168030232191086,
      "num_tokens": 61651392.0,
      "step": 27620
    },
    {
      "entropy": 5.32077317237854,
      "epoch": 3.1025439433930475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040585111964327157,
      "loss": 5.0661,
      "mean_token_accuracy": 0.20482772141695021,
      "num_tokens": 61664932.0,
      "step": 27625
    },
    {
      "entropy": 5.24095983505249,
      "epoch": 3.103105520301005,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00040581845645154877,
      "loss": 4.869,
      "mean_token_accuracy": 0.21781551390886306,
      "num_tokens": 61675719.0,
      "step": 27630
    },
    {
      "entropy": 5.36637053489685,
      "epoch": 3.103667097208963,
      "grad_norm": 1.0,
      "learning_rate": 0.0004057857890945155,
      "loss": 5.0185,
      "mean_token_accuracy": 0.21639130860567093,
      "num_tokens": 61687298.0,
      "step": 27635
    },
    {
      "entropy": 5.334862565994262,
      "epoch": 3.1042286741169205,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040575311757321203,
      "loss": 4.9887,
      "mean_token_accuracy": 0.21314700990915297,
      "num_tokens": 61697943.0,
      "step": 27640
    },
    {
      "entropy": 5.325567626953125,
      "epoch": 3.1047902510248777,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040572044188867863,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2123613566160202,
      "num_tokens": 61708897.0,
      "step": 27645
    },
    {
      "entropy": 5.2917827606201175,
      "epoch": 3.1053518279328354,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004056877620419556,
      "loss": 5.0537,
      "mean_token_accuracy": 0.2210842788219452,
      "num_tokens": 61720409.0,
      "step": 27650
    },
    {
      "entropy": 5.38825364112854,
      "epoch": 3.105913404840793,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004056550780340835,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21989912092685698,
      "num_tokens": 61731378.0,
      "step": 27655
    },
    {
      "entropy": 5.26224913597107,
      "epoch": 3.1064749817487507,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004056223898661031,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2127375602722168,
      "num_tokens": 61742713.0,
      "step": 27660
    },
    {
      "entropy": 5.319585704803467,
      "epoch": 3.107036558656708,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040558969753905514,
      "loss": 5.0028,
      "mean_token_accuracy": 0.21577365547418595,
      "num_tokens": 61753561.0,
      "step": 27665
    },
    {
      "entropy": 5.3364135265350345,
      "epoch": 3.1075981355646656,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004055570010539806,
      "loss": 4.9624,
      "mean_token_accuracy": 0.21484615355730058,
      "num_tokens": 61764966.0,
      "step": 27670
    },
    {
      "entropy": 5.287333059310913,
      "epoch": 3.1081597124726232,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004055243004119204,
      "loss": 4.9819,
      "mean_token_accuracy": 0.2095949739217758,
      "num_tokens": 61776988.0,
      "step": 27675
    },
    {
      "entropy": 5.297413110733032,
      "epoch": 3.108721289380581,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040549159561391595,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21351308673620223,
      "num_tokens": 61788327.0,
      "step": 27680
    },
    {
      "entropy": 5.256051683425904,
      "epoch": 3.109282866288538,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004054588866610085,
      "loss": 4.8635,
      "mean_token_accuracy": 0.22205168157815933,
      "num_tokens": 61799872.0,
      "step": 27685
    },
    {
      "entropy": 5.272955417633057,
      "epoch": 3.1098444431964958,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040542617355423947,
      "loss": 4.9212,
      "mean_token_accuracy": 0.21655530482530594,
      "num_tokens": 61810736.0,
      "step": 27690
    },
    {
      "entropy": 5.284741497039795,
      "epoch": 3.1104060201044534,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040539345629465043,
      "loss": 5.0227,
      "mean_token_accuracy": 0.21585240513086318,
      "num_tokens": 61821236.0,
      "step": 27695
    },
    {
      "entropy": 5.3798384189605715,
      "epoch": 3.1109675970124107,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040536073488328325,
      "loss": 4.999,
      "mean_token_accuracy": 0.21264252215623855,
      "num_tokens": 61832282.0,
      "step": 27700
    },
    {
      "entropy": 5.446009683609009,
      "epoch": 3.1115291739203683,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004053280093211796,
      "loss": 5.0366,
      "mean_token_accuracy": 0.20800196677446364,
      "num_tokens": 61842376.0,
      "step": 27705
    },
    {
      "entropy": 5.310466146469116,
      "epoch": 3.112090750828326,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004052952796093817,
      "loss": 4.9609,
      "mean_token_accuracy": 0.21695451587438583,
      "num_tokens": 61852227.0,
      "step": 27710
    },
    {
      "entropy": 5.341201400756836,
      "epoch": 3.1126523277362836,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040526254574893154,
      "loss": 4.9961,
      "mean_token_accuracy": 0.2166695401072502,
      "num_tokens": 61863119.0,
      "step": 27715
    },
    {
      "entropy": 5.242023897171021,
      "epoch": 3.113213904644241,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040522980774087143,
      "loss": 4.8889,
      "mean_token_accuracy": 0.22241268008947374,
      "num_tokens": 61874036.0,
      "step": 27720
    },
    {
      "entropy": 5.179537963867188,
      "epoch": 3.1137754815521985,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004051970655862437,
      "loss": 4.8364,
      "mean_token_accuracy": 0.22061903923749923,
      "num_tokens": 61884744.0,
      "step": 27725
    },
    {
      "entropy": 5.285757637023925,
      "epoch": 3.114337058460156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004051643192860908,
      "loss": 4.937,
      "mean_token_accuracy": 0.2115885004401207,
      "num_tokens": 61896866.0,
      "step": 27730
    },
    {
      "entropy": 5.29435625076294,
      "epoch": 3.114898635368114,
      "grad_norm": 1.0,
      "learning_rate": 0.00040513156884145564,
      "loss": 4.8355,
      "mean_token_accuracy": 0.22728189826011658,
      "num_tokens": 61907384.0,
      "step": 27735
    },
    {
      "entropy": 5.149316835403442,
      "epoch": 3.115460212276071,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004050988142533808,
      "loss": 4.7574,
      "mean_token_accuracy": 0.22716719955205916,
      "num_tokens": 61917719.0,
      "step": 27740
    },
    {
      "entropy": 5.256261348724365,
      "epoch": 3.1160217891840287,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004050660555229093,
      "loss": 4.893,
      "mean_token_accuracy": 0.21754445135593414,
      "num_tokens": 61928159.0,
      "step": 27745
    },
    {
      "entropy": 5.2382982730865475,
      "epoch": 3.1165833660919864,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004050332926510841,
      "loss": 4.852,
      "mean_token_accuracy": 0.2197740614414215,
      "num_tokens": 61938748.0,
      "step": 27750
    },
    {
      "entropy": 5.170305395126343,
      "epoch": 3.117144942999944,
      "grad_norm": 0.9375,
      "learning_rate": 0.0004050005256389484,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22953225523233414,
      "num_tokens": 61950429.0,
      "step": 27755
    },
    {
      "entropy": 5.235582447052002,
      "epoch": 3.1177065199079013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004049677544875456,
      "loss": 4.9138,
      "mean_token_accuracy": 0.21653608232736588,
      "num_tokens": 61961156.0,
      "step": 27760
    },
    {
      "entropy": 5.283325862884522,
      "epoch": 3.118268096815859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040493497919791906,
      "loss": 4.9027,
      "mean_token_accuracy": 0.2139047771692276,
      "num_tokens": 61972673.0,
      "step": 27765
    },
    {
      "entropy": 5.234465217590332,
      "epoch": 3.1188296737238166,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040490219977111237,
      "loss": 4.8755,
      "mean_token_accuracy": 0.2215936452150345,
      "num_tokens": 61984481.0,
      "step": 27770
    },
    {
      "entropy": 5.241405248641968,
      "epoch": 3.1193912506317742,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004048694162081693,
      "loss": 4.9095,
      "mean_token_accuracy": 0.2204273208975792,
      "num_tokens": 61994919.0,
      "step": 27775
    },
    {
      "entropy": 5.3199175834655765,
      "epoch": 3.1199528275397315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004048366285101336,
      "loss": 4.9434,
      "mean_token_accuracy": 0.22286106199026107,
      "num_tokens": 62005303.0,
      "step": 27780
    },
    {
      "entropy": 5.2344811916351315,
      "epoch": 3.120514404447689,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004048038366780494,
      "loss": 4.7674,
      "mean_token_accuracy": 0.22595374882221222,
      "num_tokens": 62015104.0,
      "step": 27785
    },
    {
      "entropy": 5.187202024459839,
      "epoch": 3.121075981355647,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0004047710407129606,
      "loss": 4.8747,
      "mean_token_accuracy": 0.21301120817661284,
      "num_tokens": 62025577.0,
      "step": 27790
    },
    {
      "entropy": 5.426878309249878,
      "epoch": 3.121637558263604,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004047382406159115,
      "loss": 5.0803,
      "mean_token_accuracy": 0.20761756002902984,
      "num_tokens": 62037001.0,
      "step": 27795
    },
    {
      "entropy": 5.344919633865357,
      "epoch": 3.1221991351715617,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004047054363879466,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21584350764751434,
      "num_tokens": 62049520.0,
      "step": 27800
    },
    {
      "entropy": 5.250145959854126,
      "epoch": 3.1227607120795193,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040467262803011025,
      "loss": 4.9397,
      "mean_token_accuracy": 0.21713325530290603,
      "num_tokens": 62061650.0,
      "step": 27805
    },
    {
      "entropy": 5.302205801010132,
      "epoch": 3.123322288987477,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040463981554344716,
      "loss": 5.0175,
      "mean_token_accuracy": 0.21595600098371506,
      "num_tokens": 62074433.0,
      "step": 27810
    },
    {
      "entropy": 5.335082197189331,
      "epoch": 3.123883865895434,
      "grad_norm": 1.15625,
      "learning_rate": 0.00040460699892900197,
      "loss": 4.9352,
      "mean_token_accuracy": 0.22236909419298173,
      "num_tokens": 62084885.0,
      "step": 27815
    },
    {
      "entropy": 5.278466701507568,
      "epoch": 3.124445442803392,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004045741781878198,
      "loss": 4.9693,
      "mean_token_accuracy": 0.2115195572376251,
      "num_tokens": 62096847.0,
      "step": 27820
    },
    {
      "entropy": 5.2998881340026855,
      "epoch": 3.1250070197113495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040454135332094543,
      "loss": 5.0128,
      "mean_token_accuracy": 0.21503058820962906,
      "num_tokens": 62108131.0,
      "step": 27825
    },
    {
      "entropy": 5.287036466598511,
      "epoch": 3.125568596619307,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004045085243294242,
      "loss": 4.9231,
      "mean_token_accuracy": 0.2188492476940155,
      "num_tokens": 62118450.0,
      "step": 27830
    },
    {
      "entropy": 5.303010940551758,
      "epoch": 3.1261301735272644,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004044756912143013,
      "loss": 4.9926,
      "mean_token_accuracy": 0.2161266624927521,
      "num_tokens": 62130702.0,
      "step": 27835
    },
    {
      "entropy": 5.326996326446533,
      "epoch": 3.126691750435222,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040444285397662225,
      "loss": 4.9929,
      "mean_token_accuracy": 0.2111354276537895,
      "num_tokens": 62142863.0,
      "step": 27840
    },
    {
      "entropy": 5.31247992515564,
      "epoch": 3.1272533273431797,
      "grad_norm": 0.953125,
      "learning_rate": 0.0004044100126174324,
      "loss": 4.9102,
      "mean_token_accuracy": 0.21657134592533112,
      "num_tokens": 62153787.0,
      "step": 27845
    },
    {
      "entropy": 5.324835586547851,
      "epoch": 3.1278149042511374,
      "grad_norm": 1.0,
      "learning_rate": 0.00040437716713777756,
      "loss": 4.995,
      "mean_token_accuracy": 0.21151483803987503,
      "num_tokens": 62166862.0,
      "step": 27850
    },
    {
      "entropy": 5.277343797683716,
      "epoch": 3.1283764811590946,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004043443175387036,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21813390552997589,
      "num_tokens": 62177431.0,
      "step": 27855
    },
    {
      "entropy": 5.282526302337646,
      "epoch": 3.1289380580670523,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00040431146382125644,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21377797275781632,
      "num_tokens": 62189465.0,
      "step": 27860
    },
    {
      "entropy": 5.30683126449585,
      "epoch": 3.12949963497501,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000404278605986482,
      "loss": 4.8409,
      "mean_token_accuracy": 0.22498052716255187,
      "num_tokens": 62201327.0,
      "step": 27865
    },
    {
      "entropy": 5.226938199996948,
      "epoch": 3.1300612118829676,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004042457440354267,
      "loss": 4.9341,
      "mean_token_accuracy": 0.2161559671163559,
      "num_tokens": 62212689.0,
      "step": 27870
    },
    {
      "entropy": 5.231374979019165,
      "epoch": 3.130622788790925,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004042128779691368,
      "loss": 4.9183,
      "mean_token_accuracy": 0.21652573943138123,
      "num_tokens": 62223771.0,
      "step": 27875
    },
    {
      "entropy": 5.338122987747193,
      "epoch": 3.1311843656988825,
      "grad_norm": 1.125,
      "learning_rate": 0.0004041800077886587,
      "loss": 4.9483,
      "mean_token_accuracy": 0.21862152814865113,
      "num_tokens": 62234137.0,
      "step": 27880
    },
    {
      "entropy": 5.298220825195313,
      "epoch": 3.13174594260684,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040414713349503907,
      "loss": 4.9434,
      "mean_token_accuracy": 0.21323164701461791,
      "num_tokens": 62244978.0,
      "step": 27885
    },
    {
      "entropy": 5.3568323135375975,
      "epoch": 3.1323075195147974,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040411425508932457,
      "loss": 5.0189,
      "mean_token_accuracy": 0.21377757638692857,
      "num_tokens": 62257364.0,
      "step": 27890
    },
    {
      "entropy": 5.271847248077393,
      "epoch": 3.132869096422755,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040408137257256225,
      "loss": 4.8087,
      "mean_token_accuracy": 0.22117819190025328,
      "num_tokens": 62268135.0,
      "step": 27895
    },
    {
      "entropy": 5.271409368515014,
      "epoch": 3.1334306733307127,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004040484859457988,
      "loss": 4.9902,
      "mean_token_accuracy": 0.21930910646915436,
      "num_tokens": 62279096.0,
      "step": 27900
    },
    {
      "entropy": 5.23463544845581,
      "epoch": 3.1339922502386703,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040401559521008155,
      "loss": 4.931,
      "mean_token_accuracy": 0.21130696088075637,
      "num_tokens": 62289640.0,
      "step": 27905
    },
    {
      "entropy": 5.384275579452515,
      "epoch": 3.1345538271466276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00040398270036645775,
      "loss": 5.0478,
      "mean_token_accuracy": 0.2123083084821701,
      "num_tokens": 62301085.0,
      "step": 27910
    },
    {
      "entropy": 5.218484020233154,
      "epoch": 3.135115404054585,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004039498014159748,
      "loss": 4.8098,
      "mean_token_accuracy": 0.22679661363363265,
      "num_tokens": 62312667.0,
      "step": 27915
    },
    {
      "entropy": 5.313133716583252,
      "epoch": 3.135676980962543,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040391689835968004,
      "loss": 4.9514,
      "mean_token_accuracy": 0.2182217851281166,
      "num_tokens": 62324640.0,
      "step": 27920
    },
    {
      "entropy": 5.327848672866821,
      "epoch": 3.1362385578705005,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004038839911986213,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21577666699886322,
      "num_tokens": 62335940.0,
      "step": 27925
    },
    {
      "entropy": 5.331552171707154,
      "epoch": 3.1368001347784578,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004038510799338462,
      "loss": 4.9945,
      "mean_token_accuracy": 0.21782040745019912,
      "num_tokens": 62346745.0,
      "step": 27930
    },
    {
      "entropy": 5.300714159011841,
      "epoch": 3.1373617116864154,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004038181645664029,
      "loss": 4.9662,
      "mean_token_accuracy": 0.20997847020626068,
      "num_tokens": 62357694.0,
      "step": 27935
    },
    {
      "entropy": 5.356081867218018,
      "epoch": 3.137923288594373,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040378524509733916,
      "loss": 5.0812,
      "mean_token_accuracy": 0.20799137502908707,
      "num_tokens": 62370621.0,
      "step": 27940
    },
    {
      "entropy": 5.29133415222168,
      "epoch": 3.1384848655023307,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004037523215277033,
      "loss": 4.939,
      "mean_token_accuracy": 0.2178821861743927,
      "num_tokens": 62381248.0,
      "step": 27945
    },
    {
      "entropy": 5.382280397415161,
      "epoch": 3.139046442410288,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004037193938585436,
      "loss": 4.9459,
      "mean_token_accuracy": 0.2113119289278984,
      "num_tokens": 62392102.0,
      "step": 27950
    },
    {
      "entropy": 5.3351154804229735,
      "epoch": 3.1396080193182456,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0004036864620909085,
      "loss": 4.9726,
      "mean_token_accuracy": 0.22282915264368058,
      "num_tokens": 62403254.0,
      "step": 27955
    },
    {
      "entropy": 5.242304039001465,
      "epoch": 3.1401695962262033,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004036535262258465,
      "loss": 4.853,
      "mean_token_accuracy": 0.22935593426227568,
      "num_tokens": 62413686.0,
      "step": 27960
    },
    {
      "entropy": 5.325237464904785,
      "epoch": 3.140731173134161,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040362058626440637,
      "loss": 4.9371,
      "mean_token_accuracy": 0.22143459618091582,
      "num_tokens": 62424119.0,
      "step": 27965
    },
    {
      "entropy": 5.3131202220916744,
      "epoch": 3.141292750042118,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040358764220763685,
      "loss": 4.8808,
      "mean_token_accuracy": 0.2236458793282509,
      "num_tokens": 62434900.0,
      "step": 27970
    },
    {
      "entropy": 5.265931701660156,
      "epoch": 3.141854326950076,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040355469405658687,
      "loss": 4.8293,
      "mean_token_accuracy": 0.22483887523412704,
      "num_tokens": 62446766.0,
      "step": 27975
    },
    {
      "entropy": 5.234325361251831,
      "epoch": 3.1424159038580335,
      "grad_norm": 1.09375,
      "learning_rate": 0.0004035217418123056,
      "loss": 4.896,
      "mean_token_accuracy": 0.21695386618375778,
      "num_tokens": 62457863.0,
      "step": 27980
    },
    {
      "entropy": 5.303915548324585,
      "epoch": 3.1429774807659907,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040348878547584227,
      "loss": 5.0374,
      "mean_token_accuracy": 0.2103790432214737,
      "num_tokens": 62470380.0,
      "step": 27985
    },
    {
      "entropy": 5.312794351577759,
      "epoch": 3.1435390576739484,
      "grad_norm": 0.953125,
      "learning_rate": 0.00040345582504824624,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2154131516814232,
      "num_tokens": 62482135.0,
      "step": 27990
    },
    {
      "entropy": 5.429594707489014,
      "epoch": 3.144100634581906,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0004034228605305668,
      "loss": 5.1394,
      "mean_token_accuracy": 0.20402587354183196,
      "num_tokens": 62494409.0,
      "step": 27995
    },
    {
      "entropy": 5.394413566589355,
      "epoch": 3.1446622114898637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040338989192385377,
      "loss": 5.0291,
      "mean_token_accuracy": 0.20944419205188752,
      "num_tokens": 62506247.0,
      "step": 28000
    },
    {
      "entropy": 5.383357143402099,
      "epoch": 3.145223788397821,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004033569192291567,
      "loss": 5.0167,
      "mean_token_accuracy": 0.20767561495304107,
      "num_tokens": 62517580.0,
      "step": 28005
    },
    {
      "entropy": 5.2527495384216305,
      "epoch": 3.1457853653057786,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0004033239424475256,
      "loss": 4.9613,
      "mean_token_accuracy": 0.21797734647989273,
      "num_tokens": 62529086.0,
      "step": 28010
    },
    {
      "entropy": 5.215735578536988,
      "epoch": 3.1463469422137362,
      "grad_norm": 1.0,
      "learning_rate": 0.00040329096158001033,
      "loss": 4.8682,
      "mean_token_accuracy": 0.21825608164072036,
      "num_tokens": 62540381.0,
      "step": 28015
    },
    {
      "entropy": 5.310928535461426,
      "epoch": 3.146908519121694,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0004032579766276612,
      "loss": 4.9854,
      "mean_token_accuracy": 0.20615262240171434,
      "num_tokens": 62551756.0,
      "step": 28020
    },
    {
      "entropy": 5.375712633132935,
      "epoch": 3.147470096029651,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040322498759152824,
      "loss": 4.9937,
      "mean_token_accuracy": 0.21148842573165894,
      "num_tokens": 62563666.0,
      "step": 28025
    },
    {
      "entropy": 5.423685550689697,
      "epoch": 3.1480316729376088,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040319199447266205,
      "loss": 5.0999,
      "mean_token_accuracy": 0.20465544760227203,
      "num_tokens": 62576326.0,
      "step": 28030
    },
    {
      "entropy": 5.333496332168579,
      "epoch": 3.1485932498455664,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004031589972721129,
      "loss": 4.9195,
      "mean_token_accuracy": 0.22140679955482484,
      "num_tokens": 62586911.0,
      "step": 28035
    },
    {
      "entropy": 5.230599737167358,
      "epoch": 3.149154826753524,
      "grad_norm": 1.0625,
      "learning_rate": 0.0004031259959909317,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22367650419473648,
      "num_tokens": 62596694.0,
      "step": 28040
    },
    {
      "entropy": 5.326405572891235,
      "epoch": 3.1497164036614813,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000403092990630169,
      "loss": 4.8847,
      "mean_token_accuracy": 0.22310822159051896,
      "num_tokens": 62607576.0,
      "step": 28045
    },
    {
      "entropy": 5.24544620513916,
      "epoch": 3.150277980569439,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004030599811908758,
      "loss": 4.8482,
      "mean_token_accuracy": 0.22222615778446198,
      "num_tokens": 62618015.0,
      "step": 28050
    },
    {
      "entropy": 5.206729507446289,
      "epoch": 3.1508395574773966,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004030269676741032,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21655142605304717,
      "num_tokens": 62629317.0,
      "step": 28055
    },
    {
      "entropy": 5.3193872451782225,
      "epoch": 3.1514011343853543,
      "grad_norm": 1.046875,
      "learning_rate": 0.0004029939500809023,
      "loss": 4.9747,
      "mean_token_accuracy": 0.21038335263729097,
      "num_tokens": 62640077.0,
      "step": 28060
    },
    {
      "entropy": 5.509465932846069,
      "epoch": 3.1519627112933115,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004029609284123243,
      "loss": 5.1503,
      "mean_token_accuracy": 0.20559622645378112,
      "num_tokens": 62652399.0,
      "step": 28065
    },
    {
      "entropy": 5.318646574020386,
      "epoch": 3.152524288201269,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040292790266942074,
      "loss": 4.912,
      "mean_token_accuracy": 0.21798875629901887,
      "num_tokens": 62663489.0,
      "step": 28070
    },
    {
      "entropy": 5.275661754608154,
      "epoch": 3.153085865109227,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040289487285324324,
      "loss": 5.0162,
      "mean_token_accuracy": 0.21577037870883942,
      "num_tokens": 62674975.0,
      "step": 28075
    },
    {
      "entropy": 5.422028064727783,
      "epoch": 3.153647442017184,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0004028618389648433,
      "loss": 5.074,
      "mean_token_accuracy": 0.2020912140607834,
      "num_tokens": 62686608.0,
      "step": 28080
    },
    {
      "entropy": 5.297918081283569,
      "epoch": 3.1542090189251417,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004028288010052728,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21682285517454147,
      "num_tokens": 62697844.0,
      "step": 28085
    },
    {
      "entropy": 5.19400315284729,
      "epoch": 3.1547705958330994,
      "grad_norm": 1.015625,
      "learning_rate": 0.0004027957589755837,
      "loss": 4.8574,
      "mean_token_accuracy": 0.2228354573249817,
      "num_tokens": 62709092.0,
      "step": 28090
    },
    {
      "entropy": 5.258807897567749,
      "epoch": 3.155332172741057,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00040276271287682813,
      "loss": 4.8844,
      "mean_token_accuracy": 0.2201384037733078,
      "num_tokens": 62720047.0,
      "step": 28095
    },
    {
      "entropy": 5.367953777313232,
      "epoch": 3.1558937496490143,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040272966271005816,
      "loss": 4.9573,
      "mean_token_accuracy": 0.21984607726335526,
      "num_tokens": 62729952.0,
      "step": 28100
    },
    {
      "entropy": 5.311278486251831,
      "epoch": 3.156455326556972,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040269660847632626,
      "loss": 4.9812,
      "mean_token_accuracy": 0.21385645717382432,
      "num_tokens": 62741961.0,
      "step": 28105
    },
    {
      "entropy": 5.245157957077026,
      "epoch": 3.1570169034649296,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040266355017668475,
      "loss": 4.9449,
      "mean_token_accuracy": 0.21319992393255233,
      "num_tokens": 62753552.0,
      "step": 28110
    },
    {
      "entropy": 5.225804948806763,
      "epoch": 3.1575784803728872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0004026304878121864,
      "loss": 4.8328,
      "mean_token_accuracy": 0.22636751532554628,
      "num_tokens": 62765239.0,
      "step": 28115
    },
    {
      "entropy": 5.160750436782837,
      "epoch": 3.1581400572808445,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00040259742138388366,
      "loss": 4.8993,
      "mean_token_accuracy": 0.2162322774529457,
      "num_tokens": 62776975.0,
      "step": 28120
    },
    {
      "entropy": 5.28805832862854,
      "epoch": 3.158701634188802,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040256435089282966,
      "loss": 4.8876,
      "mean_token_accuracy": 0.2200162813067436,
      "num_tokens": 62787435.0,
      "step": 28125
    },
    {
      "entropy": 5.321001482009888,
      "epoch": 3.15926321109676,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004025312763400772,
      "loss": 4.9355,
      "mean_token_accuracy": 0.21921732276678085,
      "num_tokens": 62799014.0,
      "step": 28130
    },
    {
      "entropy": 5.27852954864502,
      "epoch": 3.1598247880047174,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004024981977266794,
      "loss": 5.0108,
      "mean_token_accuracy": 0.20732743442058563,
      "num_tokens": 62810196.0,
      "step": 28135
    },
    {
      "entropy": 5.201806497573853,
      "epoch": 3.1603863649126747,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004024651150536896,
      "loss": 4.8166,
      "mean_token_accuracy": 0.22804001122713088,
      "num_tokens": 62821416.0,
      "step": 28140
    },
    {
      "entropy": 5.254514455795288,
      "epoch": 3.1609479418206323,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004024320283221611,
      "loss": 4.8776,
      "mean_token_accuracy": 0.22064041048288346,
      "num_tokens": 62831736.0,
      "step": 28145
    },
    {
      "entropy": 5.247081613540649,
      "epoch": 3.16150951872859,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040239893753314735,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2275161549448967,
      "num_tokens": 62842017.0,
      "step": 28150
    },
    {
      "entropy": 5.330385065078735,
      "epoch": 3.1620710956365476,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004023658426877021,
      "loss": 5.0631,
      "mean_token_accuracy": 0.20609376728534698,
      "num_tokens": 62853594.0,
      "step": 28155
    },
    {
      "entropy": 5.304387426376342,
      "epoch": 3.162632672544505,
      "grad_norm": 0.984375,
      "learning_rate": 0.00040233274378687897,
      "loss": 4.8897,
      "mean_token_accuracy": 0.22009450644254686,
      "num_tokens": 62864831.0,
      "step": 28160
    },
    {
      "entropy": 5.416194343566895,
      "epoch": 3.1631942494524625,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040229964083173187,
      "loss": 5.0481,
      "mean_token_accuracy": 0.2147366225719452,
      "num_tokens": 62877557.0,
      "step": 28165
    },
    {
      "entropy": 5.307600831985473,
      "epoch": 3.16375582636042,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040226653382331484,
      "loss": 4.9186,
      "mean_token_accuracy": 0.2167346805334091,
      "num_tokens": 62889162.0,
      "step": 28170
    },
    {
      "entropy": 5.201777553558349,
      "epoch": 3.1643174032683774,
      "grad_norm": 1.09375,
      "learning_rate": 0.000402233422762682,
      "loss": 4.8479,
      "mean_token_accuracy": 0.22096650749444963,
      "num_tokens": 62899376.0,
      "step": 28175
    },
    {
      "entropy": 5.286875200271607,
      "epoch": 3.164878980176335,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040220030765088764,
      "loss": 4.923,
      "mean_token_accuracy": 0.21830499470233916,
      "num_tokens": 62912448.0,
      "step": 28180
    },
    {
      "entropy": 5.21205768585205,
      "epoch": 3.1654405570842927,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004021671884889862,
      "loss": 4.8813,
      "mean_token_accuracy": 0.21809111535549164,
      "num_tokens": 62924021.0,
      "step": 28185
    },
    {
      "entropy": 5.352956008911133,
      "epoch": 3.1660021339922504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0004021340652780321,
      "loss": 4.9894,
      "mean_token_accuracy": 0.21775683611631394,
      "num_tokens": 62935330.0,
      "step": 28190
    },
    {
      "entropy": 5.306538248062134,
      "epoch": 3.1665637109002076,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004021009380190801,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22107849419116973,
      "num_tokens": 62947211.0,
      "step": 28195
    },
    {
      "entropy": 5.378201484680176,
      "epoch": 3.1671252878081653,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00040206780671318495,
      "loss": 5.0539,
      "mean_token_accuracy": 0.21398065984249115,
      "num_tokens": 62959012.0,
      "step": 28200
    },
    {
      "entropy": 5.3532383918762205,
      "epoch": 3.167686864716123,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004020346713614015,
      "loss": 5.0387,
      "mean_token_accuracy": 0.21284271478652955,
      "num_tokens": 62970608.0,
      "step": 28205
    },
    {
      "entropy": 5.382429504394532,
      "epoch": 3.1682484416240806,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00040200153196478486,
      "loss": 5.0041,
      "mean_token_accuracy": 0.22164004445075988,
      "num_tokens": 62982613.0,
      "step": 28210
    },
    {
      "entropy": 5.186114835739136,
      "epoch": 3.168810018532038,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004019683885243902,
      "loss": 4.8586,
      "mean_token_accuracy": 0.21877162009477616,
      "num_tokens": 62994394.0,
      "step": 28215
    },
    {
      "entropy": 5.282277679443359,
      "epoch": 3.1693715954399955,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004019352410412728,
      "loss": 5.016,
      "mean_token_accuracy": 0.2186909422278404,
      "num_tokens": 63007017.0,
      "step": 28220
    },
    {
      "entropy": 5.364042282104492,
      "epoch": 3.169933172347953,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040190208951648807,
      "loss": 4.9355,
      "mean_token_accuracy": 0.2136947989463806,
      "num_tokens": 63018325.0,
      "step": 28225
    },
    {
      "entropy": 5.380446338653565,
      "epoch": 3.170494749255911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004018689339510916,
      "loss": 5.0019,
      "mean_token_accuracy": 0.21804933100938798,
      "num_tokens": 63030694.0,
      "step": 28230
    },
    {
      "entropy": 5.206554269790649,
      "epoch": 3.171056326163868,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00040183577434613904,
      "loss": 4.8884,
      "mean_token_accuracy": 0.21826736778020858,
      "num_tokens": 63040889.0,
      "step": 28235
    },
    {
      "entropy": 5.19416823387146,
      "epoch": 3.1716179030718257,
      "grad_norm": 1.15625,
      "learning_rate": 0.0004018026107026863,
      "loss": 4.7888,
      "mean_token_accuracy": 0.22728036046028138,
      "num_tokens": 63051385.0,
      "step": 28240
    },
    {
      "entropy": 5.2279970169067385,
      "epoch": 3.1721794799797833,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004017694430217892,
      "loss": 4.9557,
      "mean_token_accuracy": 0.21618415862321855,
      "num_tokens": 63061989.0,
      "step": 28245
    },
    {
      "entropy": 5.33473834991455,
      "epoch": 3.172741056887741,
      "grad_norm": 1.03125,
      "learning_rate": 0.0004017362713045039,
      "loss": 4.9939,
      "mean_token_accuracy": 0.21571938842535018,
      "num_tokens": 63073031.0,
      "step": 28250
    },
    {
      "entropy": 5.256050109863281,
      "epoch": 3.173302633795698,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00040170309555188655,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22373560965061187,
      "num_tokens": 63083674.0,
      "step": 28255
    },
    {
      "entropy": 5.27005672454834,
      "epoch": 3.173864210703656,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0004016699157649936,
      "loss": 4.9167,
      "mean_token_accuracy": 0.22882162779569626,
      "num_tokens": 63094916.0,
      "step": 28260
    },
    {
      "entropy": 5.262336206436157,
      "epoch": 3.1744257876116135,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0004016367319448812,
      "loss": 5.0168,
      "mean_token_accuracy": 0.2088432118296623,
      "num_tokens": 63106307.0,
      "step": 28265
    },
    {
      "entropy": 5.246774482727051,
      "epoch": 3.1749873645195708,
      "grad_norm": 1.046875,
      "learning_rate": 0.00040160354409260625,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22427207231521606,
      "num_tokens": 63115871.0,
      "step": 28270
    },
    {
      "entropy": 5.151303148269653,
      "epoch": 3.1755489414275284,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004015703522092254,
      "loss": 4.8402,
      "mean_token_accuracy": 0.22894054055213928,
      "num_tokens": 63126282.0,
      "step": 28275
    },
    {
      "entropy": 5.19470534324646,
      "epoch": 3.176110518335486,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0004015371562957954,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21808595359325408,
      "num_tokens": 63138061.0,
      "step": 28280
    },
    {
      "entropy": 5.344504070281983,
      "epoch": 3.1766720952434437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0004015039563533732,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21624496281147004,
      "num_tokens": 63148478.0,
      "step": 28285
    },
    {
      "entropy": 5.346087503433227,
      "epoch": 3.177233672151401,
      "grad_norm": 0.96875,
      "learning_rate": 0.0004014707523830161,
      "loss": 4.9835,
      "mean_token_accuracy": 0.21876175105571746,
      "num_tokens": 63161205.0,
      "step": 28290
    },
    {
      "entropy": 5.30981478691101,
      "epoch": 3.1777952490593586,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0004014375443857811,
      "loss": 4.9062,
      "mean_token_accuracy": 0.2177453324198723,
      "num_tokens": 63173532.0,
      "step": 28295
    },
    {
      "entropy": 5.332444095611573,
      "epoch": 3.1783568259673163,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004014043323627257,
      "loss": 4.9517,
      "mean_token_accuracy": 0.2174285590648651,
      "num_tokens": 63185254.0,
      "step": 28300
    },
    {
      "entropy": 5.351208639144898,
      "epoch": 3.178918402875274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004013711163149072,
      "loss": 5.0303,
      "mean_token_accuracy": 0.20703745931386947,
      "num_tokens": 63196481.0,
      "step": 28305
    },
    {
      "entropy": 5.273060894012451,
      "epoch": 3.179479979783231,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040133789624338353,
      "loss": 4.8902,
      "mean_token_accuracy": 0.21994346827268602,
      "num_tokens": 63207411.0,
      "step": 28310
    },
    {
      "entropy": 5.261749744415283,
      "epoch": 3.180041556691189,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040130467214921216,
      "loss": 4.9273,
      "mean_token_accuracy": 0.2167239859700203,
      "num_tokens": 63219097.0,
      "step": 28315
    },
    {
      "entropy": 5.294607019424438,
      "epoch": 3.1806031335991465,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000401271444033451,
      "loss": 4.9905,
      "mean_token_accuracy": 0.21017127335071564,
      "num_tokens": 63230530.0,
      "step": 28320
    },
    {
      "entropy": 5.289042329788208,
      "epoch": 3.181164710507104,
      "grad_norm": 0.984375,
      "learning_rate": 0.0004012382118971582,
      "loss": 4.9384,
      "mean_token_accuracy": 0.22313934713602065,
      "num_tokens": 63241918.0,
      "step": 28325
    },
    {
      "entropy": 5.396295976638794,
      "epoch": 3.1817262874150614,
      "grad_norm": 1.078125,
      "learning_rate": 0.0004012049757413917,
      "loss": 5.0442,
      "mean_token_accuracy": 0.21201870888471602,
      "num_tokens": 63253821.0,
      "step": 28330
    },
    {
      "entropy": 5.327092075347901,
      "epoch": 3.182287864323019,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00040117173556720976,
      "loss": 4.9466,
      "mean_token_accuracy": 0.2165930986404419,
      "num_tokens": 63265463.0,
      "step": 28335
    },
    {
      "entropy": 5.2404218196868895,
      "epoch": 3.1828494412309767,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00040113849137567094,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22417756021022797,
      "num_tokens": 63276010.0,
      "step": 28340
    },
    {
      "entropy": 5.259672975540161,
      "epoch": 3.1834110181389343,
      "grad_norm": 0.90625,
      "learning_rate": 0.0004011052431678336,
      "loss": 4.9419,
      "mean_token_accuracy": 0.21501021683216096,
      "num_tokens": 63287179.0,
      "step": 28345
    },
    {
      "entropy": 5.274524068832397,
      "epoch": 3.1839725950468916,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004010719909447564,
      "loss": 4.9252,
      "mean_token_accuracy": 0.2222474679350853,
      "num_tokens": 63298453.0,
      "step": 28350
    },
    {
      "entropy": 5.262884664535522,
      "epoch": 3.1845341719548492,
      "grad_norm": 0.921875,
      "learning_rate": 0.0004010387347074981,
      "loss": 4.8977,
      "mean_token_accuracy": 0.2126927673816681,
      "num_tokens": 63310130.0,
      "step": 28355
    },
    {
      "entropy": 5.276423692703247,
      "epoch": 3.185095748862807,
      "grad_norm": 1.015625,
      "learning_rate": 0.00040100547445711765,
      "loss": 4.8685,
      "mean_token_accuracy": 0.22049684226512908,
      "num_tokens": 63319809.0,
      "step": 28360
    },
    {
      "entropy": 5.274505662918091,
      "epoch": 3.185657325770764,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040097221019467393,
      "loss": 4.9543,
      "mean_token_accuracy": 0.2123691186308861,
      "num_tokens": 63331263.0,
      "step": 28365
    },
    {
      "entropy": 5.290727281570435,
      "epoch": 3.1862189026787218,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004009389419212262,
      "loss": 4.9271,
      "mean_token_accuracy": 0.21972196847200393,
      "num_tokens": 63341983.0,
      "step": 28370
    },
    {
      "entropy": 5.371298122406006,
      "epoch": 3.1867804795866794,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00040090566963783375,
      "loss": 4.9591,
      "mean_token_accuracy": 0.22289785444736482,
      "num_tokens": 63352856.0,
      "step": 28375
    },
    {
      "entropy": 5.299763202667236,
      "epoch": 3.187342056494637,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040087239334555603,
      "loss": 4.9845,
      "mean_token_accuracy": 0.2197524532675743,
      "num_tokens": 63364893.0,
      "step": 28380
    },
    {
      "entropy": 5.292545747756958,
      "epoch": 3.1879036334025943,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00040083911304545234,
      "loss": 4.8885,
      "mean_token_accuracy": 0.22102829813957214,
      "num_tokens": 63375222.0,
      "step": 28385
    },
    {
      "entropy": 5.186845779418945,
      "epoch": 3.188465210310552,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00040080582873858254,
      "loss": 4.8963,
      "mean_token_accuracy": 0.21158166974782944,
      "num_tokens": 63387056.0,
      "step": 28390
    },
    {
      "entropy": 5.338265800476075,
      "epoch": 3.1890267872185096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0004007725404260064,
      "loss": 5.0041,
      "mean_token_accuracy": 0.21387791335582734,
      "num_tokens": 63397475.0,
      "step": 28395
    },
    {
      "entropy": 5.3139139175415036,
      "epoch": 3.1895883641264673,
      "grad_norm": 1.109375,
      "learning_rate": 0.0004007392481087837,
      "loss": 4.902,
      "mean_token_accuracy": 0.22098324745893477,
      "num_tokens": 63407920.0,
      "step": 28400
    },
    {
      "entropy": 5.221347713470459,
      "epoch": 3.1901499410344245,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0004007059517879747,
      "loss": 4.8448,
      "mean_token_accuracy": 0.22906207740306855,
      "num_tokens": 63418537.0,
      "step": 28405
    },
    {
      "entropy": 5.201022481918335,
      "epoch": 3.190711517942382,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0004006726514646393,
      "loss": 4.8661,
      "mean_token_accuracy": 0.21602424532175063,
      "num_tokens": 63429844.0,
      "step": 28410
    },
    {
      "entropy": 5.321454000473023,
      "epoch": 3.19127309485034,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00040063934713983803,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21796325743198394,
      "num_tokens": 63441417.0,
      "step": 28415
    },
    {
      "entropy": 5.236539602279663,
      "epoch": 3.1918346717582975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00040060603881463127,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2182061865925789,
      "num_tokens": 63451772.0,
      "step": 28420
    },
    {
      "entropy": 5.241088008880615,
      "epoch": 3.1923962486662547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00040057272649007944,
      "loss": 4.8608,
      "mean_token_accuracy": 0.2164887398481369,
      "num_tokens": 63462683.0,
      "step": 28425
    },
    {
      "entropy": 5.35709137916565,
      "epoch": 3.1929578255742124,
      "grad_norm": 1.078125,
      "learning_rate": 0.00040053941016724326,
      "loss": 4.9224,
      "mean_token_accuracy": 0.217690409719944,
      "num_tokens": 63473627.0,
      "step": 28430
    },
    {
      "entropy": 5.309332942962646,
      "epoch": 3.19351940248217,
      "grad_norm": 1.125,
      "learning_rate": 0.00040050608984718364,
      "loss": 5.02,
      "mean_token_accuracy": 0.21501439511775972,
      "num_tokens": 63485554.0,
      "step": 28435
    },
    {
      "entropy": 5.370879316329956,
      "epoch": 3.1940809793901277,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0004004727655309614,
      "loss": 4.9752,
      "mean_token_accuracy": 0.21948191672563552,
      "num_tokens": 63495881.0,
      "step": 28440
    },
    {
      "entropy": 5.310284519195557,
      "epoch": 3.194642556298085,
      "grad_norm": 0.96875,
      "learning_rate": 0.00040043943721963766,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21553446650505065,
      "num_tokens": 63506639.0,
      "step": 28445
    },
    {
      "entropy": 5.28129096031189,
      "epoch": 3.1952041332060426,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040040610491427357,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21208437979221345,
      "num_tokens": 63518482.0,
      "step": 28450
    },
    {
      "entropy": 5.250284957885742,
      "epoch": 3.1957657101140002,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0004003727686159305,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22131748497486115,
      "num_tokens": 63529019.0,
      "step": 28455
    },
    {
      "entropy": 5.305150175094605,
      "epoch": 3.1963272870219575,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00040033942832566984,
      "loss": 4.9245,
      "mean_token_accuracy": 0.2177282139658928,
      "num_tokens": 63540330.0,
      "step": 28460
    },
    {
      "entropy": 5.221702241897583,
      "epoch": 3.196888863929915,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0004003060840445531,
      "loss": 4.9214,
      "mean_token_accuracy": 0.2129831925034523,
      "num_tokens": 63552689.0,
      "step": 28465
    },
    {
      "entropy": 5.216285562515258,
      "epoch": 3.1974504408378728,
      "grad_norm": 1.0,
      "learning_rate": 0.0004002727357736422,
      "loss": 4.8842,
      "mean_token_accuracy": 0.2205333575606346,
      "num_tokens": 63562954.0,
      "step": 28470
    },
    {
      "entropy": 5.37522668838501,
      "epoch": 3.1980120177458304,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00040023938351399865,
      "loss": 4.9811,
      "mean_token_accuracy": 0.21220824271440505,
      "num_tokens": 63574697.0,
      "step": 28475
    },
    {
      "entropy": 5.309764385223389,
      "epoch": 3.1985735946537877,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0004002060272666846,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22466679215431212,
      "num_tokens": 63585060.0,
      "step": 28480
    },
    {
      "entropy": 5.271594142913818,
      "epoch": 3.1991351715617453,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004001726670327621,
      "loss": 4.9154,
      "mean_token_accuracy": 0.22406038492918015,
      "num_tokens": 63596749.0,
      "step": 28485
    },
    {
      "entropy": 5.218597888946533,
      "epoch": 3.199696748469703,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0004001393028132933,
      "loss": 4.7776,
      "mean_token_accuracy": 0.22650182545185088,
      "num_tokens": 63607629.0,
      "step": 28490
    },
    {
      "entropy": 5.205955982208252,
      "epoch": 3.2002583253776606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040010593460934065,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2267004981637001,
      "num_tokens": 63617354.0,
      "step": 28495
    },
    {
      "entropy": 5.261109781265259,
      "epoch": 3.200819902285618,
      "grad_norm": 1.03125,
      "learning_rate": 0.00040007256242196644,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21981254369020461,
      "num_tokens": 63629320.0,
      "step": 28500
    },
    {
      "entropy": 5.327747011184693,
      "epoch": 3.2013814791935755,
      "grad_norm": 1.0625,
      "learning_rate": 0.00040003918625223335,
      "loss": 4.9602,
      "mean_token_accuracy": 0.21510932743549346,
      "num_tokens": 63640453.0,
      "step": 28505
    },
    {
      "entropy": 5.269252490997315,
      "epoch": 3.201943056101533,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00040000580610120406,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22223442494869233,
      "num_tokens": 63652484.0,
      "step": 28510
    },
    {
      "entropy": 5.290737724304199,
      "epoch": 3.202504633009491,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003999724219699415,
      "loss": 4.9643,
      "mean_token_accuracy": 0.22718244343996047,
      "num_tokens": 63663084.0,
      "step": 28515
    },
    {
      "entropy": 5.2362110137939455,
      "epoch": 3.203066209917448,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003999390338595085,
      "loss": 4.9235,
      "mean_token_accuracy": 0.21759740710258485,
      "num_tokens": 63674023.0,
      "step": 28520
    },
    {
      "entropy": 5.308143949508667,
      "epoch": 3.2036277868254057,
      "grad_norm": 1.0,
      "learning_rate": 0.00039990564177096817,
      "loss": 4.9354,
      "mean_token_accuracy": 0.22350748479366303,
      "num_tokens": 63685370.0,
      "step": 28525
    },
    {
      "entropy": 5.332501363754273,
      "epoch": 3.2041893637333634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003998722457053838,
      "loss": 5.0066,
      "mean_token_accuracy": 0.2166453093290329,
      "num_tokens": 63696550.0,
      "step": 28530
    },
    {
      "entropy": 5.434557008743286,
      "epoch": 3.204750940641321,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039983884566381866,
      "loss": 5.0224,
      "mean_token_accuracy": 0.21350641995668412,
      "num_tokens": 63708040.0,
      "step": 28535
    },
    {
      "entropy": 5.309530401229859,
      "epoch": 3.2053125175492783,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003998054416473363,
      "loss": 4.9738,
      "mean_token_accuracy": 0.212677401304245,
      "num_tokens": 63719704.0,
      "step": 28540
    },
    {
      "entropy": 5.253631210327148,
      "epoch": 3.205874094457236,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039977203365700026,
      "loss": 4.9011,
      "mean_token_accuracy": 0.21515837907791138,
      "num_tokens": 63730111.0,
      "step": 28545
    },
    {
      "entropy": 5.330303621292114,
      "epoch": 3.2064356713651936,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003997386216938743,
      "loss": 4.9345,
      "mean_token_accuracy": 0.22543676644563676,
      "num_tokens": 63741054.0,
      "step": 28550
    },
    {
      "entropy": 5.323020362854004,
      "epoch": 3.206997248273151,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039970520575902216,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21873532831668854,
      "num_tokens": 63752598.0,
      "step": 28555
    },
    {
      "entropy": 5.278030920028686,
      "epoch": 3.2075588251811085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000399671785853508,
      "loss": 5.0262,
      "mean_token_accuracy": 0.2183200240135193,
      "num_tokens": 63763927.0,
      "step": 28560
    },
    {
      "entropy": 5.2653156280517575,
      "epoch": 3.208120402089066,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039963836197839584,
      "loss": 4.9094,
      "mean_token_accuracy": 0.2257863998413086,
      "num_tokens": 63775072.0,
      "step": 28565
    },
    {
      "entropy": 5.297639656066894,
      "epoch": 3.208681978997024,
      "grad_norm": 1.15625,
      "learning_rate": 0.00039960493413474977,
      "loss": 4.9008,
      "mean_token_accuracy": 0.22214146107435226,
      "num_tokens": 63786044.0,
      "step": 28570
    },
    {
      "entropy": 5.283523893356323,
      "epoch": 3.209243555904981,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039957150232363433,
      "loss": 4.9336,
      "mean_token_accuracy": 0.2213427633047104,
      "num_tokens": 63797889.0,
      "step": 28575
    },
    {
      "entropy": 5.343074941635132,
      "epoch": 3.2098051328129387,
      "grad_norm": 1.015625,
      "learning_rate": 0.000399538066546114,
      "loss": 4.9379,
      "mean_token_accuracy": 0.2119519129395485,
      "num_tokens": 63809567.0,
      "step": 28580
    },
    {
      "entropy": 5.3236589431762695,
      "epoch": 3.2103667097208963,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00039950462680325336,
      "loss": 4.9587,
      "mean_token_accuracy": 0.22052413523197173,
      "num_tokens": 63819617.0,
      "step": 28585
    },
    {
      "entropy": 5.365464591979981,
      "epoch": 3.210928286628854,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000399471183096117,
      "loss": 5.109,
      "mean_token_accuracy": 0.20851360708475114,
      "num_tokens": 63831535.0,
      "step": 28590
    },
    {
      "entropy": 5.359551858901978,
      "epoch": 3.211489863536811,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003994377354257698,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21910852789878846,
      "num_tokens": 63842629.0,
      "step": 28595
    },
    {
      "entropy": 5.402512693405152,
      "epoch": 3.212051440444769,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039940428379327703,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21630597561597825,
      "num_tokens": 63853677.0,
      "step": 28600
    },
    {
      "entropy": 5.370631361007691,
      "epoch": 3.2126130173527265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003993708281997036,
      "loss": 5.0191,
      "mean_token_accuracy": 0.20728300809860228,
      "num_tokens": 63865536.0,
      "step": 28605
    },
    {
      "entropy": 5.359984493255615,
      "epoch": 3.213174594260684,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003993373686461147,
      "loss": 4.9292,
      "mean_token_accuracy": 0.21103990823030472,
      "num_tokens": 63875849.0,
      "step": 28610
    },
    {
      "entropy": 5.331602573394775,
      "epoch": 3.2137361711686414,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003993039051335758,
      "loss": 4.9751,
      "mean_token_accuracy": 0.21622732579708098,
      "num_tokens": 63887222.0,
      "step": 28615
    },
    {
      "entropy": 5.2637989044189455,
      "epoch": 3.214297748076599,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003992704376631523,
      "loss": 4.9612,
      "mean_token_accuracy": 0.22060350626707076,
      "num_tokens": 63897969.0,
      "step": 28620
    },
    {
      "entropy": 5.3566062450408936,
      "epoch": 3.2148593249845567,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003992369662359099,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21907152831554413,
      "num_tokens": 63908913.0,
      "step": 28625
    },
    {
      "entropy": 5.286055231094361,
      "epoch": 3.2154209018925144,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003992034908529143,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21763285547494887,
      "num_tokens": 63919269.0,
      "step": 28630
    },
    {
      "entropy": 5.303280162811279,
      "epoch": 3.2159824788004716,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039917001151523144,
      "loss": 4.9107,
      "mean_token_accuracy": 0.21498042345046997,
      "num_tokens": 63930624.0,
      "step": 28635
    },
    {
      "entropy": 5.284516954421997,
      "epoch": 3.2165440557084293,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003991365282239272,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21865603625774382,
      "num_tokens": 63941571.0,
      "step": 28640
    },
    {
      "entropy": 5.326879215240479,
      "epoch": 3.217105632616387,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039910304098006775,
      "loss": 5.0062,
      "mean_token_accuracy": 0.21642879545688629,
      "num_tokens": 63953221.0,
      "step": 28645
    },
    {
      "entropy": 5.364869737625122,
      "epoch": 3.217667209524344,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003990695497847193,
      "loss": 4.931,
      "mean_token_accuracy": 0.21942937821149827,
      "num_tokens": 63965076.0,
      "step": 28650
    },
    {
      "entropy": 5.27906608581543,
      "epoch": 3.218228786432302,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003990360546389484,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22552677392959594,
      "num_tokens": 63976601.0,
      "step": 28655
    },
    {
      "entropy": 5.189037179946899,
      "epoch": 3.2187903633402595,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039900255554382126,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22350853830575942,
      "num_tokens": 63987757.0,
      "step": 28660
    },
    {
      "entropy": 5.290076065063476,
      "epoch": 3.219351940248217,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039896905250040473,
      "loss": 4.99,
      "mean_token_accuracy": 0.21741218864917755,
      "num_tokens": 64000104.0,
      "step": 28665
    },
    {
      "entropy": 5.322853946685791,
      "epoch": 3.2199135171561744,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039893554550976545,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22218135446310044,
      "num_tokens": 64010773.0,
      "step": 28670
    },
    {
      "entropy": 5.259202194213867,
      "epoch": 3.220475094064132,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003989020345729705,
      "loss": 4.8832,
      "mean_token_accuracy": 0.22180868536233903,
      "num_tokens": 64022269.0,
      "step": 28675
    },
    {
      "entropy": 5.2458815574646,
      "epoch": 3.2210366709720897,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039886851969108653,
      "loss": 4.9119,
      "mean_token_accuracy": 0.21652494221925736,
      "num_tokens": 64033622.0,
      "step": 28680
    },
    {
      "entropy": 5.22196946144104,
      "epoch": 3.2215982478800473,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039883500086518086,
      "loss": 4.9406,
      "mean_token_accuracy": 0.22054858654737472,
      "num_tokens": 64044243.0,
      "step": 28685
    },
    {
      "entropy": 5.3575128555297855,
      "epoch": 3.2221598247880046,
      "grad_norm": 1.078125,
      "learning_rate": 0.00039880147809632084,
      "loss": 4.9299,
      "mean_token_accuracy": 0.2181605204939842,
      "num_tokens": 64054452.0,
      "step": 28690
    },
    {
      "entropy": 5.322979831695557,
      "epoch": 3.222721401695962,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039876795138557363,
      "loss": 4.9242,
      "mean_token_accuracy": 0.21510860174894333,
      "num_tokens": 64065743.0,
      "step": 28695
    },
    {
      "entropy": 5.323347473144532,
      "epoch": 3.22328297860392,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003987344207340069,
      "loss": 4.9754,
      "mean_token_accuracy": 0.215836201608181,
      "num_tokens": 64077359.0,
      "step": 28700
    },
    {
      "entropy": 5.246966361999512,
      "epoch": 3.2238445555118775,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039870088614268824,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22161054760217666,
      "num_tokens": 64088281.0,
      "step": 28705
    },
    {
      "entropy": 5.234586620330811,
      "epoch": 3.2244061324198348,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039866734761268534,
      "loss": 4.9267,
      "mean_token_accuracy": 0.2176364168524742,
      "num_tokens": 64099600.0,
      "step": 28710
    },
    {
      "entropy": 5.360826587677002,
      "epoch": 3.2249677093277924,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039863380514506614,
      "loss": 4.9345,
      "mean_token_accuracy": 0.21829698085784913,
      "num_tokens": 64110381.0,
      "step": 28715
    },
    {
      "entropy": 5.320620584487915,
      "epoch": 3.22552928623575,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039860025874089865,
      "loss": 4.913,
      "mean_token_accuracy": 0.21302030086517335,
      "num_tokens": 64120878.0,
      "step": 28720
    },
    {
      "entropy": 5.181659460067749,
      "epoch": 3.2260908631437077,
      "grad_norm": 1.109375,
      "learning_rate": 0.000398566708401251,
      "loss": 4.7779,
      "mean_token_accuracy": 0.23203096836805343,
      "num_tokens": 64131427.0,
      "step": 28725
    },
    {
      "entropy": 5.2629804611206055,
      "epoch": 3.226652440051665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039853315412719144,
      "loss": 5.0023,
      "mean_token_accuracy": 0.21296768337488176,
      "num_tokens": 64142833.0,
      "step": 28730
    },
    {
      "entropy": 5.188179302215576,
      "epoch": 3.2272140169596226,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003984995959197883,
      "loss": 4.8608,
      "mean_token_accuracy": 0.22088757306337356,
      "num_tokens": 64152405.0,
      "step": 28735
    },
    {
      "entropy": 5.2758636474609375,
      "epoch": 3.2277755938675803,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003984660337801101,
      "loss": 4.8693,
      "mean_token_accuracy": 0.21734008193016052,
      "num_tokens": 64164398.0,
      "step": 28740
    },
    {
      "entropy": 5.293877172470093,
      "epoch": 3.2283371707755375,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00039843246770922557,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21711243093013763,
      "num_tokens": 64175313.0,
      "step": 28745
    },
    {
      "entropy": 5.215699243545532,
      "epoch": 3.228898747683495,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003983988977082033,
      "loss": 4.846,
      "mean_token_accuracy": 0.2212107449769974,
      "num_tokens": 64185866.0,
      "step": 28750
    },
    {
      "entropy": 5.267713022232056,
      "epoch": 3.229460324591453,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039836532377811237,
      "loss": 4.913,
      "mean_token_accuracy": 0.2184227228164673,
      "num_tokens": 64197126.0,
      "step": 28755
    },
    {
      "entropy": 5.234656238555909,
      "epoch": 3.2300219014994105,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003983317459200217,
      "loss": 4.9006,
      "mean_token_accuracy": 0.21606328338384628,
      "num_tokens": 64209271.0,
      "step": 28760
    },
    {
      "entropy": 5.3717512607574465,
      "epoch": 3.2305834784073677,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039829816413500035,
      "loss": 5.0072,
      "mean_token_accuracy": 0.2197332039475441,
      "num_tokens": 64221200.0,
      "step": 28765
    },
    {
      "entropy": 5.351526927947998,
      "epoch": 3.2311450553153254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003982645784241176,
      "loss": 4.9719,
      "mean_token_accuracy": 0.21444824635982512,
      "num_tokens": 64230835.0,
      "step": 28770
    },
    {
      "entropy": 5.303991413116455,
      "epoch": 3.231706632223283,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003982309887884429,
      "loss": 4.9067,
      "mean_token_accuracy": 0.22477735280990602,
      "num_tokens": 64242243.0,
      "step": 28775
    },
    {
      "entropy": 5.2649681091308596,
      "epoch": 3.2322682091312407,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039819739522904576,
      "loss": 4.9067,
      "mean_token_accuracy": 0.2207592770457268,
      "num_tokens": 64253362.0,
      "step": 28780
    },
    {
      "entropy": 5.251909255981445,
      "epoch": 3.232829786039198,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039816379774699573,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2179113358259201,
      "num_tokens": 64264462.0,
      "step": 28785
    },
    {
      "entropy": 5.22101674079895,
      "epoch": 3.2333913629471556,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039813019634336264,
      "loss": 4.8598,
      "mean_token_accuracy": 0.2186206743121147,
      "num_tokens": 64275153.0,
      "step": 28790
    },
    {
      "entropy": 5.27798433303833,
      "epoch": 3.2339529398551132,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003980965910192163,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22731612771749496,
      "num_tokens": 64288082.0,
      "step": 28795
    },
    {
      "entropy": 5.259137535095215,
      "epoch": 3.234514516763071,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039806298177562675,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22032736390829086,
      "num_tokens": 64298872.0,
      "step": 28800
    },
    {
      "entropy": 5.325430679321289,
      "epoch": 3.235076093671028,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039802936861366416,
      "loss": 5.0535,
      "mean_token_accuracy": 0.21286675333976746,
      "num_tokens": 64310245.0,
      "step": 28805
    },
    {
      "entropy": 5.347151899337769,
      "epoch": 3.2356376705789858,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003979957515343988,
      "loss": 5.0011,
      "mean_token_accuracy": 0.21418435275554656,
      "num_tokens": 64323681.0,
      "step": 28810
    },
    {
      "entropy": 5.222387886047363,
      "epoch": 3.2361992474869434,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000397962130538901,
      "loss": 4.9399,
      "mean_token_accuracy": 0.2233033746480942,
      "num_tokens": 64335265.0,
      "step": 28815
    },
    {
      "entropy": 5.368210506439209,
      "epoch": 3.236760824394901,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039792850562824124,
      "loss": 5.0177,
      "mean_token_accuracy": 0.21337098330259324,
      "num_tokens": 64346279.0,
      "step": 28820
    },
    {
      "entropy": 5.360521364212036,
      "epoch": 3.2373224013028583,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003978948768034902,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22200917899608613,
      "num_tokens": 64357511.0,
      "step": 28825
    },
    {
      "entropy": 5.283775472640992,
      "epoch": 3.237883978210816,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039786124406571854,
      "loss": 4.9594,
      "mean_token_accuracy": 0.2135210379958153,
      "num_tokens": 64367863.0,
      "step": 28830
    },
    {
      "entropy": 5.267098903656006,
      "epoch": 3.2384455551187736,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039782760741599734,
      "loss": 4.9168,
      "mean_token_accuracy": 0.2181365743279457,
      "num_tokens": 64379746.0,
      "step": 28835
    },
    {
      "entropy": 5.263925457000733,
      "epoch": 3.239007132026731,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003977939668553973,
      "loss": 4.8837,
      "mean_token_accuracy": 0.228728187084198,
      "num_tokens": 64390370.0,
      "step": 28840
    },
    {
      "entropy": 5.30924391746521,
      "epoch": 3.2395687089346885,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039776032238498995,
      "loss": 5.0332,
      "mean_token_accuracy": 0.2123050093650818,
      "num_tokens": 64402620.0,
      "step": 28845
    },
    {
      "entropy": 5.289881563186645,
      "epoch": 3.240130285842646,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003977266740058462,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21954806596040727,
      "num_tokens": 64412969.0,
      "step": 28850
    },
    {
      "entropy": 5.30048246383667,
      "epoch": 3.240691862750604,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003976930217190376,
      "loss": 4.96,
      "mean_token_accuracy": 0.21478240042924882,
      "num_tokens": 64423679.0,
      "step": 28855
    },
    {
      "entropy": 5.338481092453003,
      "epoch": 3.241253439658561,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003976593655256356,
      "loss": 4.9523,
      "mean_token_accuracy": 0.21956281661987304,
      "num_tokens": 64436124.0,
      "step": 28860
    },
    {
      "entropy": 5.259313011169434,
      "epoch": 3.2418150165665187,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039762570542671176,
      "loss": 4.8952,
      "mean_token_accuracy": 0.21877423971891402,
      "num_tokens": 64446494.0,
      "step": 28865
    },
    {
      "entropy": 5.283436393737793,
      "epoch": 3.2423765934744764,
      "grad_norm": 1.03125,
      "learning_rate": 0.000397592041423338,
      "loss": 4.8466,
      "mean_token_accuracy": 0.22218749821186065,
      "num_tokens": 64456152.0,
      "step": 28870
    },
    {
      "entropy": 5.294330358505249,
      "epoch": 3.242938170382434,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039755837351658605,
      "loss": 4.9948,
      "mean_token_accuracy": 0.22062428146600724,
      "num_tokens": 64468004.0,
      "step": 28875
    },
    {
      "entropy": 5.271237134933472,
      "epoch": 3.2434997472903913,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003975247017075279,
      "loss": 4.9053,
      "mean_token_accuracy": 0.2236076757311821,
      "num_tokens": 64478506.0,
      "step": 28880
    },
    {
      "entropy": 5.252885007858277,
      "epoch": 3.244061324198349,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003974910259972358,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22236464023590088,
      "num_tokens": 64489379.0,
      "step": 28885
    },
    {
      "entropy": 5.270563173294067,
      "epoch": 3.2446229011063066,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003974573463867819,
      "loss": 4.9719,
      "mean_token_accuracy": 0.22546855509281158,
      "num_tokens": 64499500.0,
      "step": 28890
    },
    {
      "entropy": 5.2169183731079105,
      "epoch": 3.2451844780142642,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039742366287723855,
      "loss": 4.8694,
      "mean_token_accuracy": 0.2292092889547348,
      "num_tokens": 64510869.0,
      "step": 28895
    },
    {
      "entropy": 5.255762386322021,
      "epoch": 3.2457460549222215,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039738997546967834,
      "loss": 4.9483,
      "mean_token_accuracy": 0.2101021870970726,
      "num_tokens": 64521581.0,
      "step": 28900
    },
    {
      "entropy": 5.3591827869415285,
      "epoch": 3.246307631830179,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039735628416517394,
      "loss": 5.03,
      "mean_token_accuracy": 0.21130781918764113,
      "num_tokens": 64532436.0,
      "step": 28905
    },
    {
      "entropy": 5.368667697906494,
      "epoch": 3.246869208738137,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039732258896479786,
      "loss": 4.9755,
      "mean_token_accuracy": 0.21624844521284103,
      "num_tokens": 64543546.0,
      "step": 28910
    },
    {
      "entropy": 5.404611015319825,
      "epoch": 3.2474307856460944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003972888898696231,
      "loss": 5.0844,
      "mean_token_accuracy": 0.21141918003559113,
      "num_tokens": 64554964.0,
      "step": 28915
    },
    {
      "entropy": 5.295389127731323,
      "epoch": 3.2479923625540517,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039725518688072276,
      "loss": 4.9164,
      "mean_token_accuracy": 0.21697306782007217,
      "num_tokens": 64564082.0,
      "step": 28920
    },
    {
      "entropy": 5.298329925537109,
      "epoch": 3.2485539394620093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039722147999916976,
      "loss": 5.0164,
      "mean_token_accuracy": 0.21353599131107331,
      "num_tokens": 64575377.0,
      "step": 28925
    },
    {
      "entropy": 5.29119725227356,
      "epoch": 3.249115516369967,
      "grad_norm": 0.86328125,
      "learning_rate": 0.0003971877692260375,
      "loss": 4.8931,
      "mean_token_accuracy": 0.2184244692325592,
      "num_tokens": 64587686.0,
      "step": 28930
    },
    {
      "entropy": 5.286176538467407,
      "epoch": 3.249677093277924,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039715405456239924,
      "loss": 4.8724,
      "mean_token_accuracy": 0.21692540347576142,
      "num_tokens": 64598501.0,
      "step": 28935
    },
    {
      "entropy": 5.268251943588257,
      "epoch": 3.250238670185882,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003971203360093286,
      "loss": 4.9149,
      "mean_token_accuracy": 0.2200179725885391,
      "num_tokens": 64609341.0,
      "step": 28940
    },
    {
      "entropy": 5.318048572540283,
      "epoch": 3.2508002470938395,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039708661356789896,
      "loss": 4.9827,
      "mean_token_accuracy": 0.2183813750743866,
      "num_tokens": 64621613.0,
      "step": 28945
    },
    {
      "entropy": 5.183610105514527,
      "epoch": 3.251361824001797,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039705288723918415,
      "loss": 4.9061,
      "mean_token_accuracy": 0.21968828588724137,
      "num_tokens": 64633358.0,
      "step": 28950
    },
    {
      "entropy": 5.239304161071777,
      "epoch": 3.2519234009097544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003970191570242582,
      "loss": 4.8834,
      "mean_token_accuracy": 0.21422467976808549,
      "num_tokens": 64644799.0,
      "step": 28955
    },
    {
      "entropy": 5.262898015975952,
      "epoch": 3.252484977817712,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000396985422924195,
      "loss": 4.8502,
      "mean_token_accuracy": 0.23074136525392533,
      "num_tokens": 64657028.0,
      "step": 28960
    },
    {
      "entropy": 5.352479457855225,
      "epoch": 3.2530465547256697,
      "grad_norm": 0.96875,
      "learning_rate": 0.00039695168494006857,
      "loss": 5.0365,
      "mean_token_accuracy": 0.2113450825214386,
      "num_tokens": 64668572.0,
      "step": 28965
    },
    {
      "entropy": 5.1988006114959715,
      "epoch": 3.2536081316336274,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003969179430729531,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22190239429473876,
      "num_tokens": 64679168.0,
      "step": 28970
    },
    {
      "entropy": 5.394726324081421,
      "epoch": 3.2541697085415846,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003968841973239231,
      "loss": 4.997,
      "mean_token_accuracy": 0.21212707459926605,
      "num_tokens": 64691571.0,
      "step": 28975
    },
    {
      "entropy": 5.343031358718872,
      "epoch": 3.2547312854495423,
      "grad_norm": 0.9140625,
      "learning_rate": 0.000396850447694053,
      "loss": 5.007,
      "mean_token_accuracy": 0.21335015147924424,
      "num_tokens": 64703659.0,
      "step": 28980
    },
    {
      "entropy": 5.263974285125732,
      "epoch": 3.2552928623575,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003968166941844174,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22594222873449327,
      "num_tokens": 64715121.0,
      "step": 28985
    },
    {
      "entropy": 5.248375177383423,
      "epoch": 3.2558544392654576,
      "grad_norm": 1.0625,
      "learning_rate": 0.000396782936796091,
      "loss": 4.9404,
      "mean_token_accuracy": 0.215643909573555,
      "num_tokens": 64726289.0,
      "step": 28990
    },
    {
      "entropy": 5.274578285217285,
      "epoch": 3.256416016173415,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003967491755301487,
      "loss": 4.8942,
      "mean_token_accuracy": 0.21912216246128083,
      "num_tokens": 64737378.0,
      "step": 28995
    },
    {
      "entropy": 5.386879539489746,
      "epoch": 3.2569775930813725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003967154103876654,
      "loss": 5.0964,
      "mean_token_accuracy": 0.20523611754179,
      "num_tokens": 64749203.0,
      "step": 29000
    },
    {
      "entropy": 5.216130781173706,
      "epoch": 3.25753916998933,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039668164136971615,
      "loss": 4.8388,
      "mean_token_accuracy": 0.22592228949069976,
      "num_tokens": 64759514.0,
      "step": 29005
    },
    {
      "entropy": 5.285856294631958,
      "epoch": 3.258100746897288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003966478684773763,
      "loss": 4.9062,
      "mean_token_accuracy": 0.225363090634346,
      "num_tokens": 64769370.0,
      "step": 29010
    },
    {
      "entropy": 5.294276094436645,
      "epoch": 3.258662323805245,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00039661409171172117,
      "loss": 4.9293,
      "mean_token_accuracy": 0.2189907193183899,
      "num_tokens": 64779930.0,
      "step": 29015
    },
    {
      "entropy": 5.247143507003784,
      "epoch": 3.2592239007132027,
      "grad_norm": 0.953125,
      "learning_rate": 0.00039658031107382615,
      "loss": 4.9771,
      "mean_token_accuracy": 0.20833487212657928,
      "num_tokens": 64791726.0,
      "step": 29020
    },
    {
      "entropy": 5.194288158416748,
      "epoch": 3.2597854776211603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039654652656476695,
      "loss": 4.794,
      "mean_token_accuracy": 0.23188262581825256,
      "num_tokens": 64802778.0,
      "step": 29025
    },
    {
      "entropy": 5.266699266433716,
      "epoch": 3.2603470545291175,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003965127381856191,
      "loss": 5.0051,
      "mean_token_accuracy": 0.21357111632823944,
      "num_tokens": 64814664.0,
      "step": 29030
    },
    {
      "entropy": 5.259763193130493,
      "epoch": 3.260908631437075,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039647894593745865,
      "loss": 4.8923,
      "mean_token_accuracy": 0.2213047742843628,
      "num_tokens": 64825775.0,
      "step": 29035
    },
    {
      "entropy": 5.344921064376831,
      "epoch": 3.261470208345033,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039644514982136136,
      "loss": 5.0591,
      "mean_token_accuracy": 0.21755662709474563,
      "num_tokens": 64837376.0,
      "step": 29040
    },
    {
      "entropy": 5.322809648513794,
      "epoch": 3.2620317852529905,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039641134983840336,
      "loss": 4.8941,
      "mean_token_accuracy": 0.21804770827293396,
      "num_tokens": 64847941.0,
      "step": 29045
    },
    {
      "entropy": 5.388068389892578,
      "epoch": 3.2625933621609478,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000396377545989661,
      "loss": 5.0222,
      "mean_token_accuracy": 0.2126004919409752,
      "num_tokens": 64858793.0,
      "step": 29050
    },
    {
      "entropy": 5.259556818008423,
      "epoch": 3.2631549390689054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039634373827621047,
      "loss": 4.9251,
      "mean_token_accuracy": 0.21661882996559143,
      "num_tokens": 64870233.0,
      "step": 29055
    },
    {
      "entropy": 5.226206731796265,
      "epoch": 3.263716515976863,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003963099266991282,
      "loss": 4.8883,
      "mean_token_accuracy": 0.21989445090293885,
      "num_tokens": 64881875.0,
      "step": 29060
    },
    {
      "entropy": 5.248084640502929,
      "epoch": 3.2642780928848207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039627611125949086,
      "loss": 4.9,
      "mean_token_accuracy": 0.22131018042564393,
      "num_tokens": 64892837.0,
      "step": 29065
    },
    {
      "entropy": 5.3932852268219,
      "epoch": 3.264839669792778,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039624229195837506,
      "loss": 5.0289,
      "mean_token_accuracy": 0.21141774356365203,
      "num_tokens": 64904773.0,
      "step": 29070
    },
    {
      "entropy": 5.3004974842071535,
      "epoch": 3.2654012467007356,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039620846879685767,
      "loss": 4.9361,
      "mean_token_accuracy": 0.21888973116874694,
      "num_tokens": 64916596.0,
      "step": 29075
    },
    {
      "entropy": 5.28334789276123,
      "epoch": 3.2659628236086933,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00039617464177601564,
      "loss": 4.9154,
      "mean_token_accuracy": 0.2195022240281105,
      "num_tokens": 64927736.0,
      "step": 29080
    },
    {
      "entropy": 5.335350179672242,
      "epoch": 3.266524400516651,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00039614081089692606,
      "loss": 5.043,
      "mean_token_accuracy": 0.21256789565086365,
      "num_tokens": 64940426.0,
      "step": 29085
    },
    {
      "entropy": 5.226320600509643,
      "epoch": 3.267085977424608,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039610697616066604,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21974678486585617,
      "num_tokens": 64951034.0,
      "step": 29090
    },
    {
      "entropy": 5.345095300674439,
      "epoch": 3.267647554332566,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039607313756831296,
      "loss": 4.9784,
      "mean_token_accuracy": 0.2139018267393112,
      "num_tokens": 64962900.0,
      "step": 29095
    },
    {
      "entropy": 5.315540504455567,
      "epoch": 3.2682091312405235,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039603929512094417,
      "loss": 4.8949,
      "mean_token_accuracy": 0.23301808834075927,
      "num_tokens": 64974820.0,
      "step": 29100
    },
    {
      "entropy": 5.214464378356934,
      "epoch": 3.268770708148481,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003960054488196373,
      "loss": 4.8606,
      "mean_token_accuracy": 0.2162912368774414,
      "num_tokens": 64985964.0,
      "step": 29105
    },
    {
      "entropy": 5.263800573348999,
      "epoch": 3.2693322850564384,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039597159866546996,
      "loss": 4.8843,
      "mean_token_accuracy": 0.2176966890692711,
      "num_tokens": 64996772.0,
      "step": 29110
    },
    {
      "entropy": 5.242616415023804,
      "epoch": 3.269893861964396,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003959377446595201,
      "loss": 4.8378,
      "mean_token_accuracy": 0.22048196494579314,
      "num_tokens": 65007296.0,
      "step": 29115
    },
    {
      "entropy": 5.22902512550354,
      "epoch": 3.2704554388723537,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039590388680286543,
      "loss": 4.8961,
      "mean_token_accuracy": 0.23121747523546218,
      "num_tokens": 65018384.0,
      "step": 29120
    },
    {
      "entropy": 5.211492538452148,
      "epoch": 3.271017015780311,
      "grad_norm": 1.125,
      "learning_rate": 0.0003958700250965842,
      "loss": 4.8277,
      "mean_token_accuracy": 0.22170943021774292,
      "num_tokens": 65029042.0,
      "step": 29125
    },
    {
      "entropy": 5.295046663284301,
      "epoch": 3.2715785926882686,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003958361595417544,
      "loss": 4.9373,
      "mean_token_accuracy": 0.21794226467609407,
      "num_tokens": 65040792.0,
      "step": 29130
    },
    {
      "entropy": 5.300461196899414,
      "epoch": 3.2721401695962262,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003958022901394545,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22641170471906663,
      "num_tokens": 65051702.0,
      "step": 29135
    },
    {
      "entropy": 5.254183149337768,
      "epoch": 3.272701746504184,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003957684168907628,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22207220047712325,
      "num_tokens": 65063288.0,
      "step": 29140
    },
    {
      "entropy": 5.318965148925781,
      "epoch": 3.273263323412141,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039573453979675793,
      "loss": 5.023,
      "mean_token_accuracy": 0.2059458389878273,
      "num_tokens": 65074913.0,
      "step": 29145
    },
    {
      "entropy": 5.290867757797241,
      "epoch": 3.2738249003200988,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039570065885851844,
      "loss": 5.0331,
      "mean_token_accuracy": 0.2147068127989769,
      "num_tokens": 65085327.0,
      "step": 29150
    },
    {
      "entropy": 5.3018097400665285,
      "epoch": 3.2743864772280564,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003956667740771231,
      "loss": 4.8919,
      "mean_token_accuracy": 0.22146945148706437,
      "num_tokens": 65095709.0,
      "step": 29155
    },
    {
      "entropy": 5.33000340461731,
      "epoch": 3.274948054136014,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003956328854536509,
      "loss": 4.9649,
      "mean_token_accuracy": 0.21847189962863922,
      "num_tokens": 65106786.0,
      "step": 29160
    },
    {
      "entropy": 5.2326826572418215,
      "epoch": 3.2755096310439713,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039559899298918094,
      "loss": 4.9196,
      "mean_token_accuracy": 0.21777488738298417,
      "num_tokens": 65117506.0,
      "step": 29165
    },
    {
      "entropy": 5.189513301849365,
      "epoch": 3.276071207951929,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003955650966847922,
      "loss": 4.8057,
      "mean_token_accuracy": 0.23231253623962403,
      "num_tokens": 65127830.0,
      "step": 29170
    },
    {
      "entropy": 5.302333068847656,
      "epoch": 3.2766327848598866,
      "grad_norm": 0.953125,
      "learning_rate": 0.000395531196541564,
      "loss": 4.9746,
      "mean_token_accuracy": 0.21494394093751906,
      "num_tokens": 65139562.0,
      "step": 29175
    },
    {
      "entropy": 5.3305806636810305,
      "epoch": 3.2771943617678443,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00039549729256057583,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21778286397457122,
      "num_tokens": 65150799.0,
      "step": 29180
    },
    {
      "entropy": 5.265044832229615,
      "epoch": 3.2777559386758015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000395463384742907,
      "loss": 4.9249,
      "mean_token_accuracy": 0.21859782040119172,
      "num_tokens": 65161458.0,
      "step": 29185
    },
    {
      "entropy": 5.256999826431274,
      "epoch": 3.278317515583759,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039542947308963733,
      "loss": 4.9031,
      "mean_token_accuracy": 0.22252779603004455,
      "num_tokens": 65172098.0,
      "step": 29190
    },
    {
      "entropy": 5.341102457046508,
      "epoch": 3.278879092491717,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003953955576018466,
      "loss": 4.9428,
      "mean_token_accuracy": 0.21419748961925505,
      "num_tokens": 65183668.0,
      "step": 29195
    },
    {
      "entropy": 5.2232757091522215,
      "epoch": 3.2794406693996745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003953616382806145,
      "loss": 4.8183,
      "mean_token_accuracy": 0.22968388646841048,
      "num_tokens": 65193920.0,
      "step": 29200
    },
    {
      "entropy": 5.276603364944458,
      "epoch": 3.2800022463076317,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00039532771512702124,
      "loss": 4.9732,
      "mean_token_accuracy": 0.21785025149583817,
      "num_tokens": 65203397.0,
      "step": 29205
    },
    {
      "entropy": 5.265945959091186,
      "epoch": 3.2805638232155894,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003952937881421468,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22108619660139084,
      "num_tokens": 65213821.0,
      "step": 29210
    },
    {
      "entropy": 5.226012182235718,
      "epoch": 3.281125400123547,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003952598573270716,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22446785867214203,
      "num_tokens": 65224034.0,
      "step": 29215
    },
    {
      "entropy": 5.178140544891358,
      "epoch": 3.2816869770315042,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003952259226828759,
      "loss": 4.87,
      "mean_token_accuracy": 0.2177480250597,
      "num_tokens": 65234879.0,
      "step": 29220
    },
    {
      "entropy": 5.291498851776123,
      "epoch": 3.282248553939462,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00039519198421064005,
      "loss": 4.8773,
      "mean_token_accuracy": 0.2257293865084648,
      "num_tokens": 65246530.0,
      "step": 29225
    },
    {
      "entropy": 5.234841918945312,
      "epoch": 3.2828101308474196,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039515804191144496,
      "loss": 4.9292,
      "mean_token_accuracy": 0.22149790823459625,
      "num_tokens": 65258102.0,
      "step": 29230
    },
    {
      "entropy": 5.2601710796356205,
      "epoch": 3.2833717077553772,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039512409578637114,
      "loss": 4.8486,
      "mean_token_accuracy": 0.23376694172620774,
      "num_tokens": 65269396.0,
      "step": 29235
    },
    {
      "entropy": 5.213965177536011,
      "epoch": 3.2839332846633345,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039509014583649947,
      "loss": 4.8185,
      "mean_token_accuracy": 0.23348421156406401,
      "num_tokens": 65280004.0,
      "step": 29240
    },
    {
      "entropy": 5.270637702941895,
      "epoch": 3.284494861571292,
      "grad_norm": 1.09375,
      "learning_rate": 0.00039505619206291106,
      "loss": 4.9312,
      "mean_token_accuracy": 0.22051673680543898,
      "num_tokens": 65291564.0,
      "step": 29245
    },
    {
      "entropy": 5.251637649536133,
      "epoch": 3.2850564384792498,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00039502223446668687,
      "loss": 4.8415,
      "mean_token_accuracy": 0.22357817888259887,
      "num_tokens": 65303032.0,
      "step": 29250
    },
    {
      "entropy": 5.172544479370117,
      "epoch": 3.2856180153872074,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003949882730489082,
      "loss": 4.7763,
      "mean_token_accuracy": 0.2251957595348358,
      "num_tokens": 65313658.0,
      "step": 29255
    },
    {
      "entropy": 5.361184024810791,
      "epoch": 3.2861795922951647,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039495430781065645,
      "loss": 5.0109,
      "mean_token_accuracy": 0.2194594547152519,
      "num_tokens": 65324593.0,
      "step": 29260
    },
    {
      "entropy": 5.344628429412841,
      "epoch": 3.2867411692031223,
      "grad_norm": 1.171875,
      "learning_rate": 0.000394920338753013,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22141306847333908,
      "num_tokens": 65335059.0,
      "step": 29265
    },
    {
      "entropy": 5.344312429428101,
      "epoch": 3.28730274611108,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003948863658770594,
      "loss": 5.0599,
      "mean_token_accuracy": 0.21767424494028093,
      "num_tokens": 65347082.0,
      "step": 29270
    },
    {
      "entropy": 5.156473779678345,
      "epoch": 3.2878643230190376,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003948523891838774,
      "loss": 4.8445,
      "mean_token_accuracy": 0.2207549810409546,
      "num_tokens": 65358644.0,
      "step": 29275
    },
    {
      "entropy": 5.291332769393921,
      "epoch": 3.288425899926995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039481840867454886,
      "loss": 4.9017,
      "mean_token_accuracy": 0.2178698167204857,
      "num_tokens": 65369652.0,
      "step": 29280
    },
    {
      "entropy": 5.312250423431396,
      "epoch": 3.2889874768349525,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003947844243501558,
      "loss": 4.8792,
      "mean_token_accuracy": 0.2238767609000206,
      "num_tokens": 65380196.0,
      "step": 29285
    },
    {
      "entropy": 5.225322246551514,
      "epoch": 3.28954905374291,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00039475043621178005,
      "loss": 4.8904,
      "mean_token_accuracy": 0.22042172402143478,
      "num_tokens": 65391135.0,
      "step": 29290
    },
    {
      "entropy": 5.31117992401123,
      "epoch": 3.290110630650868,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003947164442605041,
      "loss": 5.0383,
      "mean_token_accuracy": 0.21867503672838212,
      "num_tokens": 65402330.0,
      "step": 29295
    },
    {
      "entropy": 5.286956024169922,
      "epoch": 3.290672207558825,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003946824484974101,
      "loss": 4.9309,
      "mean_token_accuracy": 0.21749375760555267,
      "num_tokens": 65414111.0,
      "step": 29300
    },
    {
      "entropy": 5.314995574951172,
      "epoch": 3.2912337844667827,
      "grad_norm": 1.0,
      "learning_rate": 0.0003946484489235805,
      "loss": 4.94,
      "mean_token_accuracy": 0.22112611830234527,
      "num_tokens": 65424760.0,
      "step": 29305
    },
    {
      "entropy": 5.22037091255188,
      "epoch": 3.2917953613747404,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039461444554009787,
      "loss": 4.8561,
      "mean_token_accuracy": 0.22530371099710464,
      "num_tokens": 65435213.0,
      "step": 29310
    },
    {
      "entropy": 5.182293462753296,
      "epoch": 3.2923569382826976,
      "grad_norm": 0.96875,
      "learning_rate": 0.000394580438348045,
      "loss": 4.8349,
      "mean_token_accuracy": 0.2151091977953911,
      "num_tokens": 65446141.0,
      "step": 29315
    },
    {
      "entropy": 5.306439065933228,
      "epoch": 3.2929185151906553,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039454642734850447,
      "loss": 4.9521,
      "mean_token_accuracy": 0.22253102511167527,
      "num_tokens": 65457634.0,
      "step": 29320
    },
    {
      "entropy": 5.290516376495361,
      "epoch": 3.293480092098613,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003945124125425594,
      "loss": 4.9422,
      "mean_token_accuracy": 0.22143170982599258,
      "num_tokens": 65468457.0,
      "step": 29325
    },
    {
      "entropy": 5.3063336372375485,
      "epoch": 3.2940416690065706,
      "grad_norm": 1.0,
      "learning_rate": 0.0003944783939312928,
      "loss": 4.9105,
      "mean_token_accuracy": 0.2194645881652832,
      "num_tokens": 65479938.0,
      "step": 29330
    },
    {
      "entropy": 5.292670822143554,
      "epoch": 3.294603245914528,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003944443715157877,
      "loss": 4.9673,
      "mean_token_accuracy": 0.21157462745904923,
      "num_tokens": 65492030.0,
      "step": 29335
    },
    {
      "entropy": 5.370320415496826,
      "epoch": 3.2951648228224855,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003944103452971275,
      "loss": 4.9662,
      "mean_token_accuracy": 0.22012887448072432,
      "num_tokens": 65503385.0,
      "step": 29340
    },
    {
      "entropy": 5.189487934112549,
      "epoch": 3.295726399730443,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039437631527639553,
      "loss": 4.8283,
      "mean_token_accuracy": 0.22629730254411698,
      "num_tokens": 65514027.0,
      "step": 29345
    },
    {
      "entropy": 5.208918285369873,
      "epoch": 3.296287976638401,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039434228145467554,
      "loss": 4.849,
      "mean_token_accuracy": 0.2191474109888077,
      "num_tokens": 65525653.0,
      "step": 29350
    },
    {
      "entropy": 5.185562562942505,
      "epoch": 3.296849553546358,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000394308243833051,
      "loss": 4.8331,
      "mean_token_accuracy": 0.22019879072904586,
      "num_tokens": 65536040.0,
      "step": 29355
    },
    {
      "entropy": 5.369727087020874,
      "epoch": 3.2974111304543157,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039427420241260557,
      "loss": 5.0354,
      "mean_token_accuracy": 0.21384665071964265,
      "num_tokens": 65547505.0,
      "step": 29360
    },
    {
      "entropy": 5.34031867980957,
      "epoch": 3.2979727073622733,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00039424015719442335,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22208406329154967,
      "num_tokens": 65557565.0,
      "step": 29365
    },
    {
      "entropy": 5.226785707473755,
      "epoch": 3.298534284270231,
      "grad_norm": 1.0,
      "learning_rate": 0.00039420610817958836,
      "loss": 4.9272,
      "mean_token_accuracy": 0.21851796656847,
      "num_tokens": 65568729.0,
      "step": 29370
    },
    {
      "entropy": 5.280762624740601,
      "epoch": 3.299095861178188,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003941720553691846,
      "loss": 5.0156,
      "mean_token_accuracy": 0.21419240087270736,
      "num_tokens": 65580664.0,
      "step": 29375
    },
    {
      "entropy": 5.232279491424561,
      "epoch": 3.299657438086146,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003941379987642963,
      "loss": 4.7986,
      "mean_token_accuracy": 0.22948629707098006,
      "num_tokens": 65591348.0,
      "step": 29380
    },
    {
      "entropy": 5.291241931915283,
      "epoch": 3.3002190149941035,
      "grad_norm": 1.03125,
      "learning_rate": 0.000394103938366008,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2227160304784775,
      "num_tokens": 65603553.0,
      "step": 29385
    },
    {
      "entropy": 5.250101900100708,
      "epoch": 3.300780591902061,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003940698741754041,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22791292667388915,
      "num_tokens": 65613903.0,
      "step": 29390
    },
    {
      "entropy": 5.250937557220459,
      "epoch": 3.3013421688100184,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003940358061935692,
      "loss": 4.9116,
      "mean_token_accuracy": 0.22265274822711945,
      "num_tokens": 65624862.0,
      "step": 29395
    },
    {
      "entropy": 5.230651712417602,
      "epoch": 3.301903745717976,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003940017344215881,
      "loss": 4.8708,
      "mean_token_accuracy": 0.22411825656890869,
      "num_tokens": 65636828.0,
      "step": 29400
    },
    {
      "entropy": 5.299429035186767,
      "epoch": 3.3024653226259337,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039396765886054555,
      "loss": 4.9518,
      "mean_token_accuracy": 0.2198137879371643,
      "num_tokens": 65648251.0,
      "step": 29405
    },
    {
      "entropy": 5.325869512557984,
      "epoch": 3.303026899533891,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039393357951152667,
      "loss": 5.0777,
      "mean_token_accuracy": 0.21039717942476271,
      "num_tokens": 65659715.0,
      "step": 29410
    },
    {
      "entropy": 5.322327136993408,
      "epoch": 3.3035884764418486,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039389949637561646,
      "loss": 4.9507,
      "mean_token_accuracy": 0.21916542947292328,
      "num_tokens": 65670844.0,
      "step": 29415
    },
    {
      "entropy": 5.200179767608643,
      "epoch": 3.3041500533498063,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039386540945390017,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22924821227788925,
      "num_tokens": 65682404.0,
      "step": 29420
    },
    {
      "entropy": 5.281607961654663,
      "epoch": 3.304711630257764,
      "grad_norm": 1.0,
      "learning_rate": 0.0003938313187474631,
      "loss": 4.9143,
      "mean_token_accuracy": 0.21642733365297318,
      "num_tokens": 65692826.0,
      "step": 29425
    },
    {
      "entropy": 5.26664662361145,
      "epoch": 3.3052732071657216,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003937972242573908,
      "loss": 4.9424,
      "mean_token_accuracy": 0.21653160750865935,
      "num_tokens": 65704139.0,
      "step": 29430
    },
    {
      "entropy": 5.225437641143799,
      "epoch": 3.305834784073679,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003937631259847688,
      "loss": 4.8908,
      "mean_token_accuracy": 0.21747539788484574,
      "num_tokens": 65716282.0,
      "step": 29435
    },
    {
      "entropy": 5.328140926361084,
      "epoch": 3.3063963609816365,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003937290239306828,
      "loss": 5.0206,
      "mean_token_accuracy": 0.2142635241150856,
      "num_tokens": 65728086.0,
      "step": 29440
    },
    {
      "entropy": 5.340057897567749,
      "epoch": 3.306957937889594,
      "grad_norm": 0.90625,
      "learning_rate": 0.00039369491809621867,
      "loss": 5.0194,
      "mean_token_accuracy": 0.21306005120277405,
      "num_tokens": 65740584.0,
      "step": 29445
    },
    {
      "entropy": 5.296872615814209,
      "epoch": 3.3075195147975514,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003936608084824623,
      "loss": 4.9532,
      "mean_token_accuracy": 0.2169371575117111,
      "num_tokens": 65751479.0,
      "step": 29450
    },
    {
      "entropy": 5.266234254837036,
      "epoch": 3.308081091705509,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039362669509049975,
      "loss": 4.967,
      "mean_token_accuracy": 0.21790872216224672,
      "num_tokens": 65762073.0,
      "step": 29455
    },
    {
      "entropy": 5.2341350555419925,
      "epoch": 3.3086426686134667,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039359257792141717,
      "loss": 4.8616,
      "mean_token_accuracy": 0.21925610601902007,
      "num_tokens": 65772808.0,
      "step": 29460
    },
    {
      "entropy": 5.264393329620361,
      "epoch": 3.3092042455214243,
      "grad_norm": 0.9765625,
      "learning_rate": 0.000393558456976301,
      "loss": 4.9029,
      "mean_token_accuracy": 0.220741106569767,
      "num_tokens": 65783507.0,
      "step": 29465
    },
    {
      "entropy": 5.2738793849945065,
      "epoch": 3.3097658224293816,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003935243322562375,
      "loss": 4.9296,
      "mean_token_accuracy": 0.21851243823766708,
      "num_tokens": 65794250.0,
      "step": 29470
    },
    {
      "entropy": 5.268389940261841,
      "epoch": 3.310327399337339,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003934902037623133,
      "loss": 4.8992,
      "mean_token_accuracy": 0.2157019779086113,
      "num_tokens": 65805487.0,
      "step": 29475
    },
    {
      "entropy": 5.272819566726684,
      "epoch": 3.310888976245297,
      "grad_norm": 0.984375,
      "learning_rate": 0.00039345607149561513,
      "loss": 4.9533,
      "mean_token_accuracy": 0.21239990144968032,
      "num_tokens": 65818264.0,
      "step": 29480
    },
    {
      "entropy": 5.365408420562744,
      "epoch": 3.3114505531532545,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039342193545722966,
      "loss": 5.0407,
      "mean_token_accuracy": 0.223477341234684,
      "num_tokens": 65829686.0,
      "step": 29485
    },
    {
      "entropy": 5.27428126335144,
      "epoch": 3.3120121300612118,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003933877956482438,
      "loss": 4.8376,
      "mean_token_accuracy": 0.22808793932199478,
      "num_tokens": 65840800.0,
      "step": 29490
    },
    {
      "entropy": 5.299081516265869,
      "epoch": 3.3125737069691694,
      "grad_norm": 0.9375,
      "learning_rate": 0.00039335365206974465,
      "loss": 4.9443,
      "mean_token_accuracy": 0.21734229028224944,
      "num_tokens": 65853634.0,
      "step": 29495
    },
    {
      "entropy": 5.240578269958496,
      "epoch": 3.313135283877127,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003933195047228193,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2143983796238899,
      "num_tokens": 65864754.0,
      "step": 29500
    },
    {
      "entropy": 5.233815860748291,
      "epoch": 3.3136968607850843,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00039328535360855503,
      "loss": 4.9444,
      "mean_token_accuracy": 0.2168400451540947,
      "num_tokens": 65876609.0,
      "step": 29505
    },
    {
      "entropy": 5.28493824005127,
      "epoch": 3.314258437693042,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00039325119872803924,
      "loss": 4.8989,
      "mean_token_accuracy": 0.2173907056450844,
      "num_tokens": 65887583.0,
      "step": 29510
    },
    {
      "entropy": 5.281411981582641,
      "epoch": 3.3148200146009996,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039321704008235943,
      "loss": 4.848,
      "mean_token_accuracy": 0.23327846378087996,
      "num_tokens": 65898131.0,
      "step": 29515
    },
    {
      "entropy": 5.219368124008179,
      "epoch": 3.3153815915089573,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039318287767260316,
      "loss": 4.8781,
      "mean_token_accuracy": 0.2165673106908798,
      "num_tokens": 65909353.0,
      "step": 29520
    },
    {
      "entropy": 5.245532894134522,
      "epoch": 3.315943168416915,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00039314871149985834,
      "loss": 4.9114,
      "mean_token_accuracy": 0.22200802564620972,
      "num_tokens": 65920709.0,
      "step": 29525
    },
    {
      "entropy": 5.307961511611938,
      "epoch": 3.316504745324872,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003931145415652126,
      "loss": 4.9694,
      "mean_token_accuracy": 0.2234126016497612,
      "num_tokens": 65932231.0,
      "step": 29530
    },
    {
      "entropy": 5.39408130645752,
      "epoch": 3.31706632223283,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039308036786975416,
      "loss": 5.0459,
      "mean_token_accuracy": 0.21216216832399368,
      "num_tokens": 65942629.0,
      "step": 29535
    },
    {
      "entropy": 5.32193341255188,
      "epoch": 3.3176278991407875,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003930461904145709,
      "loss": 4.8625,
      "mean_token_accuracy": 0.22443986684083939,
      "num_tokens": 65954385.0,
      "step": 29540
    },
    {
      "entropy": 5.254720449447632,
      "epoch": 3.3181894760487447,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003930120092007512,
      "loss": 4.9164,
      "mean_token_accuracy": 0.22190654575824736,
      "num_tokens": 65966283.0,
      "step": 29545
    },
    {
      "entropy": 5.360729455947876,
      "epoch": 3.3187510529567024,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003929778242293834,
      "loss": 5.0718,
      "mean_token_accuracy": 0.2115269497036934,
      "num_tokens": 65978594.0,
      "step": 29550
    },
    {
      "entropy": 5.226860523223877,
      "epoch": 3.31931262986466,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003929436355015558,
      "loss": 4.8323,
      "mean_token_accuracy": 0.21881127506494522,
      "num_tokens": 65989398.0,
      "step": 29555
    },
    {
      "entropy": 5.243882703781128,
      "epoch": 3.3198742067726177,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039290944301835723,
      "loss": 4.9358,
      "mean_token_accuracy": 0.22289599627256393,
      "num_tokens": 66000545.0,
      "step": 29560
    },
    {
      "entropy": 5.278307104110718,
      "epoch": 3.320435783680575,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039287524678087617,
      "loss": 4.9205,
      "mean_token_accuracy": 0.21996400356292725,
      "num_tokens": 66011638.0,
      "step": 29565
    },
    {
      "entropy": 5.2391520023345945,
      "epoch": 3.3209973605885326,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00039284104679020154,
      "loss": 4.8821,
      "mean_token_accuracy": 0.21641977578401567,
      "num_tokens": 66023284.0,
      "step": 29570
    },
    {
      "entropy": 5.2449479579925535,
      "epoch": 3.3215589374964902,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003928068430474224,
      "loss": 4.9133,
      "mean_token_accuracy": 0.2175465404987335,
      "num_tokens": 66034099.0,
      "step": 29575
    },
    {
      "entropy": 5.203459405899048,
      "epoch": 3.322120514404448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003927726355536275,
      "loss": 4.8397,
      "mean_token_accuracy": 0.22702266722917558,
      "num_tokens": 66043885.0,
      "step": 29580
    },
    {
      "entropy": 5.278463554382324,
      "epoch": 3.322682091312405,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039273842430990634,
      "loss": 4.9767,
      "mean_token_accuracy": 0.2108403444290161,
      "num_tokens": 66054190.0,
      "step": 29585
    },
    {
      "entropy": 5.264446067810058,
      "epoch": 3.3232436682203628,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003927042093173481,
      "loss": 4.9293,
      "mean_token_accuracy": 0.22127682715654373,
      "num_tokens": 66065254.0,
      "step": 29590
    },
    {
      "entropy": 5.330303430557251,
      "epoch": 3.3238052451283204,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003926699905770421,
      "loss": 5.0188,
      "mean_token_accuracy": 0.21513386070728302,
      "num_tokens": 66076582.0,
      "step": 29595
    },
    {
      "entropy": 5.3754496574401855,
      "epoch": 3.3243668220362776,
      "grad_norm": 1.046875,
      "learning_rate": 0.000392635768090078,
      "loss": 4.9689,
      "mean_token_accuracy": 0.21615650206804277,
      "num_tokens": 66086410.0,
      "step": 29600
    },
    {
      "entropy": 5.318622159957886,
      "epoch": 3.3249283989442353,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00039260154185754545,
      "loss": 4.9897,
      "mean_token_accuracy": 0.21002089381217956,
      "num_tokens": 66098393.0,
      "step": 29605
    },
    {
      "entropy": 5.335929822921753,
      "epoch": 3.325489975852193,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003925673118805341,
      "loss": 4.9574,
      "mean_token_accuracy": 0.21979565769433976,
      "num_tokens": 66108077.0,
      "step": 29610
    },
    {
      "entropy": 5.1495294094085695,
      "epoch": 3.3260515527601506,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000392533078160134,
      "loss": 4.7231,
      "mean_token_accuracy": 0.236220346391201,
      "num_tokens": 66118008.0,
      "step": 29615
    },
    {
      "entropy": 5.274313974380493,
      "epoch": 3.3266131296681083,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00039249884069743514,
      "loss": 5.0064,
      "mean_token_accuracy": 0.2189728394150734,
      "num_tokens": 66128697.0,
      "step": 29620
    },
    {
      "entropy": 5.338425493240356,
      "epoch": 3.3271747065760655,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003924645994935276,
      "loss": 4.9418,
      "mean_token_accuracy": 0.21719348579645156,
      "num_tokens": 66139800.0,
      "step": 29625
    },
    {
      "entropy": 5.255507707595825,
      "epoch": 3.327736283484023,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003924303545495017,
      "loss": 4.8266,
      "mean_token_accuracy": 0.2282159686088562,
      "num_tokens": 66151202.0,
      "step": 29630
    },
    {
      "entropy": 5.30038366317749,
      "epoch": 3.328297860391981,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039239610586644765,
      "loss": 4.9273,
      "mean_token_accuracy": 0.21798139810562134,
      "num_tokens": 66162159.0,
      "step": 29635
    },
    {
      "entropy": 5.28494029045105,
      "epoch": 3.328859437299938,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003923618534454561,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2208869606256485,
      "num_tokens": 66174492.0,
      "step": 29640
    },
    {
      "entropy": 5.248617553710938,
      "epoch": 3.3294210142078957,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003923275972876176,
      "loss": 4.9608,
      "mean_token_accuracy": 0.22529327720403672,
      "num_tokens": 66186617.0,
      "step": 29645
    },
    {
      "entropy": 5.230626821517944,
      "epoch": 3.3299825911158534,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039229333739402303,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22380739748477935,
      "num_tokens": 66198001.0,
      "step": 29650
    },
    {
      "entropy": 5.218782520294189,
      "epoch": 3.330544168023811,
      "grad_norm": 0.96875,
      "learning_rate": 0.000392259073765763,
      "loss": 4.8699,
      "mean_token_accuracy": 0.2231778547167778,
      "num_tokens": 66208017.0,
      "step": 29655
    },
    {
      "entropy": 5.183077144622803,
      "epoch": 3.3311057449317683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039222480640392854,
      "loss": 4.857,
      "mean_token_accuracy": 0.22449653446674347,
      "num_tokens": 66219798.0,
      "step": 29660
    },
    {
      "entropy": 5.295391273498535,
      "epoch": 3.331667321839726,
      "grad_norm": 1.0,
      "learning_rate": 0.00039219053530961087,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22599872350692748,
      "num_tokens": 66230662.0,
      "step": 29665
    },
    {
      "entropy": 5.302159929275513,
      "epoch": 3.3322288987476836,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039215626048390105,
      "loss": 4.8789,
      "mean_token_accuracy": 0.2242247372865677,
      "num_tokens": 66240787.0,
      "step": 29670
    },
    {
      "entropy": 5.197985410690308,
      "epoch": 3.3327904756556412,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003921219819278905,
      "loss": 4.8097,
      "mean_token_accuracy": 0.222908815741539,
      "num_tokens": 66250937.0,
      "step": 29675
    },
    {
      "entropy": 5.338726282119751,
      "epoch": 3.3333520525635985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003920876996426706,
      "loss": 5.034,
      "mean_token_accuracy": 0.21438297182321547,
      "num_tokens": 66261322.0,
      "step": 29680
    },
    {
      "entropy": 5.255003595352173,
      "epoch": 3.333913629471556,
      "grad_norm": 1.0625,
      "learning_rate": 0.00039205341362933294,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2225881353020668,
      "num_tokens": 66271307.0,
      "step": 29685
    },
    {
      "entropy": 5.213127613067627,
      "epoch": 3.334475206379514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003920191238889691,
      "loss": 4.8251,
      "mean_token_accuracy": 0.2285887762904167,
      "num_tokens": 66283067.0,
      "step": 29690
    },
    {
      "entropy": 5.2822685718536375,
      "epoch": 3.335036783287471,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003919848304226711,
      "loss": 4.9463,
      "mean_token_accuracy": 0.21282953917980194,
      "num_tokens": 66295853.0,
      "step": 29695
    },
    {
      "entropy": 5.301497936248779,
      "epoch": 3.3355983601954287,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003919505332315307,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2229978233575821,
      "num_tokens": 66307573.0,
      "step": 29700
    },
    {
      "entropy": 5.228827476501465,
      "epoch": 3.3361599371033863,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039191623231663985,
      "loss": 4.8642,
      "mean_token_accuracy": 0.23003108203411102,
      "num_tokens": 66318100.0,
      "step": 29705
    },
    {
      "entropy": 5.27058424949646,
      "epoch": 3.336721514011344,
      "grad_norm": 1.03125,
      "learning_rate": 0.000391881927679091,
      "loss": 4.9084,
      "mean_token_accuracy": 0.2176363453269005,
      "num_tokens": 66328918.0,
      "step": 29710
    },
    {
      "entropy": 5.229003429412842,
      "epoch": 3.3372830909193016,
      "grad_norm": 1.046875,
      "learning_rate": 0.00039184761931997607,
      "loss": 4.9846,
      "mean_token_accuracy": 0.21401261240243913,
      "num_tokens": 66341124.0,
      "step": 29715
    },
    {
      "entropy": 5.323524808883667,
      "epoch": 3.337844667827259,
      "grad_norm": 1.0,
      "learning_rate": 0.0003918133072403878,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2205760583281517,
      "num_tokens": 66351033.0,
      "step": 29720
    },
    {
      "entropy": 5.274878215789795,
      "epoch": 3.3384062447352165,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00039177899144141835,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2254047065973282,
      "num_tokens": 66363338.0,
      "step": 29725
    },
    {
      "entropy": 5.247694444656372,
      "epoch": 3.338967821643174,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003917446719241606,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22448271214962007,
      "num_tokens": 66373716.0,
      "step": 29730
    },
    {
      "entropy": 5.288740491867065,
      "epoch": 3.3395293985511314,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039171034868970724,
      "loss": 4.8871,
      "mean_token_accuracy": 0.2225578770041466,
      "num_tokens": 66384448.0,
      "step": 29735
    },
    {
      "entropy": 5.1831001281738285,
      "epoch": 3.340090975459089,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00039167602173915104,
      "loss": 4.8367,
      "mean_token_accuracy": 0.2248813822865486,
      "num_tokens": 66395662.0,
      "step": 29740
    },
    {
      "entropy": 5.134633207321167,
      "epoch": 3.3406525523670467,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003916416910735851,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22277145087718964,
      "num_tokens": 66407856.0,
      "step": 29745
    },
    {
      "entropy": 5.339507389068603,
      "epoch": 3.3412141292750044,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00039160735669410246,
      "loss": 5.0133,
      "mean_token_accuracy": 0.2068480908870697,
      "num_tokens": 66419324.0,
      "step": 29750
    },
    {
      "entropy": 5.232793712615967,
      "epoch": 3.3417757061829616,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003915730186017964,
      "loss": 4.8178,
      "mean_token_accuracy": 0.22253818511962892,
      "num_tokens": 66430720.0,
      "step": 29755
    },
    {
      "entropy": 5.277332592010498,
      "epoch": 3.3423372830909193,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039153867679776023,
      "loss": 4.9253,
      "mean_token_accuracy": 0.21515196710824966,
      "num_tokens": 66441736.0,
      "step": 29760
    },
    {
      "entropy": 5.343836402893066,
      "epoch": 3.342898859998877,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003915043312830873,
      "loss": 4.9815,
      "mean_token_accuracy": 0.2216696783900261,
      "num_tokens": 66451851.0,
      "step": 29765
    },
    {
      "entropy": 5.378570985794068,
      "epoch": 3.3434604369068346,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003914699820588714,
      "loss": 5.0208,
      "mean_token_accuracy": 0.21087879538536072,
      "num_tokens": 66463105.0,
      "step": 29770
    },
    {
      "entropy": 5.3018941402435305,
      "epoch": 3.344022013814792,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039143562912620603,
      "loss": 5.0384,
      "mean_token_accuracy": 0.21687722355127334,
      "num_tokens": 66474613.0,
      "step": 29775
    },
    {
      "entropy": 5.263784170150757,
      "epoch": 3.3445835907227495,
      "grad_norm": 1.015625,
      "learning_rate": 0.000391401272486185,
      "loss": 4.9718,
      "mean_token_accuracy": 0.2149228647351265,
      "num_tokens": 66486783.0,
      "step": 29780
    },
    {
      "entropy": 5.391638374328613,
      "epoch": 3.345145167630707,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003913669121399024,
      "loss": 5.0024,
      "mean_token_accuracy": 0.21541035026311875,
      "num_tokens": 66498581.0,
      "step": 29785
    },
    {
      "entropy": 5.224271249771118,
      "epoch": 3.3457067445386643,
      "grad_norm": 1.03125,
      "learning_rate": 0.00039133254808845216,
      "loss": 4.8689,
      "mean_token_accuracy": 0.22818415015935897,
      "num_tokens": 66509480.0,
      "step": 29790
    },
    {
      "entropy": 5.245246362686157,
      "epoch": 3.346268321446622,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00039129818033292843,
      "loss": 4.8581,
      "mean_token_accuracy": 0.22205984592437744,
      "num_tokens": 66521831.0,
      "step": 29795
    },
    {
      "entropy": 5.282585096359253,
      "epoch": 3.3468298983545797,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003912638088744256,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22061897516250611,
      "num_tokens": 66533900.0,
      "step": 29800
    },
    {
      "entropy": 5.3984612941741945,
      "epoch": 3.3473914752625373,
      "grad_norm": 1.078125,
      "learning_rate": 0.000391229433714038,
      "loss": 5.0398,
      "mean_token_accuracy": 0.20800905972719191,
      "num_tokens": 66543848.0,
      "step": 29805
    },
    {
      "entropy": 5.3054646968841555,
      "epoch": 3.347953052170495,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003911950548528601,
      "loss": 4.913,
      "mean_token_accuracy": 0.21753843873739243,
      "num_tokens": 66554113.0,
      "step": 29810
    },
    {
      "entropy": 5.328948593139648,
      "epoch": 3.348514629078452,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003911606722919866,
      "loss": 4.9555,
      "mean_token_accuracy": 0.2175936222076416,
      "num_tokens": 66565202.0,
      "step": 29815
    },
    {
      "entropy": 5.312395095825195,
      "epoch": 3.34907620598641,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003911262860325123,
      "loss": 5.0079,
      "mean_token_accuracy": 0.21201431155204772,
      "num_tokens": 66575930.0,
      "step": 29820
    },
    {
      "entropy": 5.323049783706665,
      "epoch": 3.3496377828943675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003910918960755319,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2233863055706024,
      "num_tokens": 66585422.0,
      "step": 29825
    },
    {
      "entropy": 5.303429985046387,
      "epoch": 3.3501993598023248,
      "grad_norm": 1.125,
      "learning_rate": 0.0003910575024221406,
      "loss": 4.918,
      "mean_token_accuracy": 0.21919730454683303,
      "num_tokens": 66597395.0,
      "step": 29830
    },
    {
      "entropy": 5.284577989578247,
      "epoch": 3.3507609367102824,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039102310507343335,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2231140285730362,
      "num_tokens": 66608464.0,
      "step": 29835
    },
    {
      "entropy": 5.3256267547607425,
      "epoch": 3.35132251361824,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003909887040305055,
      "loss": 4.9707,
      "mean_token_accuracy": 0.2152840957045555,
      "num_tokens": 66619930.0,
      "step": 29840
    },
    {
      "entropy": 5.218983125686646,
      "epoch": 3.3518840905261977,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003909542992944523,
      "loss": 4.9239,
      "mean_token_accuracy": 0.22169507443904876,
      "num_tokens": 66632833.0,
      "step": 29845
    },
    {
      "entropy": 5.26315655708313,
      "epoch": 3.352445667434155,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003909198908663692,
      "loss": 4.9527,
      "mean_token_accuracy": 0.21887608915567397,
      "num_tokens": 66643641.0,
      "step": 29850
    },
    {
      "entropy": 5.397197008132935,
      "epoch": 3.3530072443421126,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00039088547874735187,
      "loss": 4.9971,
      "mean_token_accuracy": 0.20936987549066544,
      "num_tokens": 66653863.0,
      "step": 29855
    },
    {
      "entropy": 5.2260161399841305,
      "epoch": 3.3535688212500703,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00039085106293849594,
      "loss": 4.8507,
      "mean_token_accuracy": 0.2279619723558426,
      "num_tokens": 66665885.0,
      "step": 29860
    },
    {
      "entropy": 5.27515811920166,
      "epoch": 3.354130398158028,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039081664344089724,
      "loss": 4.9306,
      "mean_token_accuracy": 0.22203658819198607,
      "num_tokens": 66676584.0,
      "step": 29865
    },
    {
      "entropy": 5.175761699676514,
      "epoch": 3.354691975065985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003907822202556516,
      "loss": 4.7964,
      "mean_token_accuracy": 0.23188765197992325,
      "num_tokens": 66687882.0,
      "step": 29870
    },
    {
      "entropy": 5.29911994934082,
      "epoch": 3.355253551973943,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003907477933838553,
      "loss": 4.9078,
      "mean_token_accuracy": 0.21417330652475358,
      "num_tokens": 66699665.0,
      "step": 29875
    },
    {
      "entropy": 5.248944854736328,
      "epoch": 3.3558151288819005,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00039071336282660437,
      "loss": 5.0129,
      "mean_token_accuracy": 0.21342599391937256,
      "num_tokens": 66711455.0,
      "step": 29880
    },
    {
      "entropy": 5.211044454574585,
      "epoch": 3.3563767057898577,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00039067892858499496,
      "loss": 4.813,
      "mean_token_accuracy": 0.2324649140238762,
      "num_tokens": 66722369.0,
      "step": 29885
    },
    {
      "entropy": 5.262921905517578,
      "epoch": 3.3569382826978154,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003906444906601237,
      "loss": 4.9571,
      "mean_token_accuracy": 0.22257129997015,
      "num_tokens": 66733200.0,
      "step": 29890
    },
    {
      "entropy": 5.260540008544922,
      "epoch": 3.357499859605773,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000390610049053087,
      "loss": 4.865,
      "mean_token_accuracy": 0.2262023374438286,
      "num_tokens": 66744650.0,
      "step": 29895
    },
    {
      "entropy": 5.299485874176026,
      "epoch": 3.3580614365137307,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00039057560376498147,
      "loss": 4.9356,
      "mean_token_accuracy": 0.21231392174959182,
      "num_tokens": 66755712.0,
      "step": 29900
    },
    {
      "entropy": 5.309061527252197,
      "epoch": 3.3586230134216883,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003905411547969038,
      "loss": 4.8973,
      "mean_token_accuracy": 0.22537981867790222,
      "num_tokens": 66765855.0,
      "step": 29905
    },
    {
      "entropy": 5.27924861907959,
      "epoch": 3.3591845903296456,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003905067021499511,
      "loss": 4.95,
      "mean_token_accuracy": 0.22592354118824004,
      "num_tokens": 66776870.0,
      "step": 29910
    },
    {
      "entropy": 5.247446918487549,
      "epoch": 3.359746167237603,
      "grad_norm": 1.015625,
      "learning_rate": 0.00039047224582522015,
      "loss": 4.8891,
      "mean_token_accuracy": 0.22365480214357375,
      "num_tokens": 66787226.0,
      "step": 29915
    },
    {
      "entropy": 5.2572887420654295,
      "epoch": 3.360307744145561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003904377858238081,
      "loss": 4.865,
      "mean_token_accuracy": 0.22278422862291336,
      "num_tokens": 66797352.0,
      "step": 29920
    },
    {
      "entropy": 5.329019927978516,
      "epoch": 3.360869321053518,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003904033221468122,
      "loss": 5.0316,
      "mean_token_accuracy": 0.20926871448755263,
      "num_tokens": 66809227.0,
      "step": 29925
    },
    {
      "entropy": 5.322204494476319,
      "epoch": 3.3614308979614758,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003903688547953297,
      "loss": 4.931,
      "mean_token_accuracy": 0.21629197299480438,
      "num_tokens": 66821196.0,
      "step": 29930
    },
    {
      "entropy": 5.289442539215088,
      "epoch": 3.3619924748694334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003903343837704582,
      "loss": 4.9394,
      "mean_token_accuracy": 0.21767635345458985,
      "num_tokens": 66831958.0,
      "step": 29935
    },
    {
      "entropy": 5.300126695632935,
      "epoch": 3.362554051777391,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003902999090732951,
      "loss": 4.9145,
      "mean_token_accuracy": 0.22198840379714965,
      "num_tokens": 66843727.0,
      "step": 29940
    },
    {
      "entropy": 5.233318662643432,
      "epoch": 3.3631156286853483,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00039026543070493823,
      "loss": 4.8997,
      "mean_token_accuracy": 0.223402938246727,
      "num_tokens": 66854756.0,
      "step": 29945
    },
    {
      "entropy": 5.169430541992187,
      "epoch": 3.363677205593306,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003902309486664854,
      "loss": 4.8583,
      "mean_token_accuracy": 0.2271970346570015,
      "num_tokens": 66865615.0,
      "step": 29950
    },
    {
      "entropy": 5.267844152450562,
      "epoch": 3.3642387825012636,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003901964629590344,
      "loss": 4.9313,
      "mean_token_accuracy": 0.2160376489162445,
      "num_tokens": 66877262.0,
      "step": 29955
    },
    {
      "entropy": 5.402543067932129,
      "epoch": 3.3648003594092213,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003901619735836833,
      "loss": 5.0064,
      "mean_token_accuracy": 0.20811809450387955,
      "num_tokens": 66889405.0,
      "step": 29960
    },
    {
      "entropy": 5.269650077819824,
      "epoch": 3.3653619363171785,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00039012748054153044,
      "loss": 4.8745,
      "mean_token_accuracy": 0.22594607174396514,
      "num_tokens": 66900013.0,
      "step": 29965
    },
    {
      "entropy": 5.18872480392456,
      "epoch": 3.365923513225136,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003900929838336738,
      "loss": 4.7681,
      "mean_token_accuracy": 0.2301899015903473,
      "num_tokens": 66909731.0,
      "step": 29970
    },
    {
      "entropy": 5.187195110321045,
      "epoch": 3.366485090133094,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003900584834612119,
      "loss": 4.8476,
      "mean_token_accuracy": 0.22823408544063567,
      "num_tokens": 66919731.0,
      "step": 29975
    },
    {
      "entropy": 5.31118540763855,
      "epoch": 3.367046667041051,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003900239794252434,
      "loss": 4.9017,
      "mean_token_accuracy": 0.21786898225545884,
      "num_tokens": 66930169.0,
      "step": 29980
    },
    {
      "entropy": 5.263034439086914,
      "epoch": 3.3676082439490087,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003899894717268667,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22452257722616195,
      "num_tokens": 66942010.0,
      "step": 29985
    },
    {
      "entropy": 5.257385110855102,
      "epoch": 3.3681698208569664,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003899549603671807,
      "loss": 4.8709,
      "mean_token_accuracy": 0.23088784515857697,
      "num_tokens": 66953972.0,
      "step": 29990
    },
    {
      "entropy": 5.288667869567871,
      "epoch": 3.368731397764924,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00038992044534728404,
      "loss": 4.9927,
      "mean_token_accuracy": 0.21776093244552613,
      "num_tokens": 66965548.0,
      "step": 29995
    },
    {
      "entropy": 5.262576580047607,
      "epoch": 3.3692929746728817,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038988592666827594,
      "loss": 4.9063,
      "mean_token_accuracy": 0.22048855125904082,
      "num_tokens": 66976414.0,
      "step": 30000
    },
    {
      "epoch": 3.3692929746728817,
      "eval_entropy": 5.1318283283990285,
      "eval_loss": 5.096953392028809,
      "eval_mean_token_accuracy": 0.21843377524794208,
      "eval_num_tokens": 66976414.0,
      "eval_runtime": 24.0961,
      "eval_samples_per_second": 1287.052,
      "eval_steps_per_second": 160.897,
      "step": 30000
    },
    {
      "entropy": 5.184066534042358,
      "epoch": 3.369854551580839,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003898514043312553,
      "loss": 4.8761,
      "mean_token_accuracy": 0.22621785402297973,
      "num_tokens": 66987392.0,
      "step": 30005
    },
    {
      "entropy": 5.425533962249756,
      "epoch": 3.3704161284887966,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038981687833732146,
      "loss": 5.0687,
      "mean_token_accuracy": 0.21159074157476426,
      "num_tokens": 67000444.0,
      "step": 30010
    },
    {
      "entropy": 5.309379625320434,
      "epoch": 3.3709777053967542,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003897823486875737,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21984632611274718,
      "num_tokens": 67012124.0,
      "step": 30015
    },
    {
      "entropy": 5.333197116851807,
      "epoch": 3.3715392823047114,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003897478153831114,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21816502809524535,
      "num_tokens": 67024179.0,
      "step": 30020
    },
    {
      "entropy": 5.27920184135437,
      "epoch": 3.372100859212669,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038971327842503414,
      "loss": 4.8952,
      "mean_token_accuracy": 0.2229463517665863,
      "num_tokens": 67035876.0,
      "step": 30025
    },
    {
      "entropy": 5.187250232696533,
      "epoch": 3.3726624361206268,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038967873781444167,
      "loss": 4.8638,
      "mean_token_accuracy": 0.22417571395635605,
      "num_tokens": 67046499.0,
      "step": 30030
    },
    {
      "entropy": 5.255889844894409,
      "epoch": 3.3732240130285844,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038964419355243365,
      "loss": 4.9078,
      "mean_token_accuracy": 0.22226639837026596,
      "num_tokens": 67058357.0,
      "step": 30035
    },
    {
      "entropy": 5.307149362564087,
      "epoch": 3.3737855899365417,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003896096456401101,
      "loss": 4.9977,
      "mean_token_accuracy": 0.21487953811883925,
      "num_tokens": 67069753.0,
      "step": 30040
    },
    {
      "entropy": 5.242362117767334,
      "epoch": 3.3743471668444993,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038957509407857097,
      "loss": 4.8997,
      "mean_token_accuracy": 0.2226392850279808,
      "num_tokens": 67080185.0,
      "step": 30045
    },
    {
      "entropy": 5.270830535888672,
      "epoch": 3.374908743752457,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038954053886891635,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2218499854207039,
      "num_tokens": 67091382.0,
      "step": 30050
    },
    {
      "entropy": 5.38858060836792,
      "epoch": 3.3754703206604146,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003895059800122466,
      "loss": 5.0737,
      "mean_token_accuracy": 0.20049283802509307,
      "num_tokens": 67102952.0,
      "step": 30055
    },
    {
      "entropy": 5.230470657348633,
      "epoch": 3.376031897568372,
      "grad_norm": 0.96875,
      "learning_rate": 0.000389471417509662,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2211269572377205,
      "num_tokens": 67113498.0,
      "step": 30060
    },
    {
      "entropy": 5.249836492538452,
      "epoch": 3.3765934744763295,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000389436851362263,
      "loss": 4.9344,
      "mean_token_accuracy": 0.2214808478951454,
      "num_tokens": 67124591.0,
      "step": 30065
    },
    {
      "entropy": 5.293403244018554,
      "epoch": 3.377155051384287,
      "grad_norm": 1.0,
      "learning_rate": 0.00038940228157115036,
      "loss": 4.9401,
      "mean_token_accuracy": 0.21636135280132293,
      "num_tokens": 67135627.0,
      "step": 30070
    },
    {
      "entropy": 5.2452178478240965,
      "epoch": 3.3777166282922444,
      "grad_norm": 1.0,
      "learning_rate": 0.00038936770813742473,
      "loss": 4.8203,
      "mean_token_accuracy": 0.22384341210126876,
      "num_tokens": 67146084.0,
      "step": 30075
    },
    {
      "entropy": 5.266552591323853,
      "epoch": 3.378278205200202,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038933313106218694,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22304999977350234,
      "num_tokens": 67157749.0,
      "step": 30080
    },
    {
      "entropy": 5.31452054977417,
      "epoch": 3.3788397821081597,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038929855034653776,
      "loss": 5.0293,
      "mean_token_accuracy": 0.2139410898089409,
      "num_tokens": 67169386.0,
      "step": 30085
    },
    {
      "entropy": 5.230009746551514,
      "epoch": 3.3794013590161174,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038926396599157855,
      "loss": 4.9209,
      "mean_token_accuracy": 0.2200719892978668,
      "num_tokens": 67181436.0,
      "step": 30090
    },
    {
      "entropy": 5.301683807373047,
      "epoch": 3.379962935924075,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003892293779984103,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22926938831806182,
      "num_tokens": 67192576.0,
      "step": 30095
    },
    {
      "entropy": 5.374930572509766,
      "epoch": 3.3805245128320323,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003891947863681344,
      "loss": 5.0245,
      "mean_token_accuracy": 0.21586984992027283,
      "num_tokens": 67205083.0,
      "step": 30100
    },
    {
      "entropy": 5.123916530609131,
      "epoch": 3.38108608973999,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003891601911018521,
      "loss": 4.7687,
      "mean_token_accuracy": 0.22446683794260025,
      "num_tokens": 67216499.0,
      "step": 30105
    },
    {
      "entropy": 5.167118263244629,
      "epoch": 3.3816476666479476,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003891255922006652,
      "loss": 4.8436,
      "mean_token_accuracy": 0.2286264032125473,
      "num_tokens": 67228390.0,
      "step": 30110
    },
    {
      "entropy": 5.3049239158630375,
      "epoch": 3.382209243555905,
      "grad_norm": 1.03125,
      "learning_rate": 0.000389090989665675,
      "loss": 4.9207,
      "mean_token_accuracy": 0.21699696481227876,
      "num_tokens": 67239655.0,
      "step": 30115
    },
    {
      "entropy": 5.286165237426758,
      "epoch": 3.3827708204638625,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003890563834979835,
      "loss": 4.9636,
      "mean_token_accuracy": 0.21658642441034318,
      "num_tokens": 67251712.0,
      "step": 30120
    },
    {
      "entropy": 5.2873194217681885,
      "epoch": 3.38333239737182,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003890217736986926,
      "loss": 4.8911,
      "mean_token_accuracy": 0.22369159013032913,
      "num_tokens": 67262494.0,
      "step": 30125
    },
    {
      "entropy": 5.255672550201416,
      "epoch": 3.383893974279778,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038898716026890414,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22967892587184907,
      "num_tokens": 67273179.0,
      "step": 30130
    },
    {
      "entropy": 5.254700708389282,
      "epoch": 3.384455551187735,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003889525432097203,
      "loss": 4.9164,
      "mean_token_accuracy": 0.2288793921470642,
      "num_tokens": 67284031.0,
      "step": 30135
    },
    {
      "entropy": 5.366011619567871,
      "epoch": 3.3850171280956927,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003889179225222432,
      "loss": 5.0221,
      "mean_token_accuracy": 0.21286811977624892,
      "num_tokens": 67295728.0,
      "step": 30140
    },
    {
      "entropy": 5.321989297866821,
      "epoch": 3.3855787050036503,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038888329820757536,
      "loss": 4.9926,
      "mean_token_accuracy": 0.21434528827667237,
      "num_tokens": 67306880.0,
      "step": 30145
    },
    {
      "entropy": 5.250279235839844,
      "epoch": 3.386140281911608,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038884867026681906,
      "loss": 4.8643,
      "mean_token_accuracy": 0.22237369269132615,
      "num_tokens": 67317849.0,
      "step": 30150
    },
    {
      "entropy": 5.178459548950196,
      "epoch": 3.386701858819565,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003888140387010771,
      "loss": 4.8972,
      "mean_token_accuracy": 0.22362206876277924,
      "num_tokens": 67329421.0,
      "step": 30155
    },
    {
      "entropy": 5.303807163238526,
      "epoch": 3.387263435727523,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038877940351145186,
      "loss": 4.9555,
      "mean_token_accuracy": 0.223042856156826,
      "num_tokens": 67340072.0,
      "step": 30160
    },
    {
      "entropy": 5.258102750778198,
      "epoch": 3.3878250126354805,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038874476469904633,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22206594198942184,
      "num_tokens": 67351687.0,
      "step": 30165
    },
    {
      "entropy": 5.199702262878418,
      "epoch": 3.3883865895434377,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003887101222649634,
      "loss": 4.885,
      "mean_token_accuracy": 0.2201700523495674,
      "num_tokens": 67362978.0,
      "step": 30170
    },
    {
      "entropy": 5.227716684341431,
      "epoch": 3.3889481664513954,
      "grad_norm": 1.015625,
      "learning_rate": 0.000388675476210306,
      "loss": 4.8591,
      "mean_token_accuracy": 0.23061007559299468,
      "num_tokens": 67374142.0,
      "step": 30175
    },
    {
      "entropy": 5.2939074516296385,
      "epoch": 3.389509743359353,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038864082653617744,
      "loss": 4.9452,
      "mean_token_accuracy": 0.21230644583702088,
      "num_tokens": 67385196.0,
      "step": 30180
    },
    {
      "entropy": 5.234585380554199,
      "epoch": 3.3900713202673107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038860617324368085,
      "loss": 4.8945,
      "mean_token_accuracy": 0.22582553178071976,
      "num_tokens": 67395830.0,
      "step": 30185
    },
    {
      "entropy": 5.254419040679932,
      "epoch": 3.3906328971752684,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003885715163339196,
      "loss": 4.9172,
      "mean_token_accuracy": 0.21947467029094697,
      "num_tokens": 67405950.0,
      "step": 30190
    },
    {
      "entropy": 5.291035938262939,
      "epoch": 3.3911944740832256,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038853685580799726,
      "loss": 4.9016,
      "mean_token_accuracy": 0.22060259729623793,
      "num_tokens": 67417993.0,
      "step": 30195
    },
    {
      "entropy": 5.314374685287476,
      "epoch": 3.3917560509911833,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003885021916670174,
      "loss": 4.8939,
      "mean_token_accuracy": 0.22901211977005004,
      "num_tokens": 67429222.0,
      "step": 30200
    },
    {
      "entropy": 5.31543984413147,
      "epoch": 3.392317627899141,
      "grad_norm": 1.0,
      "learning_rate": 0.0003884675239120838,
      "loss": 5.0006,
      "mean_token_accuracy": 0.22000284492969513,
      "num_tokens": 67439870.0,
      "step": 30205
    },
    {
      "entropy": 5.25907793045044,
      "epoch": 3.392879204807098,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038843285254430005,
      "loss": 4.7877,
      "mean_token_accuracy": 0.2261288955807686,
      "num_tokens": 67450491.0,
      "step": 30210
    },
    {
      "entropy": 5.285482263565063,
      "epoch": 3.393440781715056,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003883981775647705,
      "loss": 4.9473,
      "mean_token_accuracy": 0.2160302773118019,
      "num_tokens": 67461432.0,
      "step": 30215
    },
    {
      "entropy": 5.276615047454834,
      "epoch": 3.3940023586230135,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003883634989745988,
      "loss": 4.9319,
      "mean_token_accuracy": 0.21808574944734574,
      "num_tokens": 67472614.0,
      "step": 30220
    },
    {
      "entropy": 5.289629030227661,
      "epoch": 3.394563935530971,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00038832881677488944,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22212473154067994,
      "num_tokens": 67484383.0,
      "step": 30225
    },
    {
      "entropy": 5.27636685371399,
      "epoch": 3.3951255124389284,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00038829413096674653,
      "loss": 4.8848,
      "mean_token_accuracy": 0.22259222567081452,
      "num_tokens": 67495697.0,
      "step": 30230
    },
    {
      "entropy": 5.22426700592041,
      "epoch": 3.395687089346886,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003882594415512746,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22265804708003997,
      "num_tokens": 67507937.0,
      "step": 30235
    },
    {
      "entropy": 5.247018194198608,
      "epoch": 3.3962486662548437,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038822474852957817,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22492894828319548,
      "num_tokens": 67519649.0,
      "step": 30240
    },
    {
      "entropy": 5.371817493438721,
      "epoch": 3.3968102431628013,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003881900519027618,
      "loss": 5.0111,
      "mean_token_accuracy": 0.21417767852544783,
      "num_tokens": 67532301.0,
      "step": 30245
    },
    {
      "entropy": 5.2986503601074215,
      "epoch": 3.3973718200707586,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003881553516719303,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22091288715600968,
      "num_tokens": 67545106.0,
      "step": 30250
    },
    {
      "entropy": 5.336609745025635,
      "epoch": 3.397933396978716,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038812064783818847,
      "loss": 4.9843,
      "mean_token_accuracy": 0.21174246519804002,
      "num_tokens": 67556188.0,
      "step": 30255
    },
    {
      "entropy": 5.222738790512085,
      "epoch": 3.398494973886674,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038808594040264127,
      "loss": 4.836,
      "mean_token_accuracy": 0.22425906509160995,
      "num_tokens": 67567707.0,
      "step": 30260
    },
    {
      "entropy": 5.222129535675049,
      "epoch": 3.399056550794631,
      "grad_norm": 1.015625,
      "learning_rate": 0.000388051229366394,
      "loss": 4.9034,
      "mean_token_accuracy": 0.21939173340797424,
      "num_tokens": 67578334.0,
      "step": 30265
    },
    {
      "entropy": 5.153694677352905,
      "epoch": 3.3996181277025888,
      "grad_norm": 1.0,
      "learning_rate": 0.00038801651473055167,
      "loss": 4.7277,
      "mean_token_accuracy": 0.227745857834816,
      "num_tokens": 67588896.0,
      "step": 30270
    },
    {
      "entropy": 5.249508857727051,
      "epoch": 3.4001797046105464,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003879817964962197,
      "loss": 4.9205,
      "mean_token_accuracy": 0.21900044232606888,
      "num_tokens": 67599032.0,
      "step": 30275
    },
    {
      "entropy": 5.286703062057495,
      "epoch": 3.400741281518504,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003879470746645035,
      "loss": 4.972,
      "mean_token_accuracy": 0.21264339983463287,
      "num_tokens": 67610095.0,
      "step": 30280
    },
    {
      "entropy": 5.340462923049927,
      "epoch": 3.4013028584264617,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038791234923650866,
      "loss": 5.0204,
      "mean_token_accuracy": 0.21207524538040162,
      "num_tokens": 67622039.0,
      "step": 30285
    },
    {
      "entropy": 5.365471839904785,
      "epoch": 3.401864435334419,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038787762021334077,
      "loss": 4.9951,
      "mean_token_accuracy": 0.21500943452119828,
      "num_tokens": 67634123.0,
      "step": 30290
    },
    {
      "entropy": 5.259051704406739,
      "epoch": 3.4024260122423766,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003878428875961057,
      "loss": 4.9552,
      "mean_token_accuracy": 0.2213268682360649,
      "num_tokens": 67644696.0,
      "step": 30295
    },
    {
      "entropy": 5.2923366069793705,
      "epoch": 3.4029875891503343,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038780815138590935,
      "loss": 5.0092,
      "mean_token_accuracy": 0.20584191083908082,
      "num_tokens": 67656193.0,
      "step": 30300
    },
    {
      "entropy": 5.227161836624146,
      "epoch": 3.4035491660582915,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003877734115838577,
      "loss": 4.8473,
      "mean_token_accuracy": 0.2247167021036148,
      "num_tokens": 67668261.0,
      "step": 30305
    },
    {
      "entropy": 5.29750428199768,
      "epoch": 3.404110742966249,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038773866819105684,
      "loss": 4.9571,
      "mean_token_accuracy": 0.21201450228691102,
      "num_tokens": 67678264.0,
      "step": 30310
    },
    {
      "entropy": 5.298026418685913,
      "epoch": 3.404672319874207,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003877039212086131,
      "loss": 4.9458,
      "mean_token_accuracy": 0.21950660794973373,
      "num_tokens": 67688455.0,
      "step": 30315
    },
    {
      "entropy": 5.2228233337402346,
      "epoch": 3.4052338967821645,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038766917063763287,
      "loss": 4.8065,
      "mean_token_accuracy": 0.2257700264453888,
      "num_tokens": 67700970.0,
      "step": 30320
    },
    {
      "entropy": 5.281900024414062,
      "epoch": 3.4057954736901217,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038763441647922246,
      "loss": 4.9531,
      "mean_token_accuracy": 0.22077552527189254,
      "num_tokens": 67711835.0,
      "step": 30325
    },
    {
      "entropy": 5.202890825271607,
      "epoch": 3.4063570505980794,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003875996587344885,
      "loss": 4.8916,
      "mean_token_accuracy": 0.22529463917016984,
      "num_tokens": 67722915.0,
      "step": 30330
    },
    {
      "entropy": 5.3511780261993405,
      "epoch": 3.406918627506037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038756489740453785,
      "loss": 4.93,
      "mean_token_accuracy": 0.22200679928064346,
      "num_tokens": 67734301.0,
      "step": 30335
    },
    {
      "entropy": 5.303980827331543,
      "epoch": 3.4074802044139947,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038753013249047706,
      "loss": 4.9631,
      "mean_token_accuracy": 0.21852849125862123,
      "num_tokens": 67745763.0,
      "step": 30340
    },
    {
      "entropy": 5.274885892868042,
      "epoch": 3.408041781321952,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003874953639934134,
      "loss": 4.9216,
      "mean_token_accuracy": 0.22400962859392165,
      "num_tokens": 67757241.0,
      "step": 30345
    },
    {
      "entropy": 5.261227035522461,
      "epoch": 3.4086033582299096,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003874605919144536,
      "loss": 4.9388,
      "mean_token_accuracy": 0.220479916036129,
      "num_tokens": 67768231.0,
      "step": 30350
    },
    {
      "entropy": 5.335046291351318,
      "epoch": 3.4091649351378672,
      "grad_norm": 0.890625,
      "learning_rate": 0.0003874258162547049,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21078290790319443,
      "num_tokens": 67780735.0,
      "step": 30355
    },
    {
      "entropy": 5.273346614837647,
      "epoch": 3.4097265120458244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003873910370152747,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22821754962205887,
      "num_tokens": 67791817.0,
      "step": 30360
    },
    {
      "entropy": 5.286165571212768,
      "epoch": 3.410288088953782,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003873562541972702,
      "loss": 4.9203,
      "mean_token_accuracy": 0.22696709185838698,
      "num_tokens": 67803591.0,
      "step": 30365
    },
    {
      "entropy": 5.287691497802735,
      "epoch": 3.4108496658617398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003873214678017991,
      "loss": 4.932,
      "mean_token_accuracy": 0.22097309678792953,
      "num_tokens": 67813250.0,
      "step": 30370
    },
    {
      "entropy": 5.287344026565552,
      "epoch": 3.4114112427696974,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003872866778299688,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22184806019067765,
      "num_tokens": 67825042.0,
      "step": 30375
    },
    {
      "entropy": 5.310749959945679,
      "epoch": 3.411972819677655,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003872518842828872,
      "loss": 4.9882,
      "mean_token_accuracy": 0.2128303974866867,
      "num_tokens": 67837535.0,
      "step": 30380
    },
    {
      "entropy": 5.297269582748413,
      "epoch": 3.4125343965856123,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038721708716166203,
      "loss": 5.0091,
      "mean_token_accuracy": 0.22286460250616075,
      "num_tokens": 67848950.0,
      "step": 30385
    },
    {
      "entropy": 5.285321426391602,
      "epoch": 3.41309597349357,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003871822864674013,
      "loss": 4.9637,
      "mean_token_accuracy": 0.21650200039148332,
      "num_tokens": 67861488.0,
      "step": 30390
    },
    {
      "entropy": 5.3389839172363285,
      "epoch": 3.4136575504015276,
      "grad_norm": 1.0,
      "learning_rate": 0.00038714748220121305,
      "loss": 5.0336,
      "mean_token_accuracy": 0.22078835517168044,
      "num_tokens": 67872627.0,
      "step": 30395
    },
    {
      "entropy": 5.273837900161743,
      "epoch": 3.414219127309485,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003871126743642055,
      "loss": 4.94,
      "mean_token_accuracy": 0.21725678443908691,
      "num_tokens": 67884033.0,
      "step": 30400
    },
    {
      "entropy": 5.267806720733643,
      "epoch": 3.4147807042174425,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038707786295748693,
      "loss": 4.9217,
      "mean_token_accuracy": 0.217074479162693,
      "num_tokens": 67895696.0,
      "step": 30405
    },
    {
      "entropy": 5.333352756500244,
      "epoch": 3.4153422811254,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038704304798216566,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2258339360356331,
      "num_tokens": 67906328.0,
      "step": 30410
    },
    {
      "entropy": 5.234513187408448,
      "epoch": 3.415903858033358,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003870082294393503,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23282802551984788,
      "num_tokens": 67917258.0,
      "step": 30415
    },
    {
      "entropy": 5.219203615188599,
      "epoch": 3.416465434941315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003869734073301495,
      "loss": 4.8721,
      "mean_token_accuracy": 0.22113756686449051,
      "num_tokens": 67927431.0,
      "step": 30420
    },
    {
      "entropy": 5.298858499526977,
      "epoch": 3.4170270118492727,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038693858165567195,
      "loss": 4.9398,
      "mean_token_accuracy": 0.22034342288970948,
      "num_tokens": 67938801.0,
      "step": 30425
    },
    {
      "entropy": 5.2393700122833256,
      "epoch": 3.4175885887572304,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003869037524170266,
      "loss": 4.8224,
      "mean_token_accuracy": 0.22581790387630463,
      "num_tokens": 67949690.0,
      "step": 30430
    },
    {
      "entropy": 5.215966796875,
      "epoch": 3.418150165665188,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038686891961532227,
      "loss": 4.8771,
      "mean_token_accuracy": 0.2232057958841324,
      "num_tokens": 67959542.0,
      "step": 30435
    },
    {
      "entropy": 5.198278474807739,
      "epoch": 3.4187117425731453,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003868340832516683,
      "loss": 4.7843,
      "mean_token_accuracy": 0.22764142006635665,
      "num_tokens": 67969759.0,
      "step": 30440
    },
    {
      "entropy": 5.195681142807007,
      "epoch": 3.419273319481103,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038679924332717363,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2221192568540573,
      "num_tokens": 67982302.0,
      "step": 30445
    },
    {
      "entropy": 5.287873268127441,
      "epoch": 3.4198348963890606,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038676439984294776,
      "loss": 4.9837,
      "mean_token_accuracy": 0.21788906902074814,
      "num_tokens": 67994452.0,
      "step": 30450
    },
    {
      "entropy": 5.34516396522522,
      "epoch": 3.420396473297018,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038672955280009996,
      "loss": 4.9684,
      "mean_token_accuracy": 0.21715096682310103,
      "num_tokens": 68006445.0,
      "step": 30455
    },
    {
      "entropy": 5.311071729660034,
      "epoch": 3.4209580502049755,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038669470219973996,
      "loss": 4.9214,
      "mean_token_accuracy": 0.21794755905866622,
      "num_tokens": 68017720.0,
      "step": 30460
    },
    {
      "entropy": 5.233233404159546,
      "epoch": 3.421519627112933,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038665984804297734,
      "loss": 4.8625,
      "mean_token_accuracy": 0.2224726378917694,
      "num_tokens": 68028199.0,
      "step": 30465
    },
    {
      "entropy": 5.261049079895019,
      "epoch": 3.4220812040208908,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003866249903309218,
      "loss": 4.9561,
      "mean_token_accuracy": 0.21542255878448485,
      "num_tokens": 68040171.0,
      "step": 30470
    },
    {
      "entropy": 5.336052656173706,
      "epoch": 3.4226427809288484,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003865901290646833,
      "loss": 4.9433,
      "mean_token_accuracy": 0.21410854309797286,
      "num_tokens": 68050484.0,
      "step": 30475
    },
    {
      "entropy": 5.33268027305603,
      "epoch": 3.4232043578368057,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038655526424537175,
      "loss": 4.965,
      "mean_token_accuracy": 0.21743730157613755,
      "num_tokens": 68062110.0,
      "step": 30480
    },
    {
      "entropy": 5.327915954589844,
      "epoch": 3.4237659347447633,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038652039587409734,
      "loss": 4.9869,
      "mean_token_accuracy": 0.21907012313604354,
      "num_tokens": 68074890.0,
      "step": 30485
    },
    {
      "entropy": 5.2610689163208,
      "epoch": 3.424327511652721,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003864855239519703,
      "loss": 4.9126,
      "mean_token_accuracy": 0.21789348423480986,
      "num_tokens": 68084909.0,
      "step": 30490
    },
    {
      "entropy": 5.163326978683472,
      "epoch": 3.424889088560678,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00038645064848010093,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22768950462341309,
      "num_tokens": 68097114.0,
      "step": 30495
    },
    {
      "entropy": 5.309282350540161,
      "epoch": 3.425450665468636,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038641576945959964,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22821509540081025,
      "num_tokens": 68108377.0,
      "step": 30500
    },
    {
      "entropy": 5.230483055114746,
      "epoch": 3.4260122423765935,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038638088689157713,
      "loss": 4.9236,
      "mean_token_accuracy": 0.22181156426668167,
      "num_tokens": 68118859.0,
      "step": 30505
    },
    {
      "entropy": 5.239348649978638,
      "epoch": 3.426573819284551,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038634600077714395,
      "loss": 4.8538,
      "mean_token_accuracy": 0.22891925573348998,
      "num_tokens": 68130151.0,
      "step": 30510
    },
    {
      "entropy": 5.254832601547241,
      "epoch": 3.4271353961925084,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003863111111174109,
      "loss": 4.8738,
      "mean_token_accuracy": 0.2192390128970146,
      "num_tokens": 68142514.0,
      "step": 30515
    },
    {
      "entropy": 5.293326377868652,
      "epoch": 3.427696973100466,
      "grad_norm": 1.0625,
      "learning_rate": 0.00038627621791348886,
      "loss": 5.007,
      "mean_token_accuracy": 0.2155739352107048,
      "num_tokens": 68153751.0,
      "step": 30520
    },
    {
      "entropy": 5.304003572463989,
      "epoch": 3.4282585500084237,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038624132116648897,
      "loss": 4.8947,
      "mean_token_accuracy": 0.22654018253087999,
      "num_tokens": 68164885.0,
      "step": 30525
    },
    {
      "entropy": 5.234183120727539,
      "epoch": 3.4288201269163814,
      "grad_norm": 1.0,
      "learning_rate": 0.0003862064208775223,
      "loss": 4.8639,
      "mean_token_accuracy": 0.22551942616701126,
      "num_tokens": 68176320.0,
      "step": 30530
    },
    {
      "entropy": 5.27961277961731,
      "epoch": 3.4293817038243386,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038617151704769994,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22424793541431426,
      "num_tokens": 68186665.0,
      "step": 30535
    },
    {
      "entropy": 5.29964861869812,
      "epoch": 3.4299432807322963,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003861366096781334,
      "loss": 5.0027,
      "mean_token_accuracy": 0.20840065628290178,
      "num_tokens": 68198011.0,
      "step": 30540
    },
    {
      "entropy": 5.380283641815185,
      "epoch": 3.430504857640254,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003861016987699342,
      "loss": 5.0345,
      "mean_token_accuracy": 0.2178317442536354,
      "num_tokens": 68209659.0,
      "step": 30545
    },
    {
      "entropy": 5.279834413528443,
      "epoch": 3.431066434548211,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038606678432421366,
      "loss": 4.8496,
      "mean_token_accuracy": 0.2279307022690773,
      "num_tokens": 68220424.0,
      "step": 30550
    },
    {
      "entropy": 5.206638765335083,
      "epoch": 3.431628011456169,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003860318663420837,
      "loss": 4.8869,
      "mean_token_accuracy": 0.21757474839687346,
      "num_tokens": 68231413.0,
      "step": 30555
    },
    {
      "entropy": 5.252436542510987,
      "epoch": 3.4321895883641265,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038599694482465606,
      "loss": 4.8455,
      "mean_token_accuracy": 0.2297004133462906,
      "num_tokens": 68242435.0,
      "step": 30560
    },
    {
      "entropy": 5.294650936126709,
      "epoch": 3.432751165272084,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038596201977304263,
      "loss": 4.8802,
      "mean_token_accuracy": 0.23015250861644745,
      "num_tokens": 68254462.0,
      "step": 30565
    },
    {
      "entropy": 5.266968774795532,
      "epoch": 3.433312742180042,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038592709118835547,
      "loss": 4.94,
      "mean_token_accuracy": 0.21961122751235962,
      "num_tokens": 68265842.0,
      "step": 30570
    },
    {
      "entropy": 5.242256832122803,
      "epoch": 3.433874319087999,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038589215907170664,
      "loss": 4.9059,
      "mean_token_accuracy": 0.22130930423736572,
      "num_tokens": 68277117.0,
      "step": 30575
    },
    {
      "entropy": 5.349419832229614,
      "epoch": 3.4344358959959567,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003858572234242086,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21869643777608871,
      "num_tokens": 68287869.0,
      "step": 30580
    },
    {
      "entropy": 5.340730667114258,
      "epoch": 3.4349974729039143,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038582228424697347,
      "loss": 4.8983,
      "mean_token_accuracy": 0.21893491446971894,
      "num_tokens": 68300711.0,
      "step": 30585
    },
    {
      "entropy": 5.284315681457519,
      "epoch": 3.4355590498118715,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003857873415411138,
      "loss": 5.0068,
      "mean_token_accuracy": 0.21852849274873734,
      "num_tokens": 68311997.0,
      "step": 30590
    },
    {
      "entropy": 5.227691984176635,
      "epoch": 3.436120626719829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003857523953077423,
      "loss": 4.8658,
      "mean_token_accuracy": 0.22492787539958953,
      "num_tokens": 68322129.0,
      "step": 30595
    },
    {
      "entropy": 5.2269329071044925,
      "epoch": 3.436682203627787,
      "grad_norm": 1.0,
      "learning_rate": 0.00038571744554797156,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22569518238306047,
      "num_tokens": 68334202.0,
      "step": 30600
    },
    {
      "entropy": 5.234831857681274,
      "epoch": 3.4372437805357445,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038568249226291433,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22549338340759278,
      "num_tokens": 68345370.0,
      "step": 30605
    },
    {
      "entropy": 5.216380023956299,
      "epoch": 3.4378053574437017,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038564753545368375,
      "loss": 4.7781,
      "mean_token_accuracy": 0.22635187804698945,
      "num_tokens": 68355703.0,
      "step": 30610
    },
    {
      "entropy": 5.2230690002441404,
      "epoch": 3.4383669343516594,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038561257512139264,
      "loss": 4.9516,
      "mean_token_accuracy": 0.21225782930850984,
      "num_tokens": 68367412.0,
      "step": 30615
    },
    {
      "entropy": 5.27530574798584,
      "epoch": 3.438928511259617,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038557761126715426,
      "loss": 4.8791,
      "mean_token_accuracy": 0.22623851895332336,
      "num_tokens": 68378074.0,
      "step": 30620
    },
    {
      "entropy": 5.328730344772339,
      "epoch": 3.4394900881675747,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003855426438920818,
      "loss": 4.957,
      "mean_token_accuracy": 0.2148025795817375,
      "num_tokens": 68388730.0,
      "step": 30625
    },
    {
      "entropy": 5.242838382720947,
      "epoch": 3.440051665075532,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038550767299728875,
      "loss": 4.8336,
      "mean_token_accuracy": 0.2240378737449646,
      "num_tokens": 68399172.0,
      "step": 30630
    },
    {
      "entropy": 5.257505130767822,
      "epoch": 3.4406132419834896,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038547269858388853,
      "loss": 4.962,
      "mean_token_accuracy": 0.21906629353761672,
      "num_tokens": 68411704.0,
      "step": 30635
    },
    {
      "entropy": 5.243538522720337,
      "epoch": 3.4411748188914473,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003854377206529947,
      "loss": 4.9088,
      "mean_token_accuracy": 0.21812504082918166,
      "num_tokens": 68422774.0,
      "step": 30640
    },
    {
      "entropy": 5.243760585784912,
      "epoch": 3.4417363957994045,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000385402739205721,
      "loss": 4.8014,
      "mean_token_accuracy": 0.23051120787858964,
      "num_tokens": 68432964.0,
      "step": 30645
    },
    {
      "entropy": 5.234975051879883,
      "epoch": 3.442297972707362,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003853677542431813,
      "loss": 4.9114,
      "mean_token_accuracy": 0.21539413928985596,
      "num_tokens": 68445875.0,
      "step": 30650
    },
    {
      "entropy": 5.332870292663574,
      "epoch": 3.44285954961532,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003853327657664895,
      "loss": 4.9798,
      "mean_token_accuracy": 0.21340992897748948,
      "num_tokens": 68457972.0,
      "step": 30655
    },
    {
      "entropy": 5.246852445602417,
      "epoch": 3.4434211265232775,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003852977737767596,
      "loss": 4.8291,
      "mean_token_accuracy": 0.22457252591848373,
      "num_tokens": 68468825.0,
      "step": 30660
    },
    {
      "entropy": 5.184511423110962,
      "epoch": 3.443982703431235,
      "grad_norm": 1.125,
      "learning_rate": 0.0003852627782751059,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22972948104143143,
      "num_tokens": 68479427.0,
      "step": 30665
    },
    {
      "entropy": 5.15514554977417,
      "epoch": 3.4445442803391924,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003852277792626425,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22444251477718352,
      "num_tokens": 68490542.0,
      "step": 30670
    },
    {
      "entropy": 5.216920566558838,
      "epoch": 3.44510585724715,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003851927767404839,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2259128138422966,
      "num_tokens": 68501396.0,
      "step": 30675
    },
    {
      "entropy": 5.348768806457519,
      "epoch": 3.4456674341551077,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003851577707097445,
      "loss": 5.056,
      "mean_token_accuracy": 0.20822470039129257,
      "num_tokens": 68512710.0,
      "step": 30680
    },
    {
      "entropy": 5.27172703742981,
      "epoch": 3.446229011063065,
      "grad_norm": 0.953125,
      "learning_rate": 0.000385122761171539,
      "loss": 4.9109,
      "mean_token_accuracy": 0.21307166963815688,
      "num_tokens": 68524289.0,
      "step": 30685
    },
    {
      "entropy": 5.251991844177246,
      "epoch": 3.4467905879710226,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00038508774812698207,
      "loss": 4.8823,
      "mean_token_accuracy": 0.2180471420288086,
      "num_tokens": 68536783.0,
      "step": 30690
    },
    {
      "entropy": 5.2672872066497805,
      "epoch": 3.44735216487898,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003850527315771885,
      "loss": 4.9506,
      "mean_token_accuracy": 0.22064568847417831,
      "num_tokens": 68548113.0,
      "step": 30695
    },
    {
      "entropy": 5.222851085662842,
      "epoch": 3.447913741786938,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038501771152327333,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22539587765932084,
      "num_tokens": 68560234.0,
      "step": 30700
    },
    {
      "entropy": 5.239931058883667,
      "epoch": 3.448475318694895,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038498268796635155,
      "loss": 4.8744,
      "mean_token_accuracy": 0.22143177986145018,
      "num_tokens": 68571712.0,
      "step": 30705
    },
    {
      "entropy": 5.206453561782837,
      "epoch": 3.4490368956028528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038494766090753824,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22622640579938888,
      "num_tokens": 68583166.0,
      "step": 30710
    },
    {
      "entropy": 5.2089227676391605,
      "epoch": 3.4495984725108104,
      "grad_norm": 0.9375,
      "learning_rate": 0.00038491263034794886,
      "loss": 4.7743,
      "mean_token_accuracy": 0.23521511852741242,
      "num_tokens": 68594026.0,
      "step": 30715
    },
    {
      "entropy": 5.156444358825683,
      "epoch": 3.450160049418768,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003848775962886987,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22301087230443956,
      "num_tokens": 68604046.0,
      "step": 30720
    },
    {
      "entropy": 5.17240080833435,
      "epoch": 3.4507216263267253,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003848425587309032,
      "loss": 4.8373,
      "mean_token_accuracy": 0.22955212593078614,
      "num_tokens": 68615575.0,
      "step": 30725
    },
    {
      "entropy": 5.3071332454681395,
      "epoch": 3.451283203234683,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038480751767567807,
      "loss": 4.9791,
      "mean_token_accuracy": 0.2183932662010193,
      "num_tokens": 68626805.0,
      "step": 30730
    },
    {
      "entropy": 5.41607666015625,
      "epoch": 3.4518447801426406,
      "grad_norm": 0.91015625,
      "learning_rate": 0.000384772473124139,
      "loss": 4.9955,
      "mean_token_accuracy": 0.2150215893983841,
      "num_tokens": 68638802.0,
      "step": 30735
    },
    {
      "entropy": 5.294073629379272,
      "epoch": 3.452406357050598,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003847374250774018,
      "loss": 4.9865,
      "mean_token_accuracy": 0.2145303875207901,
      "num_tokens": 68651137.0,
      "step": 30740
    },
    {
      "entropy": 5.213368844985962,
      "epoch": 3.4529679339585555,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038470237353658237,
      "loss": 4.8397,
      "mean_token_accuracy": 0.22877264469861985,
      "num_tokens": 68661268.0,
      "step": 30745
    },
    {
      "entropy": 5.134306001663208,
      "epoch": 3.453529510866513,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038466731850279683,
      "loss": 4.8131,
      "mean_token_accuracy": 0.2231605812907219,
      "num_tokens": 68672334.0,
      "step": 30750
    },
    {
      "entropy": 5.257324838638306,
      "epoch": 3.454091087774471,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038463225997716137,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2253659948706627,
      "num_tokens": 68683765.0,
      "step": 30755
    },
    {
      "entropy": 5.4528357028961185,
      "epoch": 3.4546526646824285,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00038459719796079227,
      "loss": 5.0335,
      "mean_token_accuracy": 0.21109310835599898,
      "num_tokens": 68694969.0,
      "step": 30760
    },
    {
      "entropy": 5.327714824676514,
      "epoch": 3.4552142415903857,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003845621324548059,
      "loss": 5.0098,
      "mean_token_accuracy": 0.22120005935430526,
      "num_tokens": 68708586.0,
      "step": 30765
    },
    {
      "entropy": 5.295125675201416,
      "epoch": 3.4557758184983434,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003845270634603186,
      "loss": 4.944,
      "mean_token_accuracy": 0.2166455551981926,
      "num_tokens": 68719649.0,
      "step": 30770
    },
    {
      "entropy": 5.292798233032227,
      "epoch": 3.456337395406301,
      "grad_norm": 1.09375,
      "learning_rate": 0.00038449199097844724,
      "loss": 4.9422,
      "mean_token_accuracy": 0.21653105765581132,
      "num_tokens": 68729491.0,
      "step": 30775
    },
    {
      "entropy": 5.25622787475586,
      "epoch": 3.4568989723142582,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003844569150103084,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2244524911046028,
      "num_tokens": 68741087.0,
      "step": 30780
    },
    {
      "entropy": 5.22101321220398,
      "epoch": 3.457460549222216,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003844218355570189,
      "loss": 4.8548,
      "mean_token_accuracy": 0.2268972247838974,
      "num_tokens": 68752699.0,
      "step": 30785
    },
    {
      "entropy": 5.271330308914185,
      "epoch": 3.4580221261301736,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003843867526196957,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22223399430513383,
      "num_tokens": 68763066.0,
      "step": 30790
    },
    {
      "entropy": 5.238379955291748,
      "epoch": 3.4585837030381312,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003843516661994559,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22283164262771607,
      "num_tokens": 68773878.0,
      "step": 30795
    },
    {
      "entropy": 5.18544487953186,
      "epoch": 3.4591452799460884,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003843165762974167,
      "loss": 4.9024,
      "mean_token_accuracy": 0.22227375209331512,
      "num_tokens": 68783864.0,
      "step": 30800
    },
    {
      "entropy": 5.227670288085937,
      "epoch": 3.459706856854046,
      "grad_norm": 1.0,
      "learning_rate": 0.00038428148291469515,
      "loss": 4.8516,
      "mean_token_accuracy": 0.22803699672222139,
      "num_tokens": 68794280.0,
      "step": 30805
    },
    {
      "entropy": 5.302725839614868,
      "epoch": 3.4602684337620038,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000384246386052409,
      "loss": 5.0297,
      "mean_token_accuracy": 0.21949234306812287,
      "num_tokens": 68806125.0,
      "step": 30810
    },
    {
      "entropy": 5.289509963989258,
      "epoch": 3.4608300106699614,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003842112857116754,
      "loss": 4.8718,
      "mean_token_accuracy": 0.22515960037708282,
      "num_tokens": 68816742.0,
      "step": 30815
    },
    {
      "entropy": 5.213838672637939,
      "epoch": 3.4613915875779186,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003841761818936122,
      "loss": 4.8998,
      "mean_token_accuracy": 0.22193594127893448,
      "num_tokens": 68828504.0,
      "step": 30820
    },
    {
      "entropy": 5.3366272926330565,
      "epoch": 3.4619531644858763,
      "grad_norm": 1.03125,
      "learning_rate": 0.000384141074599337,
      "loss": 4.981,
      "mean_token_accuracy": 0.2153831899166107,
      "num_tokens": 68839993.0,
      "step": 30825
    },
    {
      "entropy": 5.185944843292236,
      "epoch": 3.462514741393834,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00038410596382996764,
      "loss": 4.8033,
      "mean_token_accuracy": 0.22958105355501174,
      "num_tokens": 68850395.0,
      "step": 30830
    },
    {
      "entropy": 5.226319265365601,
      "epoch": 3.463076318301791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003840708495866221,
      "loss": 4.7974,
      "mean_token_accuracy": 0.2273060694336891,
      "num_tokens": 68860621.0,
      "step": 30835
    },
    {
      "entropy": 5.2510405540466305,
      "epoch": 3.463637895209749,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003840357318704183,
      "loss": 4.9563,
      "mean_token_accuracy": 0.21692779362201692,
      "num_tokens": 68872838.0,
      "step": 30840
    },
    {
      "entropy": 5.280448293685913,
      "epoch": 3.4641994721177065,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003840006106824746,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22414372563362123,
      "num_tokens": 68883372.0,
      "step": 30845
    },
    {
      "entropy": 5.277575731277466,
      "epoch": 3.464761049025664,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003839654860239091,
      "loss": 4.8731,
      "mean_token_accuracy": 0.22333725541830063,
      "num_tokens": 68894148.0,
      "step": 30850
    },
    {
      "entropy": 5.286434698104858,
      "epoch": 3.465322625933622,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003839303578958404,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22449158430099486,
      "num_tokens": 68904381.0,
      "step": 30855
    },
    {
      "entropy": 5.299822616577148,
      "epoch": 3.465884202841579,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038389522629938676,
      "loss": 5.018,
      "mean_token_accuracy": 0.2088698536157608,
      "num_tokens": 68916354.0,
      "step": 30860
    },
    {
      "entropy": 5.242871809005737,
      "epoch": 3.4664457797495367,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003838600912356669,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22088201344013214,
      "num_tokens": 68927551.0,
      "step": 30865
    },
    {
      "entropy": 5.3193224430084225,
      "epoch": 3.4670073566574944,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003838249527057995,
      "loss": 4.9733,
      "mean_token_accuracy": 0.22038194984197618,
      "num_tokens": 68939528.0,
      "step": 30870
    },
    {
      "entropy": 5.251523494720459,
      "epoch": 3.4675689335654516,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00038378981071090346,
      "loss": 4.844,
      "mean_token_accuracy": 0.223726586997509,
      "num_tokens": 68951634.0,
      "step": 30875
    },
    {
      "entropy": 5.295159816741943,
      "epoch": 3.4681305104734093,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038375466525209764,
      "loss": 4.8616,
      "mean_token_accuracy": 0.2321859672665596,
      "num_tokens": 68962259.0,
      "step": 30880
    },
    {
      "entropy": 5.235637903213501,
      "epoch": 3.468692087381367,
      "grad_norm": 1.0,
      "learning_rate": 0.00038371951633050106,
      "loss": 4.9672,
      "mean_token_accuracy": 0.22139922976493837,
      "num_tokens": 68973185.0,
      "step": 30885
    },
    {
      "entropy": 5.2722498893737795,
      "epoch": 3.4692536642893246,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038368436394723294,
      "loss": 4.9488,
      "mean_token_accuracy": 0.2148461163043976,
      "num_tokens": 68984052.0,
      "step": 30890
    },
    {
      "entropy": 5.273041439056397,
      "epoch": 3.469815241197282,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038364920810341255,
      "loss": 4.9669,
      "mean_token_accuracy": 0.2210913524031639,
      "num_tokens": 68993730.0,
      "step": 30895
    },
    {
      "entropy": 5.358463382720947,
      "epoch": 3.4703768181052395,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003836140488001592,
      "loss": 4.9377,
      "mean_token_accuracy": 0.21546520292758942,
      "num_tokens": 69006229.0,
      "step": 30900
    },
    {
      "entropy": 5.272105503082275,
      "epoch": 3.470938395013197,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00038357888603859236,
      "loss": 4.9433,
      "mean_token_accuracy": 0.2221418783068657,
      "num_tokens": 69018518.0,
      "step": 30905
    },
    {
      "entropy": 5.3151304721832275,
      "epoch": 3.471499971921155,
      "grad_norm": 0.83203125,
      "learning_rate": 0.00038354371981983175,
      "loss": 4.9663,
      "mean_token_accuracy": 0.20871005654335023,
      "num_tokens": 69031113.0,
      "step": 30910
    },
    {
      "entropy": 5.369360303878784,
      "epoch": 3.472061548829112,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003835085501449969,
      "loss": 5.0633,
      "mean_token_accuracy": 0.21228526085615157,
      "num_tokens": 69042038.0,
      "step": 30915
    },
    {
      "entropy": 5.182710742950439,
      "epoch": 3.4726231257370697,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038347337701520785,
      "loss": 4.8277,
      "mean_token_accuracy": 0.23522704243659973,
      "num_tokens": 69053675.0,
      "step": 30920
    },
    {
      "entropy": 5.250348138809204,
      "epoch": 3.4731847026450273,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00038343820043158436,
      "loss": 4.96,
      "mean_token_accuracy": 0.21672401279211045,
      "num_tokens": 69064475.0,
      "step": 30925
    },
    {
      "entropy": 5.260398244857788,
      "epoch": 3.4737462795529845,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003834030203952465,
      "loss": 4.9103,
      "mean_token_accuracy": 0.22183620184659958,
      "num_tokens": 69076069.0,
      "step": 30930
    },
    {
      "entropy": 5.341762351989746,
      "epoch": 3.474307856460942,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003833678369073144,
      "loss": 5.0084,
      "mean_token_accuracy": 0.2150321289896965,
      "num_tokens": 69087144.0,
      "step": 30935
    },
    {
      "entropy": 5.253537034988403,
      "epoch": 3.4748694333689,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003833326499689084,
      "loss": 4.8519,
      "mean_token_accuracy": 0.2196589455008507,
      "num_tokens": 69098035.0,
      "step": 30940
    },
    {
      "entropy": 5.243789911270142,
      "epoch": 3.4754310102768575,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003832974595811488,
      "loss": 4.9543,
      "mean_token_accuracy": 0.22130059599876403,
      "num_tokens": 69111722.0,
      "step": 30945
    },
    {
      "entropy": 5.276240015029908,
      "epoch": 3.475992587184815,
      "grad_norm": 0.984375,
      "learning_rate": 0.000383262265745156,
      "loss": 4.8602,
      "mean_token_accuracy": 0.22474250346422195,
      "num_tokens": 69122228.0,
      "step": 30950
    },
    {
      "entropy": 5.267228889465332,
      "epoch": 3.4765541640927724,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038322706846205073,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22213305681943893,
      "num_tokens": 69134150.0,
      "step": 30955
    },
    {
      "entropy": 5.1890647411346436,
      "epoch": 3.47711574100073,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00038319186773295364,
      "loss": 4.9381,
      "mean_token_accuracy": 0.21547018885612487,
      "num_tokens": 69145757.0,
      "step": 30960
    },
    {
      "entropy": 5.329764366149902,
      "epoch": 3.4776773179086877,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003831566635589856,
      "loss": 5.024,
      "mean_token_accuracy": 0.21998834758996963,
      "num_tokens": 69157545.0,
      "step": 30965
    },
    {
      "entropy": 5.301944351196289,
      "epoch": 3.478238894816645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003831214559412674,
      "loss": 4.8836,
      "mean_token_accuracy": 0.22264661490917206,
      "num_tokens": 69169070.0,
      "step": 30970
    },
    {
      "entropy": 5.2532586574554445,
      "epoch": 3.4788004717246026,
      "grad_norm": 1.0,
      "learning_rate": 0.00038308624488092005,
      "loss": 4.8742,
      "mean_token_accuracy": 0.22752752900123596,
      "num_tokens": 69178903.0,
      "step": 30975
    },
    {
      "entropy": 5.2002252578735355,
      "epoch": 3.4793620486325603,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003830510303790647,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22416468113660812,
      "num_tokens": 69190489.0,
      "step": 30980
    },
    {
      "entropy": 5.224518775939941,
      "epoch": 3.479923625540518,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038301581243682263,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22175789773464202,
      "num_tokens": 69200574.0,
      "step": 30985
    },
    {
      "entropy": 5.243176651000977,
      "epoch": 3.480485202448475,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003829805910553153,
      "loss": 4.8252,
      "mean_token_accuracy": 0.22703802287578584,
      "num_tokens": 69210597.0,
      "step": 30990
    },
    {
      "entropy": 5.2870738983154295,
      "epoch": 3.481046779356433,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00038294536623566407,
      "loss": 4.9459,
      "mean_token_accuracy": 0.21698232740163803,
      "num_tokens": 69221392.0,
      "step": 30995
    },
    {
      "entropy": 5.1246870994567875,
      "epoch": 3.4816083562643905,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003829101379789905,
      "loss": 4.83,
      "mean_token_accuracy": 0.22183130234479903,
      "num_tokens": 69232825.0,
      "step": 31000
    },
    {
      "entropy": 5.215247058868409,
      "epoch": 3.482169933172348,
      "grad_norm": 1.0,
      "learning_rate": 0.0003828749062864162,
      "loss": 4.9043,
      "mean_token_accuracy": 0.22556720972061156,
      "num_tokens": 69243989.0,
      "step": 31005
    },
    {
      "entropy": 5.285440969467163,
      "epoch": 3.4827315100803053,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003828396711590631,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21542710512876512,
      "num_tokens": 69255828.0,
      "step": 31010
    },
    {
      "entropy": 5.276329517364502,
      "epoch": 3.483293086988263,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038280443259805307,
      "loss": 4.9328,
      "mean_token_accuracy": 0.21787946075201034,
      "num_tokens": 69266496.0,
      "step": 31015
    },
    {
      "entropy": 5.314446544647216,
      "epoch": 3.4838546638962207,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003827691906045081,
      "loss": 4.9816,
      "mean_token_accuracy": 0.2119390308856964,
      "num_tokens": 69277884.0,
      "step": 31020
    },
    {
      "entropy": 5.337122535705566,
      "epoch": 3.484416240804178,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003827339451795503,
      "loss": 5.0621,
      "mean_token_accuracy": 0.21284154802560806,
      "num_tokens": 69290660.0,
      "step": 31025
    },
    {
      "entropy": 5.3916980743408205,
      "epoch": 3.4849778177121356,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003826986963243018,
      "loss": 5.0245,
      "mean_token_accuracy": 0.23072112500667571,
      "num_tokens": 69302310.0,
      "step": 31030
    },
    {
      "entropy": 5.299539089202881,
      "epoch": 3.485539394620093,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003826634440398852,
      "loss": 4.9595,
      "mean_token_accuracy": 0.22101353406906127,
      "num_tokens": 69313804.0,
      "step": 31035
    },
    {
      "entropy": 5.248970556259155,
      "epoch": 3.486100971528051,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038262818832742276,
      "loss": 4.8893,
      "mean_token_accuracy": 0.22227118015289307,
      "num_tokens": 69325372.0,
      "step": 31040
    },
    {
      "entropy": 5.250659084320068,
      "epoch": 3.4866625484360085,
      "grad_norm": 1.0,
      "learning_rate": 0.00038259292918803696,
      "loss": 4.8837,
      "mean_token_accuracy": 0.21443521827459336,
      "num_tokens": 69335835.0,
      "step": 31045
    },
    {
      "entropy": 5.271352910995484,
      "epoch": 3.4872241253439658,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003825576666228506,
      "loss": 4.8954,
      "mean_token_accuracy": 0.22067563384771346,
      "num_tokens": 69346895.0,
      "step": 31050
    },
    {
      "entropy": 5.222805547714233,
      "epoch": 3.4877857022519234,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003825224006329864,
      "loss": 4.9127,
      "mean_token_accuracy": 0.21952128112316133,
      "num_tokens": 69357649.0,
      "step": 31055
    },
    {
      "entropy": 5.2394115924835205,
      "epoch": 3.488347279159881,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003824871312195673,
      "loss": 4.907,
      "mean_token_accuracy": 0.21940683275461198,
      "num_tokens": 69367973.0,
      "step": 31060
    },
    {
      "entropy": 5.298606300354004,
      "epoch": 3.4889088560678383,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003824518583837163,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22575387954711915,
      "num_tokens": 69379411.0,
      "step": 31065
    },
    {
      "entropy": 5.319088792800903,
      "epoch": 3.489470432975796,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003824165821265564,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21664888858795167,
      "num_tokens": 69391261.0,
      "step": 31070
    },
    {
      "entropy": 5.187022733688354,
      "epoch": 3.4900320098837536,
      "grad_norm": 1.0,
      "learning_rate": 0.0003823813024492108,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22849617302417755,
      "num_tokens": 69401712.0,
      "step": 31075
    },
    {
      "entropy": 5.169594478607178,
      "epoch": 3.4905935867917113,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000382346019352803,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2351055845618248,
      "num_tokens": 69412612.0,
      "step": 31080
    },
    {
      "entropy": 5.338055801391602,
      "epoch": 3.4911551636996685,
      "grad_norm": 1.0,
      "learning_rate": 0.0003823107328384562,
      "loss": 4.952,
      "mean_token_accuracy": 0.22397396713495255,
      "num_tokens": 69424042.0,
      "step": 31085
    },
    {
      "entropy": 5.269130420684815,
      "epoch": 3.491716740607626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003822754429072942,
      "loss": 4.9182,
      "mean_token_accuracy": 0.21799152940511704,
      "num_tokens": 69434667.0,
      "step": 31090
    },
    {
      "entropy": 5.320447969436645,
      "epoch": 3.492278317515584,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003822401495604403,
      "loss": 5.0065,
      "mean_token_accuracy": 0.22028126418590546,
      "num_tokens": 69447015.0,
      "step": 31095
    },
    {
      "entropy": 5.235875368118286,
      "epoch": 3.4928398944235415,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00038220485279901855,
      "loss": 4.9086,
      "mean_token_accuracy": 0.21935324668884276,
      "num_tokens": 69457585.0,
      "step": 31100
    },
    {
      "entropy": 5.281552124023437,
      "epoch": 3.4934014713314987,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038216955262415273,
      "loss": 4.953,
      "mean_token_accuracy": 0.21529059261083602,
      "num_tokens": 69467751.0,
      "step": 31105
    },
    {
      "entropy": 5.3670487880706785,
      "epoch": 3.4939630482394564,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00038213424903696673,
      "loss": 5.0058,
      "mean_token_accuracy": 0.21764868199825288,
      "num_tokens": 69478898.0,
      "step": 31110
    },
    {
      "entropy": 5.348469591140747,
      "epoch": 3.494524625147414,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00038209894203858477,
      "loss": 4.9545,
      "mean_token_accuracy": 0.21624079048633577,
      "num_tokens": 69488367.0,
      "step": 31115
    },
    {
      "entropy": 5.238032960891724,
      "epoch": 3.4950862020553712,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038206363163013093,
      "loss": 4.9307,
      "mean_token_accuracy": 0.2222967654466629,
      "num_tokens": 69500446.0,
      "step": 31120
    },
    {
      "entropy": 5.325333881378174,
      "epoch": 3.495647778963329,
      "grad_norm": 0.88671875,
      "learning_rate": 0.00038202831781272944,
      "loss": 4.9633,
      "mean_token_accuracy": 0.21658910512924195,
      "num_tokens": 69512383.0,
      "step": 31125
    },
    {
      "entropy": 5.292768573760986,
      "epoch": 3.4962093558712866,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00038199300058750494,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22986701875925064,
      "num_tokens": 69523238.0,
      "step": 31130
    },
    {
      "entropy": 5.373762607574463,
      "epoch": 3.4967709327792442,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038195767995558174,
      "loss": 5.0424,
      "mean_token_accuracy": 0.2102284848690033,
      "num_tokens": 69535196.0,
      "step": 31135
    },
    {
      "entropy": 5.256552124023438,
      "epoch": 3.497332509687202,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003819223559180845,
      "loss": 4.8454,
      "mean_token_accuracy": 0.22656999677419662,
      "num_tokens": 69546896.0,
      "step": 31140
    },
    {
      "entropy": 5.193078899383545,
      "epoch": 3.497894086595159,
      "grad_norm": 0.984375,
      "learning_rate": 0.000381887028476138,
      "loss": 4.8493,
      "mean_token_accuracy": 0.21589453667402267,
      "num_tokens": 69558230.0,
      "step": 31145
    },
    {
      "entropy": 5.1638486862182615,
      "epoch": 3.4984556635031168,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038185169763086703,
      "loss": 4.8699,
      "mean_token_accuracy": 0.22710805833339692,
      "num_tokens": 69568891.0,
      "step": 31150
    },
    {
      "entropy": 5.288677740097046,
      "epoch": 3.4990172404110744,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003818163633833966,
      "loss": 4.9796,
      "mean_token_accuracy": 0.21472001671791077,
      "num_tokens": 69579988.0,
      "step": 31155
    },
    {
      "entropy": 5.293351936340332,
      "epoch": 3.4995788173190316,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00038178102573485173,
      "loss": 4.8619,
      "mean_token_accuracy": 0.225670625269413,
      "num_tokens": 69590714.0,
      "step": 31160
    },
    {
      "entropy": 5.2094367980957035,
      "epoch": 3.5001403942269893,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003817456846863575,
      "loss": 4.8546,
      "mean_token_accuracy": 0.231102654337883,
      "num_tokens": 69601755.0,
      "step": 31165
    },
    {
      "entropy": 5.1854113101959225,
      "epoch": 3.500701971134947,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00038171034023903937,
      "loss": 4.857,
      "mean_token_accuracy": 0.22345185428857803,
      "num_tokens": 69612862.0,
      "step": 31170
    },
    {
      "entropy": 5.338063955307007,
      "epoch": 3.5012635480429046,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003816749923940226,
      "loss": 4.9605,
      "mean_token_accuracy": 0.21644991636276245,
      "num_tokens": 69624746.0,
      "step": 31175
    },
    {
      "entropy": 5.245156526565552,
      "epoch": 3.5018251249508623,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038163964115243276,
      "loss": 4.7874,
      "mean_token_accuracy": 0.22973092943429946,
      "num_tokens": 69634477.0,
      "step": 31180
    },
    {
      "entropy": 5.236641788482666,
      "epoch": 3.5023867018588195,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038160428651539535,
      "loss": 4.9477,
      "mean_token_accuracy": 0.21559301763772964,
      "num_tokens": 69645913.0,
      "step": 31185
    },
    {
      "entropy": 5.241494989395141,
      "epoch": 3.502948278766777,
      "grad_norm": 0.984375,
      "learning_rate": 0.00038156892848403606,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22181134521961213,
      "num_tokens": 69657727.0,
      "step": 31190
    },
    {
      "entropy": 5.184851884841919,
      "epoch": 3.503509855674735,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003815335670594808,
      "loss": 4.7534,
      "mean_token_accuracy": 0.2463948592543602,
      "num_tokens": 69668649.0,
      "step": 31195
    },
    {
      "entropy": 5.279690217971802,
      "epoch": 3.504071432582692,
      "grad_norm": 1.0,
      "learning_rate": 0.00038149820224285546,
      "loss": 4.9525,
      "mean_token_accuracy": 0.21845057606697083,
      "num_tokens": 69680286.0,
      "step": 31200
    },
    {
      "entropy": 5.243025779724121,
      "epoch": 3.5046330094906497,
      "grad_norm": 1.109375,
      "learning_rate": 0.000381462834035286,
      "loss": 4.9025,
      "mean_token_accuracy": 0.22045980989933014,
      "num_tokens": 69689801.0,
      "step": 31205
    },
    {
      "entropy": 5.269993829727173,
      "epoch": 3.5051945863986074,
      "grad_norm": 0.8671875,
      "learning_rate": 0.0003814274624378986,
      "loss": 4.9027,
      "mean_token_accuracy": 0.22730293720960618,
      "num_tokens": 69700898.0,
      "step": 31210
    },
    {
      "entropy": 5.260734891891479,
      "epoch": 3.5057561633065646,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003813920874518196,
      "loss": 4.8389,
      "mean_token_accuracy": 0.2302117571234703,
      "num_tokens": 69711969.0,
      "step": 31215
    },
    {
      "entropy": 5.252000904083252,
      "epoch": 3.5063177402145222,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003813567090781752,
      "loss": 4.9385,
      "mean_token_accuracy": 0.221102374792099,
      "num_tokens": 69723458.0,
      "step": 31220
    },
    {
      "entropy": 5.253841543197632,
      "epoch": 3.50687931712248,
      "grad_norm": 0.9296875,
      "learning_rate": 0.000381321327318092,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22049285620450973,
      "num_tokens": 69735527.0,
      "step": 31225
    },
    {
      "entropy": 5.364801931381225,
      "epoch": 3.5074408940304376,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003812859421726964,
      "loss": 5.0083,
      "mean_token_accuracy": 0.22103057503700257,
      "num_tokens": 69748829.0,
      "step": 31230
    },
    {
      "entropy": 5.268930387496948,
      "epoch": 3.5080024709383952,
      "grad_norm": 1.078125,
      "learning_rate": 0.00038125055364311526,
      "loss": 4.7869,
      "mean_token_accuracy": 0.22767316102981566,
      "num_tokens": 69758970.0,
      "step": 31235
    },
    {
      "entropy": 5.246203374862671,
      "epoch": 3.5085640478463525,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003812151617304753,
      "loss": 4.9604,
      "mean_token_accuracy": 0.21572318077087402,
      "num_tokens": 69770082.0,
      "step": 31240
    },
    {
      "entropy": 5.23568172454834,
      "epoch": 3.50912562475431,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003811797664359033,
      "loss": 4.8952,
      "mean_token_accuracy": 0.220041923224926,
      "num_tokens": 69781386.0,
      "step": 31245
    },
    {
      "entropy": 5.330697345733642,
      "epoch": 3.5096872016622678,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038114436776052644,
      "loss": 4.9402,
      "mean_token_accuracy": 0.22074733674526215,
      "num_tokens": 69792571.0,
      "step": 31250
    },
    {
      "entropy": 5.343241739273071,
      "epoch": 3.510248778570225,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003811089657054717,
      "loss": 4.9752,
      "mean_token_accuracy": 0.22168799191713334,
      "num_tokens": 69804737.0,
      "step": 31255
    },
    {
      "entropy": 5.344256925582886,
      "epoch": 3.5108103554781827,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003810735602718664,
      "loss": 5.0301,
      "mean_token_accuracy": 0.2193372666835785,
      "num_tokens": 69816584.0,
      "step": 31260
    },
    {
      "entropy": 5.156927871704101,
      "epoch": 3.5113719323861403,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003810381514608378,
      "loss": 4.79,
      "mean_token_accuracy": 0.23711262345314027,
      "num_tokens": 69828657.0,
      "step": 31265
    },
    {
      "entropy": 5.268977642059326,
      "epoch": 3.511933509294098,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003810027392735133,
      "loss": 4.9634,
      "mean_token_accuracy": 0.2234957128763199,
      "num_tokens": 69839780.0,
      "step": 31270
    },
    {
      "entropy": 5.272912502288818,
      "epoch": 3.5124950862020556,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003809673237110204,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22357662618160248,
      "num_tokens": 69850452.0,
      "step": 31275
    },
    {
      "entropy": 5.279479026794434,
      "epoch": 3.513056663110013,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003809319047744869,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21757915169000625,
      "num_tokens": 69861840.0,
      "step": 31280
    },
    {
      "entropy": 5.309338331222534,
      "epoch": 3.5136182400179705,
      "grad_norm": 0.953125,
      "learning_rate": 0.00038089648246504045,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22274551093578338,
      "num_tokens": 69872653.0,
      "step": 31285
    },
    {
      "entropy": 5.255258464813233,
      "epoch": 3.514179816925928,
      "grad_norm": 1.109375,
      "learning_rate": 0.00038086105678380896,
      "loss": 4.9328,
      "mean_token_accuracy": 0.2218051463365555,
      "num_tokens": 69883381.0,
      "step": 31290
    },
    {
      "entropy": 5.309391164779663,
      "epoch": 3.5147413938338854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038082562773192026,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21547392010688782,
      "num_tokens": 69894732.0,
      "step": 31295
    },
    {
      "entropy": 5.356016159057617,
      "epoch": 3.515302970741843,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00038079019531050257,
      "loss": 4.9477,
      "mean_token_accuracy": 0.22176139205694198,
      "num_tokens": 69905645.0,
      "step": 31300
    },
    {
      "entropy": 5.358301782608033,
      "epoch": 3.5158645476498007,
      "grad_norm": 1.015625,
      "learning_rate": 0.00038075475952068405,
      "loss": 4.9531,
      "mean_token_accuracy": 0.21660911291837692,
      "num_tokens": 69917176.0,
      "step": 31305
    },
    {
      "entropy": 5.265824842453003,
      "epoch": 3.516426124557758,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000380719320363593,
      "loss": 5.0043,
      "mean_token_accuracy": 0.2185557708144188,
      "num_tokens": 69928923.0,
      "step": 31310
    },
    {
      "entropy": 5.2381532192230225,
      "epoch": 3.5169877014657156,
      "grad_norm": 1.046875,
      "learning_rate": 0.00038068387784035774,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22218625992536545,
      "num_tokens": 69938981.0,
      "step": 31315
    },
    {
      "entropy": 5.282177066802978,
      "epoch": 3.5175492783736733,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003806484319521069,
      "loss": 4.8592,
      "mean_token_accuracy": 0.22034336626529694,
      "num_tokens": 69949265.0,
      "step": 31320
    },
    {
      "entropy": 5.149493217468262,
      "epoch": 3.518110855281631,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00038061298269996893,
      "loss": 4.779,
      "mean_token_accuracy": 0.23319485932588577,
      "num_tokens": 69961399.0,
      "step": 31325
    },
    {
      "entropy": 5.21382737159729,
      "epoch": 3.5186724321895886,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003805775300850727,
      "loss": 4.8871,
      "mean_token_accuracy": 0.21964284628629685,
      "num_tokens": 69971601.0,
      "step": 31330
    },
    {
      "entropy": 5.17744140625,
      "epoch": 3.519234009097546,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003805420741085469,
      "loss": 4.9058,
      "mean_token_accuracy": 0.2094968780875206,
      "num_tokens": 69982375.0,
      "step": 31335
    },
    {
      "entropy": 5.305513858795166,
      "epoch": 3.5197955860055035,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003805066147715205,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22871977984905242,
      "num_tokens": 69994787.0,
      "step": 31340
    },
    {
      "entropy": 5.371834850311279,
      "epoch": 3.520357162913461,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003804711520751226,
      "loss": 5.083,
      "mean_token_accuracy": 0.21034492552280426,
      "num_tokens": 70006077.0,
      "step": 31345
    },
    {
      "entropy": 5.2852253913879395,
      "epoch": 3.5209187398214183,
      "grad_norm": 0.96875,
      "learning_rate": 0.00038043568602048227,
      "loss": 4.8848,
      "mean_token_accuracy": 0.21774938851594924,
      "num_tokens": 70016655.0,
      "step": 31350
    },
    {
      "entropy": 5.292302131652832,
      "epoch": 3.521480316729376,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003804002166087289,
      "loss": 4.8053,
      "mean_token_accuracy": 0.22682450115680694,
      "num_tokens": 70026857.0,
      "step": 31355
    },
    {
      "entropy": 5.210028076171875,
      "epoch": 3.5220418936373337,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00038036474384099174,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22169434279203415,
      "num_tokens": 70038311.0,
      "step": 31360
    },
    {
      "entropy": 5.180806398391724,
      "epoch": 3.5226034705452913,
      "grad_norm": 1.0,
      "learning_rate": 0.0003803292677184002,
      "loss": 4.8857,
      "mean_token_accuracy": 0.21967903822660445,
      "num_tokens": 70048613.0,
      "step": 31365
    },
    {
      "entropy": 5.354026412963867,
      "epoch": 3.523165047453249,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00038029378824208394,
      "loss": 4.9115,
      "mean_token_accuracy": 0.22357098907232284,
      "num_tokens": 70058686.0,
      "step": 31370
    },
    {
      "entropy": 5.298495626449585,
      "epoch": 3.523726624361206,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003802583054131727,
      "loss": 4.9478,
      "mean_token_accuracy": 0.21729585528373718,
      "num_tokens": 70068845.0,
      "step": 31375
    },
    {
      "entropy": 5.203185653686523,
      "epoch": 3.524288201269164,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003802228192327961,
      "loss": 4.896,
      "mean_token_accuracy": 0.2182785004377365,
      "num_tokens": 70080223.0,
      "step": 31380
    },
    {
      "entropy": 5.324680185317993,
      "epoch": 3.5248497781771215,
      "grad_norm": 1.03125,
      "learning_rate": 0.00038018732970208414,
      "loss": 4.9716,
      "mean_token_accuracy": 0.21725079268217087,
      "num_tokens": 70091729.0,
      "step": 31385
    },
    {
      "entropy": 5.2492162704467775,
      "epoch": 3.5254113550850787,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003801518368221668,
      "loss": 4.8833,
      "mean_token_accuracy": 0.2154928371310234,
      "num_tokens": 70103172.0,
      "step": 31390
    },
    {
      "entropy": 5.292818641662597,
      "epoch": 3.5259729319930364,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003801163405941743,
      "loss": 4.9943,
      "mean_token_accuracy": 0.21241905242204667,
      "num_tokens": 70114303.0,
      "step": 31395
    },
    {
      "entropy": 5.2457976818084715,
      "epoch": 3.526534508900994,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00038008084101923665,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2214284896850586,
      "num_tokens": 70126802.0,
      "step": 31400
    },
    {
      "entropy": 5.336067247390747,
      "epoch": 3.5270960858089513,
      "grad_norm": 0.921875,
      "learning_rate": 0.00038004533809848435,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22416297644376754,
      "num_tokens": 70138768.0,
      "step": 31405
    },
    {
      "entropy": 5.2100754737854,
      "epoch": 3.527657662716909,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00038000983183304765,
      "loss": 4.8692,
      "mean_token_accuracy": 0.22201130986213685,
      "num_tokens": 70148411.0,
      "step": 31410
    },
    {
      "entropy": 5.190782165527343,
      "epoch": 3.5282192396248666,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003799743222240572,
      "loss": 4.8548,
      "mean_token_accuracy": 0.22186959385871888,
      "num_tokens": 70159788.0,
      "step": 31415
    },
    {
      "entropy": 5.270484590530396,
      "epoch": 3.5287808165328243,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003799388092726436,
      "loss": 4.817,
      "mean_token_accuracy": 0.22892435193061828,
      "num_tokens": 70169707.0,
      "step": 31420
    },
    {
      "entropy": 5.247503757476807,
      "epoch": 3.529342393440782,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037990329297993763,
      "loss": 4.821,
      "mean_token_accuracy": 0.22797299772500992,
      "num_tokens": 70180958.0,
      "step": 31425
    },
    {
      "entropy": 5.172607374191284,
      "epoch": 3.529903970348739,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003798677733470701,
      "loss": 4.8457,
      "mean_token_accuracy": 0.21702926754951476,
      "num_tokens": 70192848.0,
      "step": 31430
    },
    {
      "entropy": 5.215887880325317,
      "epoch": 3.530465547256697,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037983225037517207,
      "loss": 4.8756,
      "mean_token_accuracy": 0.22190503478050233,
      "num_tokens": 70204034.0,
      "step": 31435
    },
    {
      "entropy": 5.291938781738281,
      "epoch": 3.5310271241646545,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037979672406537445,
      "loss": 4.9241,
      "mean_token_accuracy": 0.21853047311306,
      "num_tokens": 70214274.0,
      "step": 31440
    },
    {
      "entropy": 5.285398674011231,
      "epoch": 3.5315887010726117,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003797611944188086,
      "loss": 4.9324,
      "mean_token_accuracy": 0.2177039086818695,
      "num_tokens": 70223890.0,
      "step": 31445
    },
    {
      "entropy": 5.306841611862183,
      "epoch": 3.5321502779805694,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037972566143660566,
      "loss": 4.9329,
      "mean_token_accuracy": 0.21936523765325547,
      "num_tokens": 70234529.0,
      "step": 31450
    },
    {
      "entropy": 5.139993333816529,
      "epoch": 3.532711854888527,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003796901251198969,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22200731486082076,
      "num_tokens": 70246900.0,
      "step": 31455
    },
    {
      "entropy": 5.343884801864624,
      "epoch": 3.5332734317964847,
      "grad_norm": 1.0625,
      "learning_rate": 0.000379654585469814,
      "loss": 5.04,
      "mean_token_accuracy": 0.21525174528360366,
      "num_tokens": 70257942.0,
      "step": 31460
    },
    {
      "entropy": 5.343915271759033,
      "epoch": 3.5338350087044423,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003796190424874886,
      "loss": 4.9593,
      "mean_token_accuracy": 0.22097791582345963,
      "num_tokens": 70270297.0,
      "step": 31465
    },
    {
      "entropy": 5.3003171443939205,
      "epoch": 3.5343965856123996,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003795834961740523,
      "loss": 4.9862,
      "mean_token_accuracy": 0.21566267907619477,
      "num_tokens": 70281542.0,
      "step": 31470
    },
    {
      "entropy": 5.225385332107544,
      "epoch": 3.534958162520357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037954794653063695,
      "loss": 4.901,
      "mean_token_accuracy": 0.21599526852369308,
      "num_tokens": 70290957.0,
      "step": 31475
    },
    {
      "entropy": 5.260422801971435,
      "epoch": 3.535519739428315,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003795123935583743,
      "loss": 4.928,
      "mean_token_accuracy": 0.2272691920399666,
      "num_tokens": 70302796.0,
      "step": 31480
    },
    {
      "entropy": 5.2723253726959225,
      "epoch": 3.536081316336272,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003794768372583966,
      "loss": 4.8668,
      "mean_token_accuracy": 0.22694734185934068,
      "num_tokens": 70313709.0,
      "step": 31485
    },
    {
      "entropy": 5.3060425281524655,
      "epoch": 3.5366428932442298,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037944127763183587,
      "loss": 4.9443,
      "mean_token_accuracy": 0.22256160229444505,
      "num_tokens": 70324651.0,
      "step": 31490
    },
    {
      "entropy": 5.3426416397094725,
      "epoch": 3.5372044701521874,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037940571467982436,
      "loss": 5.0427,
      "mean_token_accuracy": 0.21060896068811416,
      "num_tokens": 70335810.0,
      "step": 31495
    },
    {
      "entropy": 5.206396293640137,
      "epoch": 3.5377660470601446,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003793701484034944,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22464808821678162,
      "num_tokens": 70347171.0,
      "step": 31500
    },
    {
      "entropy": 5.209252405166626,
      "epoch": 3.5383276239681023,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037933457880397843,
      "loss": 4.8009,
      "mean_token_accuracy": 0.22968938201665878,
      "num_tokens": 70359122.0,
      "step": 31505
    },
    {
      "entropy": 5.1525859355926515,
      "epoch": 3.53888920087606,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037929900588240897,
      "loss": 4.7906,
      "mean_token_accuracy": 0.22593820244073867,
      "num_tokens": 70369691.0,
      "step": 31510
    },
    {
      "entropy": 5.238293552398682,
      "epoch": 3.5394507777840176,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003792634296399187,
      "loss": 4.9033,
      "mean_token_accuracy": 0.2232735797762871,
      "num_tokens": 70379602.0,
      "step": 31515
    },
    {
      "entropy": 5.2822113037109375,
      "epoch": 3.5400123546919753,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00037922785007764043,
      "loss": 4.9535,
      "mean_token_accuracy": 0.21569757759571076,
      "num_tokens": 70391819.0,
      "step": 31520
    },
    {
      "entropy": 5.376216983795166,
      "epoch": 3.5405739315999325,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000379192267196707,
      "loss": 4.9701,
      "mean_token_accuracy": 0.22262802124023437,
      "num_tokens": 70402078.0,
      "step": 31525
    },
    {
      "entropy": 5.2195271968841555,
      "epoch": 3.54113550850789,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037915668099825135,
      "loss": 4.8766,
      "mean_token_accuracy": 0.21723756045103074,
      "num_tokens": 70413829.0,
      "step": 31530
    },
    {
      "entropy": 5.1297770023345945,
      "epoch": 3.541697085415848,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037912109148340663,
      "loss": 4.8023,
      "mean_token_accuracy": 0.22305222004652023,
      "num_tokens": 70426221.0,
      "step": 31535
    },
    {
      "entropy": 5.36114182472229,
      "epoch": 3.542258662323805,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00037908549865330596,
      "loss": 5.0098,
      "mean_token_accuracy": 0.2109961062669754,
      "num_tokens": 70437215.0,
      "step": 31540
    },
    {
      "entropy": 5.246729230880737,
      "epoch": 3.5428202392317627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037904990250908256,
      "loss": 4.7907,
      "mean_token_accuracy": 0.22558653503656387,
      "num_tokens": 70448918.0,
      "step": 31545
    },
    {
      "entropy": 5.2188623428344725,
      "epoch": 3.5433818161397204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037901430305187,
      "loss": 4.9196,
      "mean_token_accuracy": 0.2186692476272583,
      "num_tokens": 70460713.0,
      "step": 31550
    },
    {
      "entropy": 5.237751388549805,
      "epoch": 3.543943393047678,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037897870028280167,
      "loss": 4.8941,
      "mean_token_accuracy": 0.22018941938877107,
      "num_tokens": 70471460.0,
      "step": 31555
    },
    {
      "entropy": 5.238285875320434,
      "epoch": 3.5445049699556357,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003789430942030112,
      "loss": 4.9099,
      "mean_token_accuracy": 0.2234008103609085,
      "num_tokens": 70482359.0,
      "step": 31560
    },
    {
      "entropy": 5.30925235748291,
      "epoch": 3.545066546863593,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037890748481363216,
      "loss": 4.979,
      "mean_token_accuracy": 0.21494434922933578,
      "num_tokens": 70493272.0,
      "step": 31565
    },
    {
      "entropy": 5.285362195968628,
      "epoch": 3.5456281237715506,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003788718721157987,
      "loss": 4.8725,
      "mean_token_accuracy": 0.22772180289030075,
      "num_tokens": 70504118.0,
      "step": 31570
    },
    {
      "entropy": 5.359103155136109,
      "epoch": 3.5461897006795082,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003788362561106445,
      "loss": 5.0155,
      "mean_token_accuracy": 0.21703399121761321,
      "num_tokens": 70515379.0,
      "step": 31575
    },
    {
      "entropy": 5.269361305236816,
      "epoch": 3.5467512775874654,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037880063679930355,
      "loss": 4.9134,
      "mean_token_accuracy": 0.21881508231163024,
      "num_tokens": 70526047.0,
      "step": 31580
    },
    {
      "entropy": 5.29920597076416,
      "epoch": 3.547312854495423,
      "grad_norm": 1.0,
      "learning_rate": 0.00037876501418291015,
      "loss": 4.8822,
      "mean_token_accuracy": 0.2311604753136635,
      "num_tokens": 70537614.0,
      "step": 31585
    },
    {
      "entropy": 5.2817731380462645,
      "epoch": 3.5478744314033808,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003787293882625984,
      "loss": 4.8821,
      "mean_token_accuracy": 0.2220386043190956,
      "num_tokens": 70548048.0,
      "step": 31590
    },
    {
      "entropy": 5.1972309112548825,
      "epoch": 3.548436008311338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003786937590395027,
      "loss": 4.7848,
      "mean_token_accuracy": 0.23091035187244416,
      "num_tokens": 70558668.0,
      "step": 31595
    },
    {
      "entropy": 5.255833673477173,
      "epoch": 3.5489975852192956,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003786581265147575,
      "loss": 4.9992,
      "mean_token_accuracy": 0.21684378534555435,
      "num_tokens": 70570044.0,
      "step": 31600
    },
    {
      "entropy": 5.231251525878906,
      "epoch": 3.5495591621272533,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003786224906894974,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23179724365472792,
      "num_tokens": 70580857.0,
      "step": 31605
    },
    {
      "entropy": 5.164800119400025,
      "epoch": 3.550120739035211,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000378586851564857,
      "loss": 4.7764,
      "mean_token_accuracy": 0.22789042294025422,
      "num_tokens": 70592476.0,
      "step": 31610
    },
    {
      "entropy": 5.222618770599365,
      "epoch": 3.5506823159431686,
      "grad_norm": 1.015625,
      "learning_rate": 0.000378551209141971,
      "loss": 4.8599,
      "mean_token_accuracy": 0.22453185319900512,
      "num_tokens": 70603614.0,
      "step": 31615
    },
    {
      "entropy": 5.263078260421753,
      "epoch": 3.551243892851126,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003785155634219744,
      "loss": 4.9336,
      "mean_token_accuracy": 0.20799701362848283,
      "num_tokens": 70615239.0,
      "step": 31620
    },
    {
      "entropy": 5.3030296802520756,
      "epoch": 3.5518054697590835,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003784799144060021,
      "loss": 4.9055,
      "mean_token_accuracy": 0.22167692184448243,
      "num_tokens": 70625785.0,
      "step": 31625
    },
    {
      "entropy": 5.297403955459595,
      "epoch": 3.552367046667041,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037844426209518916,
      "loss": 4.9175,
      "mean_token_accuracy": 0.2219441667199135,
      "num_tokens": 70636942.0,
      "step": 31630
    },
    {
      "entropy": 5.212947130203247,
      "epoch": 3.5529286235749984,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037840860649067084,
      "loss": 4.7861,
      "mean_token_accuracy": 0.22593677788972855,
      "num_tokens": 70648106.0,
      "step": 31635
    },
    {
      "entropy": 5.197220802307129,
      "epoch": 3.553490200482956,
      "grad_norm": 0.984375,
      "learning_rate": 0.00037837294759358235,
      "loss": 4.8617,
      "mean_token_accuracy": 0.22580180168151856,
      "num_tokens": 70658850.0,
      "step": 31640
    },
    {
      "entropy": 5.327583694458008,
      "epoch": 3.5540517773909137,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037833728540505914,
      "loss": 4.9625,
      "mean_token_accuracy": 0.21310903131961823,
      "num_tokens": 70671060.0,
      "step": 31645
    },
    {
      "entropy": 5.213363599777222,
      "epoch": 3.5546133542988714,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003783016199262367,
      "loss": 4.797,
      "mean_token_accuracy": 0.23176826983690263,
      "num_tokens": 70682186.0,
      "step": 31650
    },
    {
      "entropy": 5.200950384140015,
      "epoch": 3.555174931206829,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003782659511582505,
      "loss": 4.8749,
      "mean_token_accuracy": 0.22303504049777984,
      "num_tokens": 70692622.0,
      "step": 31655
    },
    {
      "entropy": 5.174408769607544,
      "epoch": 3.5557365081147863,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003782302791022365,
      "loss": 4.8451,
      "mean_token_accuracy": 0.22674063891172408,
      "num_tokens": 70703293.0,
      "step": 31660
    },
    {
      "entropy": 5.169470500946045,
      "epoch": 3.556298085022744,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037819460375933034,
      "loss": 4.7868,
      "mean_token_accuracy": 0.22596292942762375,
      "num_tokens": 70714961.0,
      "step": 31665
    },
    {
      "entropy": 5.309544515609741,
      "epoch": 3.5568596619307016,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003781589251306679,
      "loss": 4.9801,
      "mean_token_accuracy": 0.21475697606801986,
      "num_tokens": 70727030.0,
      "step": 31670
    },
    {
      "entropy": 5.299641466140747,
      "epoch": 3.557421238838659,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003781232432173853,
      "loss": 4.871,
      "mean_token_accuracy": 0.21857506185770034,
      "num_tokens": 70738331.0,
      "step": 31675
    },
    {
      "entropy": 5.224859571456909,
      "epoch": 3.5579828157466165,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00037808755802061874,
      "loss": 4.9247,
      "mean_token_accuracy": 0.2143504425883293,
      "num_tokens": 70748347.0,
      "step": 31680
    },
    {
      "entropy": 5.285446834564209,
      "epoch": 3.558544392654574,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003780518695415043,
      "loss": 4.9205,
      "mean_token_accuracy": 0.22881782799959183,
      "num_tokens": 70759905.0,
      "step": 31685
    },
    {
      "entropy": 5.362414073944092,
      "epoch": 3.5591059695625313,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037801617778117835,
      "loss": 4.9664,
      "mean_token_accuracy": 0.22226063758134842,
      "num_tokens": 70770538.0,
      "step": 31690
    },
    {
      "entropy": 5.24901933670044,
      "epoch": 3.559667546470489,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037798048274077735,
      "loss": 4.978,
      "mean_token_accuracy": 0.21919480860233306,
      "num_tokens": 70781881.0,
      "step": 31695
    },
    {
      "entropy": 5.137257528305054,
      "epoch": 3.5602291233784467,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037794478442143786,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2278719425201416,
      "num_tokens": 70792299.0,
      "step": 31700
    },
    {
      "entropy": 5.299091243743897,
      "epoch": 3.5607907002864043,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037790908282429647,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2250654548406601,
      "num_tokens": 70802470.0,
      "step": 31705
    },
    {
      "entropy": 5.30302619934082,
      "epoch": 3.561352277194362,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037787337795049,
      "loss": 4.9813,
      "mean_token_accuracy": 0.2216277152299881,
      "num_tokens": 70813502.0,
      "step": 31710
    },
    {
      "entropy": 5.220501041412353,
      "epoch": 3.561913854102319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037783766980115525,
      "loss": 4.8165,
      "mean_token_accuracy": 0.2204776406288147,
      "num_tokens": 70823917.0,
      "step": 31715
    },
    {
      "entropy": 5.278582286834717,
      "epoch": 3.562475431010277,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037780195837742927,
      "loss": 4.9989,
      "mean_token_accuracy": 0.21323640197515487,
      "num_tokens": 70834503.0,
      "step": 31720
    },
    {
      "entropy": 5.281810760498047,
      "epoch": 3.5630370079182345,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037776624368044895,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22523510158061982,
      "num_tokens": 70844954.0,
      "step": 31725
    },
    {
      "entropy": 5.33897819519043,
      "epoch": 3.5635985848261917,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037773052571135157,
      "loss": 5.0186,
      "mean_token_accuracy": 0.21718216091394424,
      "num_tokens": 70856142.0,
      "step": 31730
    },
    {
      "entropy": 5.2574952125549315,
      "epoch": 3.5641601617341494,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003776948044712745,
      "loss": 4.8958,
      "mean_token_accuracy": 0.22379774302244188,
      "num_tokens": 70866880.0,
      "step": 31735
    },
    {
      "entropy": 5.268126821517944,
      "epoch": 3.564721738642107,
      "grad_norm": 1.0625,
      "learning_rate": 0.000377659079961355,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22769761085510254,
      "num_tokens": 70876836.0,
      "step": 31740
    },
    {
      "entropy": 5.2444499969482425,
      "epoch": 3.5652833155500647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037762335218273057,
      "loss": 4.8962,
      "mean_token_accuracy": 0.22097367495298387,
      "num_tokens": 70889430.0,
      "step": 31745
    },
    {
      "entropy": 5.186123466491699,
      "epoch": 3.5658448924580224,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003775876211365388,
      "loss": 4.7917,
      "mean_token_accuracy": 0.22603993564844133,
      "num_tokens": 70899291.0,
      "step": 31750
    },
    {
      "entropy": 5.262025451660156,
      "epoch": 3.5664064693659796,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003775518868239174,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22410556972026824,
      "num_tokens": 70910596.0,
      "step": 31755
    },
    {
      "entropy": 5.300336694717407,
      "epoch": 3.5669680462739373,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037751614924600417,
      "loss": 4.9049,
      "mean_token_accuracy": 0.2239297956228256,
      "num_tokens": 70921969.0,
      "step": 31760
    },
    {
      "entropy": 5.266104316711425,
      "epoch": 3.567529623181895,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000377480408403937,
      "loss": 4.8584,
      "mean_token_accuracy": 0.22293806225061416,
      "num_tokens": 70933372.0,
      "step": 31765
    },
    {
      "entropy": 5.306376123428345,
      "epoch": 3.568091200089852,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003774446642988538,
      "loss": 5.0097,
      "mean_token_accuracy": 0.21460439413785934,
      "num_tokens": 70944989.0,
      "step": 31770
    },
    {
      "entropy": 5.247355842590332,
      "epoch": 3.56865277699781,
      "grad_norm": 1.125,
      "learning_rate": 0.00037740891693189286,
      "loss": 4.9511,
      "mean_token_accuracy": 0.22172735780477523,
      "num_tokens": 70957516.0,
      "step": 31775
    },
    {
      "entropy": 5.369787502288818,
      "epoch": 3.5692143539057675,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003773731663041922,
      "loss": 4.9779,
      "mean_token_accuracy": 0.21640739142894744,
      "num_tokens": 70969485.0,
      "step": 31780
    },
    {
      "entropy": 5.307403707504273,
      "epoch": 3.5697759308137247,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037733741241689024,
      "loss": 4.9029,
      "mean_token_accuracy": 0.2191530853509903,
      "num_tokens": 70980335.0,
      "step": 31785
    },
    {
      "entropy": 5.215348768234253,
      "epoch": 3.5703375077216823,
      "grad_norm": 1.0,
      "learning_rate": 0.0003773016552711254,
      "loss": 4.8558,
      "mean_token_accuracy": 0.22398925870656966,
      "num_tokens": 70991870.0,
      "step": 31790
    },
    {
      "entropy": 5.253462076187134,
      "epoch": 3.57089908462964,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037726589486803617,
      "loss": 5.013,
      "mean_token_accuracy": 0.21576717495918274,
      "num_tokens": 71003906.0,
      "step": 31795
    },
    {
      "entropy": 5.194365930557251,
      "epoch": 3.5714606615375977,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037723013120876113,
      "loss": 4.8601,
      "mean_token_accuracy": 0.22086064964532853,
      "num_tokens": 71015280.0,
      "step": 31800
    },
    {
      "entropy": 5.299495029449463,
      "epoch": 3.5720222384455553,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037719436429443915,
      "loss": 4.9282,
      "mean_token_accuracy": 0.22220003008842468,
      "num_tokens": 71026299.0,
      "step": 31805
    },
    {
      "entropy": 5.37266321182251,
      "epoch": 3.5725838153535125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037715859412620895,
      "loss": 4.9856,
      "mean_token_accuracy": 0.21926547586917877,
      "num_tokens": 71037704.0,
      "step": 31810
    },
    {
      "entropy": 5.281594467163086,
      "epoch": 3.57314539226147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003771228207052095,
      "loss": 4.9288,
      "mean_token_accuracy": 0.22022275626659393,
      "num_tokens": 71048359.0,
      "step": 31815
    },
    {
      "entropy": 5.249031972885132,
      "epoch": 3.573706969169428,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037708704403257974,
      "loss": 4.9013,
      "mean_token_accuracy": 0.226918725669384,
      "num_tokens": 71059723.0,
      "step": 31820
    },
    {
      "entropy": 5.159786367416382,
      "epoch": 3.574268546077385,
      "grad_norm": 0.96875,
      "learning_rate": 0.000377051264109459,
      "loss": 4.8104,
      "mean_token_accuracy": 0.2300056979060173,
      "num_tokens": 71071437.0,
      "step": 31825
    },
    {
      "entropy": 5.199548244476318,
      "epoch": 3.5748301229853428,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003770154809369864,
      "loss": 4.7394,
      "mean_token_accuracy": 0.2362395480275154,
      "num_tokens": 71082544.0,
      "step": 31830
    },
    {
      "entropy": 5.233361101150512,
      "epoch": 3.5753916998933004,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037697969451630135,
      "loss": 4.847,
      "mean_token_accuracy": 0.22386083304882048,
      "num_tokens": 71092429.0,
      "step": 31835
    },
    {
      "entropy": 5.192736721038818,
      "epoch": 3.575953276801258,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00037694390484854323,
      "loss": 4.897,
      "mean_token_accuracy": 0.22112292647361756,
      "num_tokens": 71103154.0,
      "step": 31840
    },
    {
      "entropy": 5.240165710449219,
      "epoch": 3.5765148537092157,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003769081119348517,
      "loss": 4.826,
      "mean_token_accuracy": 0.23392991870641708,
      "num_tokens": 71112921.0,
      "step": 31845
    },
    {
      "entropy": 5.278556251525879,
      "epoch": 3.577076430617173,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037687231577636626,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22167955487966537,
      "num_tokens": 71123281.0,
      "step": 31850
    },
    {
      "entropy": 5.29417724609375,
      "epoch": 3.5776380075251306,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003768365163742269,
      "loss": 5.0156,
      "mean_token_accuracy": 0.2132173091173172,
      "num_tokens": 71134358.0,
      "step": 31855
    },
    {
      "entropy": 5.345920085906982,
      "epoch": 3.5781995844330883,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037680071372957327,
      "loss": 4.9259,
      "mean_token_accuracy": 0.21993509382009507,
      "num_tokens": 71146805.0,
      "step": 31860
    },
    {
      "entropy": 5.274719619750977,
      "epoch": 3.5787611613410455,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037676490784354543,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22490054070949556,
      "num_tokens": 71157388.0,
      "step": 31865
    },
    {
      "entropy": 5.071559429168701,
      "epoch": 3.579322738249003,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003767290987172835,
      "loss": 4.7226,
      "mean_token_accuracy": 0.23105214089155196,
      "num_tokens": 71168601.0,
      "step": 31870
    },
    {
      "entropy": 5.210654878616333,
      "epoch": 3.579884315156961,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00037669328635192757,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2233909100294113,
      "num_tokens": 71180796.0,
      "step": 31875
    },
    {
      "entropy": 5.199983930587768,
      "epoch": 3.580445892064918,
      "grad_norm": 1.0,
      "learning_rate": 0.000376657470748618,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22891200333833694,
      "num_tokens": 71191686.0,
      "step": 31880
    },
    {
      "entropy": 5.216412782669067,
      "epoch": 3.5810074689728757,
      "grad_norm": 1.0,
      "learning_rate": 0.0003766216519084952,
      "loss": 4.7166,
      "mean_token_accuracy": 0.2313363179564476,
      "num_tokens": 71202454.0,
      "step": 31885
    },
    {
      "entropy": 5.248447465896606,
      "epoch": 3.5815690458808334,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003765858298326994,
      "loss": 4.8139,
      "mean_token_accuracy": 0.2326980859041214,
      "num_tokens": 71213729.0,
      "step": 31890
    },
    {
      "entropy": 5.33596568107605,
      "epoch": 3.582130622788791,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037655000452237154,
      "loss": 5.0602,
      "mean_token_accuracy": 0.2139044836163521,
      "num_tokens": 71226810.0,
      "step": 31895
    },
    {
      "entropy": 5.213614797592163,
      "epoch": 3.5826921996967487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003765141759786521,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23270394951105117,
      "num_tokens": 71237233.0,
      "step": 31900
    },
    {
      "entropy": 5.236610698699951,
      "epoch": 3.583253776604706,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037647834420268195,
      "loss": 4.8653,
      "mean_token_accuracy": 0.23275849521160125,
      "num_tokens": 71248678.0,
      "step": 31905
    },
    {
      "entropy": 5.1304127216339115,
      "epoch": 3.5838153535126636,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003764425091956019,
      "loss": 4.8327,
      "mean_token_accuracy": 0.22917104959487916,
      "num_tokens": 71259098.0,
      "step": 31910
    },
    {
      "entropy": 5.209886789321899,
      "epoch": 3.584376930420621,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037640667095855307,
      "loss": 4.8059,
      "mean_token_accuracy": 0.22991552650928498,
      "num_tokens": 71270082.0,
      "step": 31915
    },
    {
      "entropy": 5.2838883876800535,
      "epoch": 3.5849385073285784,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037637082949267654,
      "loss": 4.9033,
      "mean_token_accuracy": 0.22843223214149475,
      "num_tokens": 71280975.0,
      "step": 31920
    },
    {
      "entropy": 5.260417175292969,
      "epoch": 3.585500084236536,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003763349847991134,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2322683423757553,
      "num_tokens": 71292552.0,
      "step": 31925
    },
    {
      "entropy": 5.3163916110992435,
      "epoch": 3.5860616611444938,
      "grad_norm": 0.96875,
      "learning_rate": 0.000376299136879005,
      "loss": 4.9961,
      "mean_token_accuracy": 0.20777003318071366,
      "num_tokens": 71304653.0,
      "step": 31930
    },
    {
      "entropy": 5.231069803237915,
      "epoch": 3.5866232380524514,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037626328573349296,
      "loss": 4.9266,
      "mean_token_accuracy": 0.22244669049978255,
      "num_tokens": 71316734.0,
      "step": 31935
    },
    {
      "entropy": 5.359453201293945,
      "epoch": 3.587184814960409,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003762274313637185,
      "loss": 5.0659,
      "mean_token_accuracy": 0.20528346598148345,
      "num_tokens": 71328228.0,
      "step": 31940
    },
    {
      "entropy": 5.296473264694214,
      "epoch": 3.5877463918683663,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003761915737708234,
      "loss": 4.8787,
      "mean_token_accuracy": 0.22799051851034163,
      "num_tokens": 71339921.0,
      "step": 31945
    },
    {
      "entropy": 5.290425062179565,
      "epoch": 3.588307968776324,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00037615571295594936,
      "loss": 4.9422,
      "mean_token_accuracy": 0.2157745197415352,
      "num_tokens": 71350995.0,
      "step": 31950
    },
    {
      "entropy": 5.223694896697998,
      "epoch": 3.5888695456842816,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003761198489202382,
      "loss": 4.9169,
      "mean_token_accuracy": 0.21945166289806367,
      "num_tokens": 71362546.0,
      "step": 31955
    },
    {
      "entropy": 5.31573371887207,
      "epoch": 3.589431122592239,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003760839816648317,
      "loss": 4.9743,
      "mean_token_accuracy": 0.2217928260564804,
      "num_tokens": 71373431.0,
      "step": 31960
    },
    {
      "entropy": 5.325287294387818,
      "epoch": 3.5899926995001965,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037604811119087213,
      "loss": 4.9109,
      "mean_token_accuracy": 0.22719673663377762,
      "num_tokens": 71384488.0,
      "step": 31965
    },
    {
      "entropy": 5.247766208648682,
      "epoch": 3.590554276408154,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003760122374995015,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22266782224178314,
      "num_tokens": 71396558.0,
      "step": 31970
    },
    {
      "entropy": 5.154648542404175,
      "epoch": 3.5911158533161114,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003759763605918621,
      "loss": 4.805,
      "mean_token_accuracy": 0.23222731649875641,
      "num_tokens": 71407107.0,
      "step": 31975
    },
    {
      "entropy": 5.290012311935425,
      "epoch": 3.591677430224069,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037594048046909616,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21833115816116333,
      "num_tokens": 71419801.0,
      "step": 31980
    },
    {
      "entropy": 5.351615571975708,
      "epoch": 3.5922390071320267,
      "grad_norm": 1.0,
      "learning_rate": 0.0003759045971323462,
      "loss": 5.0345,
      "mean_token_accuracy": 0.21638135761022567,
      "num_tokens": 71431740.0,
      "step": 31985
    },
    {
      "entropy": 5.281636190414429,
      "epoch": 3.5928005840399844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037586871058275476,
      "loss": 4.9102,
      "mean_token_accuracy": 0.2215317189693451,
      "num_tokens": 71443514.0,
      "step": 31990
    },
    {
      "entropy": 5.21028699874878,
      "epoch": 3.593362160947942,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00037583282082146434,
      "loss": 4.91,
      "mean_token_accuracy": 0.223555226624012,
      "num_tokens": 71454240.0,
      "step": 31995
    },
    {
      "entropy": 5.206971931457519,
      "epoch": 3.5939237378558992,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037579692784961785,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22233678698539733,
      "num_tokens": 71464707.0,
      "step": 32000
    },
    {
      "entropy": 5.306769704818725,
      "epoch": 3.594485314763857,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003757610316683581,
      "loss": 4.917,
      "mean_token_accuracy": 0.21709832549095154,
      "num_tokens": 71475931.0,
      "step": 32005
    },
    {
      "entropy": 5.320517539978027,
      "epoch": 3.5950468916718146,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037572513227882807,
      "loss": 4.907,
      "mean_token_accuracy": 0.21442112922668458,
      "num_tokens": 71487851.0,
      "step": 32010
    },
    {
      "entropy": 5.279571485519409,
      "epoch": 3.595608468579772,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037568922968217065,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22502810806035994,
      "num_tokens": 71498144.0,
      "step": 32015
    },
    {
      "entropy": 5.2333972454071045,
      "epoch": 3.5961700454877294,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037565332387952914,
      "loss": 4.8773,
      "mean_token_accuracy": 0.22032810896635055,
      "num_tokens": 71508074.0,
      "step": 32020
    },
    {
      "entropy": 5.224322128295898,
      "epoch": 3.596731622395687,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037561741487204684,
      "loss": 4.8912,
      "mean_token_accuracy": 0.22584731876850128,
      "num_tokens": 71519609.0,
      "step": 32025
    },
    {
      "entropy": 5.181972599029541,
      "epoch": 3.5972931993036448,
      "grad_norm": 1.0,
      "learning_rate": 0.0003755815026608669,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2347597137093544,
      "num_tokens": 71530256.0,
      "step": 32030
    },
    {
      "entropy": 5.330523681640625,
      "epoch": 3.5978547762116024,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037554558724713305,
      "loss": 4.9142,
      "mean_token_accuracy": 0.21901614367961883,
      "num_tokens": 71540383.0,
      "step": 32035
    },
    {
      "entropy": 5.2799256324768065,
      "epoch": 3.5984163531195597,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003755096686319886,
      "loss": 4.8991,
      "mean_token_accuracy": 0.22149838209152223,
      "num_tokens": 71551739.0,
      "step": 32040
    },
    {
      "entropy": 5.28434386253357,
      "epoch": 3.5989779300275173,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037547374681657736,
      "loss": 4.9072,
      "mean_token_accuracy": 0.22890360355377198,
      "num_tokens": 71562429.0,
      "step": 32045
    },
    {
      "entropy": 5.310551881790161,
      "epoch": 3.599539506935475,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037543782180204303,
      "loss": 4.9514,
      "mean_token_accuracy": 0.21481739580631257,
      "num_tokens": 71573179.0,
      "step": 32050
    },
    {
      "entropy": 5.278874158859253,
      "epoch": 3.600101083843432,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037540189358952955,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2135407954454422,
      "num_tokens": 71584223.0,
      "step": 32055
    },
    {
      "entropy": 5.2636524677276615,
      "epoch": 3.60066266075139,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003753659621801808,
      "loss": 4.924,
      "mean_token_accuracy": 0.22104128301143647,
      "num_tokens": 71595166.0,
      "step": 32060
    },
    {
      "entropy": 5.267606496810913,
      "epoch": 3.6012242376593475,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037533002757514095,
      "loss": 4.9511,
      "mean_token_accuracy": 0.2195666551589966,
      "num_tokens": 71606290.0,
      "step": 32065
    },
    {
      "entropy": 5.268284606933594,
      "epoch": 3.6017858145673047,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000375294089775554,
      "loss": 4.9331,
      "mean_token_accuracy": 0.2176086738705635,
      "num_tokens": 71617545.0,
      "step": 32070
    },
    {
      "entropy": 5.26483941078186,
      "epoch": 3.6023473914752624,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003752581487825644,
      "loss": 4.8996,
      "mean_token_accuracy": 0.22694341987371444,
      "num_tokens": 71627949.0,
      "step": 32075
    },
    {
      "entropy": 5.309035778045654,
      "epoch": 3.60290896838322,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037522220459731644,
      "loss": 4.8948,
      "mean_token_accuracy": 0.21938399821519852,
      "num_tokens": 71639017.0,
      "step": 32080
    },
    {
      "entropy": 5.189262819290161,
      "epoch": 3.6034705452911777,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003751862572209546,
      "loss": 4.7814,
      "mean_token_accuracy": 0.22494042068719863,
      "num_tokens": 71649189.0,
      "step": 32085
    },
    {
      "entropy": 5.181165742874145,
      "epoch": 3.6040321221991354,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037515030665462345,
      "loss": 4.8318,
      "mean_token_accuracy": 0.22142908722162247,
      "num_tokens": 71660083.0,
      "step": 32090
    },
    {
      "entropy": 5.193778133392334,
      "epoch": 3.6045936991070926,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00037511435289946775,
      "loss": 4.8116,
      "mean_token_accuracy": 0.2274103507399559,
      "num_tokens": 71670874.0,
      "step": 32095
    },
    {
      "entropy": 5.333415889739991,
      "epoch": 3.6051552760150503,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003750783959566322,
      "loss": 4.9345,
      "mean_token_accuracy": 0.221538008749485,
      "num_tokens": 71681479.0,
      "step": 32100
    },
    {
      "entropy": 5.218907403945923,
      "epoch": 3.605716852923008,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003750424358272617,
      "loss": 4.8695,
      "mean_token_accuracy": 0.23022721856832504,
      "num_tokens": 71693865.0,
      "step": 32105
    },
    {
      "entropy": 5.167920255661011,
      "epoch": 3.606278429830965,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003750064725125013,
      "loss": 4.8099,
      "mean_token_accuracy": 0.22718888372182847,
      "num_tokens": 71704513.0,
      "step": 32110
    },
    {
      "entropy": 5.202990627288818,
      "epoch": 3.606840006738923,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037497050601349605,
      "loss": 4.892,
      "mean_token_accuracy": 0.22173561453819274,
      "num_tokens": 71715444.0,
      "step": 32115
    },
    {
      "entropy": 5.27533483505249,
      "epoch": 3.6074015836468805,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037493453633139103,
      "loss": 4.8311,
      "mean_token_accuracy": 0.22950416356325148,
      "num_tokens": 71725452.0,
      "step": 32120
    },
    {
      "entropy": 5.233268213272095,
      "epoch": 3.607963160554838,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037489856346733163,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22279053628444673,
      "num_tokens": 71736536.0,
      "step": 32125
    },
    {
      "entropy": 5.205441904067993,
      "epoch": 3.608524737462796,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003748625874224632,
      "loss": 4.8974,
      "mean_token_accuracy": 0.22190771251916885,
      "num_tokens": 71748420.0,
      "step": 32130
    },
    {
      "entropy": 5.244811964035034,
      "epoch": 3.609086314370753,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037482660819793134,
      "loss": 4.8855,
      "mean_token_accuracy": 0.2279008850455284,
      "num_tokens": 71758779.0,
      "step": 32135
    },
    {
      "entropy": 5.364521217346192,
      "epoch": 3.6096478912787107,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003747906257948814,
      "loss": 5.0767,
      "mean_token_accuracy": 0.2129412040114403,
      "num_tokens": 71771042.0,
      "step": 32140
    },
    {
      "entropy": 5.265772914886474,
      "epoch": 3.6102094681866683,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003747546402144594,
      "loss": 4.8194,
      "mean_token_accuracy": 0.23059678077697754,
      "num_tokens": 71782234.0,
      "step": 32145
    },
    {
      "entropy": 5.268557167053222,
      "epoch": 3.6107710450946255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037471865145781094,
      "loss": 4.848,
      "mean_token_accuracy": 0.22833679467439652,
      "num_tokens": 71793115.0,
      "step": 32150
    },
    {
      "entropy": 5.148587036132812,
      "epoch": 3.611332622002583,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037468265952608196,
      "loss": 4.7847,
      "mean_token_accuracy": 0.2306506708264351,
      "num_tokens": 71804466.0,
      "step": 32155
    },
    {
      "entropy": 5.245529222488403,
      "epoch": 3.611894198910541,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003746466644204183,
      "loss": 4.963,
      "mean_token_accuracy": 0.21258774548768997,
      "num_tokens": 71815248.0,
      "step": 32160
    },
    {
      "entropy": 5.331658601760864,
      "epoch": 3.612455775818498,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003746106661419663,
      "loss": 4.9335,
      "mean_token_accuracy": 0.22403114587068557,
      "num_tokens": 71825816.0,
      "step": 32165
    },
    {
      "entropy": 5.308961486816406,
      "epoch": 3.6130173527264557,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000374574664691872,
      "loss": 4.922,
      "mean_token_accuracy": 0.22170036286115646,
      "num_tokens": 71837918.0,
      "step": 32170
    },
    {
      "entropy": 5.326984834671021,
      "epoch": 3.6135789296344134,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037453866007128176,
      "loss": 4.9979,
      "mean_token_accuracy": 0.21587269455194474,
      "num_tokens": 71848736.0,
      "step": 32175
    },
    {
      "entropy": 5.198880958557129,
      "epoch": 3.614140506542371,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037450265228134203,
      "loss": 4.807,
      "mean_token_accuracy": 0.2284349262714386,
      "num_tokens": 71859903.0,
      "step": 32180
    },
    {
      "entropy": 5.245363044738769,
      "epoch": 3.6147020834503287,
      "grad_norm": 0.96875,
      "learning_rate": 0.00037446664132319925,
      "loss": 4.921,
      "mean_token_accuracy": 0.21410340815782547,
      "num_tokens": 71871656.0,
      "step": 32185
    },
    {
      "entropy": 5.28866982460022,
      "epoch": 3.615263660358286,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000374430627198,
      "loss": 4.9023,
      "mean_token_accuracy": 0.2190606564283371,
      "num_tokens": 71883282.0,
      "step": 32190
    },
    {
      "entropy": 5.292694854736328,
      "epoch": 3.6158252372662436,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037439460990689106,
      "loss": 4.8425,
      "mean_token_accuracy": 0.22486384958028793,
      "num_tokens": 71894473.0,
      "step": 32195
    },
    {
      "entropy": 5.236652374267578,
      "epoch": 3.6163868141742013,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003743585894510192,
      "loss": 5.0146,
      "mean_token_accuracy": 0.21234564632177352,
      "num_tokens": 71905555.0,
      "step": 32200
    },
    {
      "entropy": 5.201453018188476,
      "epoch": 3.6169483910821585,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037432256583153127,
      "loss": 4.8712,
      "mean_token_accuracy": 0.2172892913222313,
      "num_tokens": 71916599.0,
      "step": 32205
    },
    {
      "entropy": 5.318310022354126,
      "epoch": 3.617509967990116,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037428653904957433,
      "loss": 4.8977,
      "mean_token_accuracy": 0.22441176921129227,
      "num_tokens": 71926173.0,
      "step": 32210
    },
    {
      "entropy": 5.331372261047363,
      "epoch": 3.618071544898074,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003742505091062955,
      "loss": 4.9695,
      "mean_token_accuracy": 0.21905336230993272,
      "num_tokens": 71937389.0,
      "step": 32215
    },
    {
      "entropy": 5.238848400115967,
      "epoch": 3.6186331218060315,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003742144760028419,
      "loss": 4.9349,
      "mean_token_accuracy": 0.21938580721616746,
      "num_tokens": 71949216.0,
      "step": 32220
    },
    {
      "entropy": 5.3409752368927,
      "epoch": 3.619194698713989,
      "grad_norm": 1.03125,
      "learning_rate": 0.000374178439740361,
      "loss": 4.9765,
      "mean_token_accuracy": 0.21750541478395463,
      "num_tokens": 71959861.0,
      "step": 32225
    },
    {
      "entropy": 5.2853107929229735,
      "epoch": 3.6197562756219464,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003741424003200001,
      "loss": 4.9311,
      "mean_token_accuracy": 0.2167435571551323,
      "num_tokens": 71970043.0,
      "step": 32230
    },
    {
      "entropy": 5.2767784118652346,
      "epoch": 3.620317852529904,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037410635774290667,
      "loss": 4.8889,
      "mean_token_accuracy": 0.2260584980249405,
      "num_tokens": 71981096.0,
      "step": 32235
    },
    {
      "entropy": 5.240108346939087,
      "epoch": 3.6208794294378617,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003740703120102284,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22520883083343507,
      "num_tokens": 71992759.0,
      "step": 32240
    },
    {
      "entropy": 5.238313388824463,
      "epoch": 3.621441006345819,
      "grad_norm": 1.046875,
      "learning_rate": 0.000374034263123113,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22623490840196608,
      "num_tokens": 72003187.0,
      "step": 32245
    },
    {
      "entropy": 5.388102626800537,
      "epoch": 3.6220025832537766,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037399821108270823,
      "loss": 5.0558,
      "mean_token_accuracy": 0.21108227223157883,
      "num_tokens": 72013712.0,
      "step": 32250
    },
    {
      "entropy": 5.404075479507446,
      "epoch": 3.622564160161734,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000373962155890162,
      "loss": 5.0199,
      "mean_token_accuracy": 0.21621423065662385,
      "num_tokens": 72024693.0,
      "step": 32255
    },
    {
      "entropy": 5.239833831787109,
      "epoch": 3.6231257370696914,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003739260975466223,
      "loss": 4.8607,
      "mean_token_accuracy": 0.230918350815773,
      "num_tokens": 72036578.0,
      "step": 32260
    },
    {
      "entropy": 5.310135173797607,
      "epoch": 3.623687313977649,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037389003605323737,
      "loss": 4.9904,
      "mean_token_accuracy": 0.2172440692782402,
      "num_tokens": 72047574.0,
      "step": 32265
    },
    {
      "entropy": 5.289339256286621,
      "epoch": 3.6242488908856068,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00037385397141115535,
      "loss": 4.8961,
      "mean_token_accuracy": 0.22875420451164247,
      "num_tokens": 72059998.0,
      "step": 32270
    },
    {
      "entropy": 5.314942979812622,
      "epoch": 3.6248104677935644,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037381790362152453,
      "loss": 4.995,
      "mean_token_accuracy": 0.21705308109521865,
      "num_tokens": 72070753.0,
      "step": 32275
    },
    {
      "entropy": 5.231231546401977,
      "epoch": 3.625372044701522,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00037378183268549326,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22546031028032304,
      "num_tokens": 72081718.0,
      "step": 32280
    },
    {
      "entropy": 5.230281496047974,
      "epoch": 3.6259336216094793,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003737457586042102,
      "loss": 4.8329,
      "mean_token_accuracy": 0.22596203535795212,
      "num_tokens": 72092403.0,
      "step": 32285
    },
    {
      "entropy": 5.2428168773651125,
      "epoch": 3.626495198517437,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003737096813788238,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22697987854480745,
      "num_tokens": 72103711.0,
      "step": 32290
    },
    {
      "entropy": 5.32008204460144,
      "epoch": 3.6270567754253946,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003736736010104829,
      "loss": 4.9566,
      "mean_token_accuracy": 0.22139948159456252,
      "num_tokens": 72115095.0,
      "step": 32295
    },
    {
      "entropy": 5.2884186744689945,
      "epoch": 3.627618352333352,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003736375175003363,
      "loss": 4.9702,
      "mean_token_accuracy": 0.2184678703546524,
      "num_tokens": 72126559.0,
      "step": 32300
    },
    {
      "entropy": 5.267751598358155,
      "epoch": 3.6281799292413095,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00037360143084953285,
      "loss": 4.9324,
      "mean_token_accuracy": 0.22149666845798494,
      "num_tokens": 72137752.0,
      "step": 32305
    },
    {
      "entropy": 5.206169128417969,
      "epoch": 3.628741506149267,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003735653410592217,
      "loss": 4.8137,
      "mean_token_accuracy": 0.22710942327976227,
      "num_tokens": 72148888.0,
      "step": 32310
    },
    {
      "entropy": 5.217497682571411,
      "epoch": 3.629303083057225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037352924813055185,
      "loss": 4.8894,
      "mean_token_accuracy": 0.22916607558727264,
      "num_tokens": 72159607.0,
      "step": 32315
    },
    {
      "entropy": 5.301958274841309,
      "epoch": 3.6298646599651825,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003734931520646724,
      "loss": 4.9244,
      "mean_token_accuracy": 0.22333256006240845,
      "num_tokens": 72171204.0,
      "step": 32320
    },
    {
      "entropy": 5.237124919891357,
      "epoch": 3.6304262368731397,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037345705286273285,
      "loss": 4.9294,
      "mean_token_accuracy": 0.22151149958372116,
      "num_tokens": 72182125.0,
      "step": 32325
    },
    {
      "entropy": 5.243053436279297,
      "epoch": 3.6309878137810974,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003734209505258825,
      "loss": 4.8864,
      "mean_token_accuracy": 0.2231354296207428,
      "num_tokens": 72193093.0,
      "step": 32330
    },
    {
      "entropy": 5.324638414382934,
      "epoch": 3.631549390689055,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00037338484505527103,
      "loss": 4.9872,
      "mean_token_accuracy": 0.21519596129655838,
      "num_tokens": 72204235.0,
      "step": 32335
    },
    {
      "entropy": 5.243725776672363,
      "epoch": 3.6321109675970122,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00037334873645204785,
      "loss": 4.9099,
      "mean_token_accuracy": 0.22345665693283082,
      "num_tokens": 72215160.0,
      "step": 32340
    },
    {
      "entropy": 5.2603325843811035,
      "epoch": 3.63267254450497,
      "grad_norm": 0.953125,
      "learning_rate": 0.00037331262471736285,
      "loss": 4.883,
      "mean_token_accuracy": 0.2254118338227272,
      "num_tokens": 72227517.0,
      "step": 32345
    },
    {
      "entropy": 5.3336272716522215,
      "epoch": 3.6332341214129276,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037327650985236564,
      "loss": 4.9937,
      "mean_token_accuracy": 0.22228700369596482,
      "num_tokens": 72238616.0,
      "step": 32350
    },
    {
      "entropy": 5.330601072311401,
      "epoch": 3.633795698320885,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003732403918582063,
      "loss": 4.9144,
      "mean_token_accuracy": 0.2231555014848709,
      "num_tokens": 72250252.0,
      "step": 32355
    },
    {
      "entropy": 5.311851119995117,
      "epoch": 3.6343572752288424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037320427073603475,
      "loss": 4.9452,
      "mean_token_accuracy": 0.2128622427582741,
      "num_tokens": 72261197.0,
      "step": 32360
    },
    {
      "entropy": 5.199191379547119,
      "epoch": 3.6349188521368,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00037316814648700113,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22469248622655869,
      "num_tokens": 72273838.0,
      "step": 32365
    },
    {
      "entropy": 5.190258026123047,
      "epoch": 3.6354804290447578,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003731320191122557,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23060538470745087,
      "num_tokens": 72285047.0,
      "step": 32370
    },
    {
      "entropy": 5.238623905181885,
      "epoch": 3.6360420059527154,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003730958886129486,
      "loss": 4.9486,
      "mean_token_accuracy": 0.220402330160141,
      "num_tokens": 72295883.0,
      "step": 32375
    },
    {
      "entropy": 5.283397054672241,
      "epoch": 3.6366035828606726,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037305975499023046,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22398523539304732,
      "num_tokens": 72307586.0,
      "step": 32380
    },
    {
      "entropy": 5.227158880233764,
      "epoch": 3.6371651597686303,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0003730236182452517,
      "loss": 4.8441,
      "mean_token_accuracy": 0.22317242324352266,
      "num_tokens": 72318501.0,
      "step": 32385
    },
    {
      "entropy": 5.233042049407959,
      "epoch": 3.637726736676588,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003729874783791629,
      "loss": 4.9185,
      "mean_token_accuracy": 0.22348786741495133,
      "num_tokens": 72330178.0,
      "step": 32390
    },
    {
      "entropy": 5.299942684173584,
      "epoch": 3.638288313584545,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003729513353931148,
      "loss": 4.9846,
      "mean_token_accuracy": 0.22332036197185517,
      "num_tokens": 72340304.0,
      "step": 32395
    },
    {
      "entropy": 5.285574817657471,
      "epoch": 3.638849890492503,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003729151892882581,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22237800657749177,
      "num_tokens": 72350353.0,
      "step": 32400
    },
    {
      "entropy": 5.272806406021118,
      "epoch": 3.6394114674004605,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003728790400657439,
      "loss": 4.849,
      "mean_token_accuracy": 0.23230017125606536,
      "num_tokens": 72361293.0,
      "step": 32405
    },
    {
      "entropy": 5.221298742294311,
      "epoch": 3.639973044308418,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037284288772672315,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23089137077331542,
      "num_tokens": 72372013.0,
      "step": 32410
    },
    {
      "entropy": 5.209573698043823,
      "epoch": 3.640534621216376,
      "grad_norm": 1.0,
      "learning_rate": 0.00037280673227234687,
      "loss": 4.936,
      "mean_token_accuracy": 0.21679614782333373,
      "num_tokens": 72382611.0,
      "step": 32415
    },
    {
      "entropy": 5.234482002258301,
      "epoch": 3.641096198124333,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003727705737037662,
      "loss": 4.8716,
      "mean_token_accuracy": 0.2242620199918747,
      "num_tokens": 72393701.0,
      "step": 32420
    },
    {
      "entropy": 5.402223300933838,
      "epoch": 3.6416577750322907,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003727344120221326,
      "loss": 5.0263,
      "mean_token_accuracy": 0.21180225759744645,
      "num_tokens": 72404168.0,
      "step": 32425
    },
    {
      "entropy": 5.322690677642822,
      "epoch": 3.6422193519402484,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00037269824722859743,
      "loss": 4.9168,
      "mean_token_accuracy": 0.22102581858634948,
      "num_tokens": 72415184.0,
      "step": 32430
    },
    {
      "entropy": 5.232743978500366,
      "epoch": 3.6427809288482056,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003726620793243121,
      "loss": 4.8884,
      "mean_token_accuracy": 0.22275093495845794,
      "num_tokens": 72426189.0,
      "step": 32435
    },
    {
      "entropy": 5.2333451271057125,
      "epoch": 3.6433425057561633,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003726259083104284,
      "loss": 4.9483,
      "mean_token_accuracy": 0.22241219133138657,
      "num_tokens": 72436741.0,
      "step": 32440
    },
    {
      "entropy": 5.329328489303589,
      "epoch": 3.643904082664121,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037258973418809784,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22368508875370025,
      "num_tokens": 72448610.0,
      "step": 32445
    },
    {
      "entropy": 5.232359218597412,
      "epoch": 3.644465659572078,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003725535569584723,
      "loss": 4.9165,
      "mean_token_accuracy": 0.22224762439727783,
      "num_tokens": 72460789.0,
      "step": 32450
    },
    {
      "entropy": 5.220217370986939,
      "epoch": 3.645027236480036,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037251737662270367,
      "loss": 4.9047,
      "mean_token_accuracy": 0.21904105544090272,
      "num_tokens": 72471065.0,
      "step": 32455
    },
    {
      "entropy": 5.239330387115478,
      "epoch": 3.6455888133879935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000372481193181944,
      "loss": 4.8157,
      "mean_token_accuracy": 0.2303273409605026,
      "num_tokens": 72481788.0,
      "step": 32460
    },
    {
      "entropy": 5.259529113769531,
      "epoch": 3.646150390295951,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00037244500663734524,
      "loss": 4.857,
      "mean_token_accuracy": 0.22815678268671036,
      "num_tokens": 72493840.0,
      "step": 32465
    },
    {
      "entropy": 5.201085710525513,
      "epoch": 3.6467119672039088,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003724088169900597,
      "loss": 4.9045,
      "mean_token_accuracy": 0.22039886862039565,
      "num_tokens": 72505002.0,
      "step": 32470
    },
    {
      "entropy": 5.1949443340301515,
      "epoch": 3.647273544111866,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037237262424123975,
      "loss": 4.7987,
      "mean_token_accuracy": 0.22536830902099608,
      "num_tokens": 72515917.0,
      "step": 32475
    },
    {
      "entropy": 5.283746433258057,
      "epoch": 3.6478351210198237,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00037233642839203763,
      "loss": 4.9409,
      "mean_token_accuracy": 0.2206746220588684,
      "num_tokens": 72527500.0,
      "step": 32480
    },
    {
      "entropy": 5.135203170776367,
      "epoch": 3.6483966979277813,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003723002294436059,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22736327350139618,
      "num_tokens": 72538402.0,
      "step": 32485
    },
    {
      "entropy": 5.256746864318847,
      "epoch": 3.6489582748357385,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037226402739709704,
      "loss": 4.9098,
      "mean_token_accuracy": 0.21677203923463823,
      "num_tokens": 72550260.0,
      "step": 32490
    },
    {
      "entropy": 5.35023136138916,
      "epoch": 3.649519851743696,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003722278222536639,
      "loss": 4.964,
      "mean_token_accuracy": 0.2193658247590065,
      "num_tokens": 72562335.0,
      "step": 32495
    },
    {
      "entropy": 5.2102649211883545,
      "epoch": 3.650081428651654,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037219161401445914,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22196228206157684,
      "num_tokens": 72572620.0,
      "step": 32500
    },
    {
      "entropy": 5.198238325119019,
      "epoch": 3.6506430055596115,
      "grad_norm": 0.890625,
      "learning_rate": 0.00037215540268063573,
      "loss": 4.956,
      "mean_token_accuracy": 0.2193707376718521,
      "num_tokens": 72584487.0,
      "step": 32505
    },
    {
      "entropy": 5.279346466064453,
      "epoch": 3.651204582467569,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003721191882533466,
      "loss": 4.9672,
      "mean_token_accuracy": 0.21761613786220552,
      "num_tokens": 72596191.0,
      "step": 32510
    },
    {
      "entropy": 5.274904251098633,
      "epoch": 3.6517661593755264,
      "grad_norm": 1.1875,
      "learning_rate": 0.0003720829707337449,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22006738781929017,
      "num_tokens": 72605901.0,
      "step": 32515
    },
    {
      "entropy": 5.252533054351806,
      "epoch": 3.652327736283484,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003720467501229836,
      "loss": 4.899,
      "mean_token_accuracy": 0.22211642414331437,
      "num_tokens": 72617231.0,
      "step": 32520
    },
    {
      "entropy": 5.361422395706176,
      "epoch": 3.6528893131914417,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003720105264222163,
      "loss": 4.9885,
      "mean_token_accuracy": 0.21658455431461335,
      "num_tokens": 72628777.0,
      "step": 32525
    },
    {
      "entropy": 5.281489562988281,
      "epoch": 3.653450890099399,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00037197429963259615,
      "loss": 4.9248,
      "mean_token_accuracy": 0.22237463146448136,
      "num_tokens": 72639670.0,
      "step": 32530
    },
    {
      "entropy": 5.202487850189209,
      "epoch": 3.6540124670073566,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003719380697552767,
      "loss": 4.7865,
      "mean_token_accuracy": 0.22992908358573913,
      "num_tokens": 72651010.0,
      "step": 32535
    },
    {
      "entropy": 5.37108154296875,
      "epoch": 3.6545740439153143,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003719018367914115,
      "loss": 5.0061,
      "mean_token_accuracy": 0.21413020938634872,
      "num_tokens": 72662900.0,
      "step": 32540
    },
    {
      "entropy": 5.297215938568115,
      "epoch": 3.6551356208232715,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003718656007421542,
      "loss": 4.9891,
      "mean_token_accuracy": 0.2232266843318939,
      "num_tokens": 72674348.0,
      "step": 32545
    },
    {
      "entropy": 5.261736154556274,
      "epoch": 3.655697197731229,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037182936160865866,
      "loss": 4.8135,
      "mean_token_accuracy": 0.2250980779528618,
      "num_tokens": 72685771.0,
      "step": 32550
    },
    {
      "entropy": 5.233519983291626,
      "epoch": 3.656258774639187,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003717931193920788,
      "loss": 4.8272,
      "mean_token_accuracy": 0.22491241097450257,
      "num_tokens": 72697121.0,
      "step": 32555
    },
    {
      "entropy": 5.167683410644531,
      "epoch": 3.6568203515471445,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003717568740935683,
      "loss": 4.874,
      "mean_token_accuracy": 0.22885120362043382,
      "num_tokens": 72707245.0,
      "step": 32560
    },
    {
      "entropy": 5.223340177536011,
      "epoch": 3.657381928455102,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003717206257142815,
      "loss": 4.8452,
      "mean_token_accuracy": 0.22742449641227722,
      "num_tokens": 72718328.0,
      "step": 32565
    },
    {
      "entropy": 5.308238983154297,
      "epoch": 3.6579435053630593,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003716843742553724,
      "loss": 4.9701,
      "mean_token_accuracy": 0.2184201255440712,
      "num_tokens": 72729192.0,
      "step": 32570
    },
    {
      "entropy": 5.223246002197266,
      "epoch": 3.658505082271017,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003716481197179953,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2151833087205887,
      "num_tokens": 72740213.0,
      "step": 32575
    },
    {
      "entropy": 5.206720018386841,
      "epoch": 3.6590666591789747,
      "grad_norm": 0.9375,
      "learning_rate": 0.00037161186210330464,
      "loss": 4.7941,
      "mean_token_accuracy": 0.2265511840581894,
      "num_tokens": 72752477.0,
      "step": 32580
    },
    {
      "entropy": 5.320250225067139,
      "epoch": 3.659628236086932,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00037157560141245485,
      "loss": 4.8717,
      "mean_token_accuracy": 0.23088551163673401,
      "num_tokens": 72763926.0,
      "step": 32585
    },
    {
      "entropy": 5.283384656906128,
      "epoch": 3.6601898129948895,
      "grad_norm": 1.046875,
      "learning_rate": 0.00037153933764660034,
      "loss": 4.9155,
      "mean_token_accuracy": 0.22208296209573747,
      "num_tokens": 72774922.0,
      "step": 32590
    },
    {
      "entropy": 5.190053176879883,
      "epoch": 3.660751389902847,
      "grad_norm": 1.015625,
      "learning_rate": 0.000371503070806896,
      "loss": 4.8542,
      "mean_token_accuracy": 0.22451552748680115,
      "num_tokens": 72786362.0,
      "step": 32595
    },
    {
      "entropy": 5.2451272964477536,
      "epoch": 3.661312966810805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00037146680089449643,
      "loss": 4.8989,
      "mean_token_accuracy": 0.22244699895381928,
      "num_tokens": 72798337.0,
      "step": 32600
    },
    {
      "entropy": 5.282134532928467,
      "epoch": 3.6618745437187625,
      "grad_norm": 1.015625,
      "learning_rate": 0.00037143052791055654,
      "loss": 4.8138,
      "mean_token_accuracy": 0.22767258584499359,
      "num_tokens": 72808923.0,
      "step": 32605
    },
    {
      "entropy": 5.18156852722168,
      "epoch": 3.6624361206267197,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003713942518562311,
      "loss": 4.8452,
      "mean_token_accuracy": 0.2253683477640152,
      "num_tokens": 72820148.0,
      "step": 32610
    },
    {
      "entropy": 5.242206287384033,
      "epoch": 3.6629976975346774,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003713579727326754,
      "loss": 4.9283,
      "mean_token_accuracy": 0.2257738471031189,
      "num_tokens": 72830412.0,
      "step": 32615
    },
    {
      "entropy": 5.275548458099365,
      "epoch": 3.663559274442635,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037132169054104446,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23247001320123672,
      "num_tokens": 72840857.0,
      "step": 32620
    },
    {
      "entropy": 5.197249555587769,
      "epoch": 3.6641208513505923,
      "grad_norm": 1.0,
      "learning_rate": 0.00037128540528249355,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22283941507339478,
      "num_tokens": 72852269.0,
      "step": 32625
    },
    {
      "entropy": 5.1912281036376955,
      "epoch": 3.66468242825855,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000371249116958178,
      "loss": 4.9427,
      "mean_token_accuracy": 0.2204911917448044,
      "num_tokens": 72863536.0,
      "step": 32630
    },
    {
      "entropy": 5.2075573921203615,
      "epoch": 3.6652440051665076,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00037121282556925326,
      "loss": 4.896,
      "mean_token_accuracy": 0.22924120724201202,
      "num_tokens": 72873835.0,
      "step": 32635
    },
    {
      "entropy": 5.292690753936768,
      "epoch": 3.665805582074465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003711765311168748,
      "loss": 4.7876,
      "mean_token_accuracy": 0.22511350810527803,
      "num_tokens": 72883916.0,
      "step": 32640
    },
    {
      "entropy": 5.236888837814331,
      "epoch": 3.6663671589824225,
      "grad_norm": 1.09375,
      "learning_rate": 0.00037114023360219824,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22088168412446976,
      "num_tokens": 72895928.0,
      "step": 32645
    },
    {
      "entropy": 5.2171694278717045,
      "epoch": 3.66692873589038,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00037110393302637937,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22088685333728791,
      "num_tokens": 72908138.0,
      "step": 32650
    },
    {
      "entropy": 5.258910036087036,
      "epoch": 3.667490312798338,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00037106762939057404,
      "loss": 4.8912,
      "mean_token_accuracy": 0.2243919253349304,
      "num_tokens": 72918326.0,
      "step": 32655
    },
    {
      "entropy": 5.294994974136353,
      "epoch": 3.6680518897062955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003710313226959381,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22296328842639923,
      "num_tokens": 72929140.0,
      "step": 32660
    },
    {
      "entropy": 5.276230096817017,
      "epoch": 3.6686134666142527,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003709950129436276,
      "loss": 4.9123,
      "mean_token_accuracy": 0.23183191567659378,
      "num_tokens": 72941355.0,
      "step": 32665
    },
    {
      "entropy": 5.293270492553711,
      "epoch": 3.6691750435222104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003709587001347987,
      "loss": 4.933,
      "mean_token_accuracy": 0.2190885215997696,
      "num_tokens": 72951841.0,
      "step": 32670
    },
    {
      "entropy": 5.178249502182007,
      "epoch": 3.669736620430168,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00037092238427060756,
      "loss": 4.854,
      "mean_token_accuracy": 0.21870125830173492,
      "num_tokens": 72963843.0,
      "step": 32675
    },
    {
      "entropy": 5.260785961151123,
      "epoch": 3.6702981973381252,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003708860653522104,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2272256061434746,
      "num_tokens": 72973709.0,
      "step": 32680
    },
    {
      "entropy": 5.171728992462159,
      "epoch": 3.670859774246083,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003708497433807639,
      "loss": 4.7126,
      "mean_token_accuracy": 0.24068481475114822,
      "num_tokens": 72984658.0,
      "step": 32685
    },
    {
      "entropy": 5.187235116958618,
      "epoch": 3.6714213511540406,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00037081341835742433,
      "loss": 4.9446,
      "mean_token_accuracy": 0.22017366588115692,
      "num_tokens": 72995745.0,
      "step": 32690
    },
    {
      "entropy": 5.24003791809082,
      "epoch": 3.671982928061998,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003707770902833483,
      "loss": 4.9435,
      "mean_token_accuracy": 0.21896506249904632,
      "num_tokens": 73007619.0,
      "step": 32695
    },
    {
      "entropy": 5.2695403575897215,
      "epoch": 3.672544504969956,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003707407591596926,
      "loss": 4.7991,
      "mean_token_accuracy": 0.22673967480659485,
      "num_tokens": 73017779.0,
      "step": 32700
    },
    {
      "entropy": 5.250861167907715,
      "epoch": 3.673106081877913,
      "grad_norm": 1.046875,
      "learning_rate": 0.000370704424987614,
      "loss": 4.8976,
      "mean_token_accuracy": 0.225007501244545,
      "num_tokens": 73028246.0,
      "step": 32705
    },
    {
      "entropy": 5.259636545181275,
      "epoch": 3.6736676587858708,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003706680877682694,
      "loss": 4.9555,
      "mean_token_accuracy": 0.20951247960329056,
      "num_tokens": 73038896.0,
      "step": 32710
    },
    {
      "entropy": 5.2902915477752686,
      "epoch": 3.6742292356938284,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00037063174750281573,
      "loss": 4.9714,
      "mean_token_accuracy": 0.21572768539190293,
      "num_tokens": 73050960.0,
      "step": 32715
    },
    {
      "entropy": 5.338661861419678,
      "epoch": 3.6747908126017856,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003705954041924102,
      "loss": 4.8547,
      "mean_token_accuracy": 0.2297824427485466,
      "num_tokens": 73061537.0,
      "step": 32720
    },
    {
      "entropy": 5.2872344017028805,
      "epoch": 3.6753523895097433,
      "grad_norm": 1.03125,
      "learning_rate": 0.00037055905783821,
      "loss": 4.9849,
      "mean_token_accuracy": 0.21899515390396118,
      "num_tokens": 73072357.0,
      "step": 32725
    },
    {
      "entropy": 5.220153093338013,
      "epoch": 3.675913966417701,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00037052270844137216,
      "loss": 4.8211,
      "mean_token_accuracy": 0.2314888596534729,
      "num_tokens": 73083527.0,
      "step": 32730
    },
    {
      "entropy": 5.20243239402771,
      "epoch": 3.676475543325658,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003704863560030543,
      "loss": 4.806,
      "mean_token_accuracy": 0.22536287009716033,
      "num_tokens": 73095068.0,
      "step": 32735
    },
    {
      "entropy": 5.155792188644409,
      "epoch": 3.677037120233616,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003704500005244139,
      "loss": 4.7898,
      "mean_token_accuracy": 0.22587635666131972,
      "num_tokens": 73106318.0,
      "step": 32740
    },
    {
      "entropy": 5.2412073612213135,
      "epoch": 3.6775986971415735,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00037041364200660835,
      "loss": 4.9149,
      "mean_token_accuracy": 0.22318706214427947,
      "num_tokens": 73118433.0,
      "step": 32745
    },
    {
      "entropy": 5.297665739059449,
      "epoch": 3.678160274049531,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00037037728045079553,
      "loss": 4.9101,
      "mean_token_accuracy": 0.2201499953866005,
      "num_tokens": 73129887.0,
      "step": 32750
    },
    {
      "entropy": 5.16417555809021,
      "epoch": 3.678721850957489,
      "grad_norm": 0.96875,
      "learning_rate": 0.000370340915858133,
      "loss": 4.8275,
      "mean_token_accuracy": 0.22460512965917587,
      "num_tokens": 73141529.0,
      "step": 32755
    },
    {
      "entropy": 5.212078189849853,
      "epoch": 3.679283427865446,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003703045482297788,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2259417727589607,
      "num_tokens": 73151317.0,
      "step": 32760
    },
    {
      "entropy": 5.14879994392395,
      "epoch": 3.6798450047734037,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003702681775668908,
      "loss": 4.7527,
      "mean_token_accuracy": 0.2281192809343338,
      "num_tokens": 73162382.0,
      "step": 32765
    },
    {
      "entropy": 5.322485446929932,
      "epoch": 3.6804065816813614,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003702318038706272,
      "loss": 4.983,
      "mean_token_accuracy": 0.2204865872859955,
      "num_tokens": 73173142.0,
      "step": 32770
    },
    {
      "entropy": 5.238940048217773,
      "epoch": 3.6809681585893186,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003701954271421459,
      "loss": 4.8931,
      "mean_token_accuracy": 0.221041177213192,
      "num_tokens": 73185074.0,
      "step": 32775
    },
    {
      "entropy": 5.313161468505859,
      "epoch": 3.6815297354972762,
      "grad_norm": 1.0625,
      "learning_rate": 0.00037015904738260533,
      "loss": 4.915,
      "mean_token_accuracy": 0.22787125110626222,
      "num_tokens": 73195806.0,
      "step": 32780
    },
    {
      "entropy": 5.362853527069092,
      "epoch": 3.682091312405234,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003701226645931637,
      "loss": 5.0311,
      "mean_token_accuracy": 0.2072199910879135,
      "num_tokens": 73207213.0,
      "step": 32785
    },
    {
      "entropy": 5.245966577529908,
      "epoch": 3.6826528893131916,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003700862787749796,
      "loss": 4.9039,
      "mean_token_accuracy": 0.22192890495061873,
      "num_tokens": 73219715.0,
      "step": 32790
    },
    {
      "entropy": 5.229845333099365,
      "epoch": 3.6832144662211492,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003700498899292115,
      "loss": 4.8573,
      "mean_token_accuracy": 0.22509486228227615,
      "num_tokens": 73229868.0,
      "step": 32795
    },
    {
      "entropy": 5.241810083389282,
      "epoch": 3.6837760431291064,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003700134980570181,
      "loss": 4.8569,
      "mean_token_accuracy": 0.22464149743318557,
      "num_tokens": 73240347.0,
      "step": 32800
    },
    {
      "entropy": 5.272997760772705,
      "epoch": 3.684337620037064,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003699771031595579,
      "loss": 4.8706,
      "mean_token_accuracy": 0.2273360326886177,
      "num_tokens": 73250165.0,
      "step": 32805
    },
    {
      "entropy": 5.285821104049683,
      "epoch": 3.6848991969450218,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036994070523799005,
      "loss": 4.8799,
      "mean_token_accuracy": 0.2230353370308876,
      "num_tokens": 73261320.0,
      "step": 32810
    },
    {
      "entropy": 5.248336982727051,
      "epoch": 3.685460773852979,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036990430429347334,
      "loss": 4.9563,
      "mean_token_accuracy": 0.22150513678789138,
      "num_tokens": 73272342.0,
      "step": 32815
    },
    {
      "entropy": 5.3099769115447994,
      "epoch": 3.6860223507609367,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036986790032716673,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22816501259803773,
      "num_tokens": 73283790.0,
      "step": 32820
    },
    {
      "entropy": 5.218863821029663,
      "epoch": 3.6865839276688943,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003698314933402292,
      "loss": 4.899,
      "mean_token_accuracy": 0.22446243166923524,
      "num_tokens": 73295869.0,
      "step": 32825
    },
    {
      "entropy": 5.1685772895812985,
      "epoch": 3.6871455045768515,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003697950833338204,
      "loss": 4.871,
      "mean_token_accuracy": 0.219385327398777,
      "num_tokens": 73307518.0,
      "step": 32830
    },
    {
      "entropy": 5.26043734550476,
      "epoch": 3.687707081484809,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036975867030909933,
      "loss": 4.9028,
      "mean_token_accuracy": 0.22212535440921782,
      "num_tokens": 73318419.0,
      "step": 32835
    },
    {
      "entropy": 5.252279996871948,
      "epoch": 3.688268658392767,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036972225426722545,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2228752091526985,
      "num_tokens": 73328094.0,
      "step": 32840
    },
    {
      "entropy": 5.24697961807251,
      "epoch": 3.6888302353007245,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003696858352093582,
      "loss": 4.8919,
      "mean_token_accuracy": 0.2319771409034729,
      "num_tokens": 73339156.0,
      "step": 32845
    },
    {
      "entropy": 5.266514730453491,
      "epoch": 3.689391812208682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036964941313665715,
      "loss": 4.7927,
      "mean_token_accuracy": 0.2291173979640007,
      "num_tokens": 73350740.0,
      "step": 32850
    },
    {
      "entropy": 5.241711759567261,
      "epoch": 3.6899533891166394,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003696129880502823,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22512691766023635,
      "num_tokens": 73360745.0,
      "step": 32855
    },
    {
      "entropy": 5.286617565155029,
      "epoch": 3.690514966024597,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000369576559951393,
      "loss": 4.9269,
      "mean_token_accuracy": 0.2238624334335327,
      "num_tokens": 73372580.0,
      "step": 32860
    },
    {
      "entropy": 5.250747966766357,
      "epoch": 3.6910765429325547,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036954012884114946,
      "loss": 4.8899,
      "mean_token_accuracy": 0.2244539424777031,
      "num_tokens": 73383970.0,
      "step": 32865
    },
    {
      "entropy": 5.211357688903808,
      "epoch": 3.691638119840512,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003695036947207115,
      "loss": 4.8393,
      "mean_token_accuracy": 0.2321056082844734,
      "num_tokens": 73394472.0,
      "step": 32870
    },
    {
      "entropy": 5.258908605575561,
      "epoch": 3.6921996967484696,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036946725759123936,
      "loss": 4.9315,
      "mean_token_accuracy": 0.21574908047914504,
      "num_tokens": 73405415.0,
      "step": 32875
    },
    {
      "entropy": 5.3672511100769045,
      "epoch": 3.6927612736564273,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003694308174538929,
      "loss": 4.8898,
      "mean_token_accuracy": 0.21964219063520432,
      "num_tokens": 73416583.0,
      "step": 32880
    },
    {
      "entropy": 5.339581823348999,
      "epoch": 3.693322850564385,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036939437430983274,
      "loss": 4.9804,
      "mean_token_accuracy": 0.2201366201043129,
      "num_tokens": 73427408.0,
      "step": 32885
    },
    {
      "entropy": 5.185599994659424,
      "epoch": 3.6938844274723426,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036935792816021905,
      "loss": 4.8493,
      "mean_token_accuracy": 0.23006113320589067,
      "num_tokens": 73438509.0,
      "step": 32890
    },
    {
      "entropy": 5.230399322509766,
      "epoch": 3.6944460043803,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036932147900621234,
      "loss": 4.8703,
      "mean_token_accuracy": 0.2274269863963127,
      "num_tokens": 73449197.0,
      "step": 32895
    },
    {
      "entropy": 5.305664205551148,
      "epoch": 3.6950075812882575,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00036928502684897314,
      "loss": 4.9517,
      "mean_token_accuracy": 0.21571084707975388,
      "num_tokens": 73460777.0,
      "step": 32900
    },
    {
      "entropy": 5.359545040130615,
      "epoch": 3.695569158196215,
      "grad_norm": 1.0,
      "learning_rate": 0.00036924857168966216,
      "loss": 4.94,
      "mean_token_accuracy": 0.21557899415493012,
      "num_tokens": 73471747.0,
      "step": 32905
    },
    {
      "entropy": 5.1399843215942385,
      "epoch": 3.6961307351041723,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003692121135294401,
      "loss": 4.7294,
      "mean_token_accuracy": 0.23356136232614516,
      "num_tokens": 73481258.0,
      "step": 32910
    },
    {
      "entropy": 5.207405853271484,
      "epoch": 3.69669231201213,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00036917565236946776,
      "loss": 4.8734,
      "mean_token_accuracy": 0.2275660142302513,
      "num_tokens": 73492073.0,
      "step": 32915
    },
    {
      "entropy": 5.204788732528686,
      "epoch": 3.6972538889200877,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036913918821090616,
      "loss": 4.8985,
      "mean_token_accuracy": 0.22253136187791825,
      "num_tokens": 73502813.0,
      "step": 32920
    },
    {
      "entropy": 5.3001110553741455,
      "epoch": 3.697815465828045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00036910272105491625,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2242044597864151,
      "num_tokens": 73514417.0,
      "step": 32925
    },
    {
      "entropy": 5.2805956363677975,
      "epoch": 3.6983770427360025,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036906625090265925,
      "loss": 4.8426,
      "mean_token_accuracy": 0.22941878885030748,
      "num_tokens": 73525636.0,
      "step": 32930
    },
    {
      "entropy": 5.227637100219726,
      "epoch": 3.69893861964396,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036902977775529636,
      "loss": 4.8953,
      "mean_token_accuracy": 0.2249802067875862,
      "num_tokens": 73536557.0,
      "step": 32935
    },
    {
      "entropy": 5.204141616821289,
      "epoch": 3.699500196551918,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003689933016139888,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23623855262994767,
      "num_tokens": 73547910.0,
      "step": 32940
    },
    {
      "entropy": 5.138375854492187,
      "epoch": 3.7000617734598755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003689568224798981,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23324548751115798,
      "num_tokens": 73557583.0,
      "step": 32945
    },
    {
      "entropy": 5.256542825698853,
      "epoch": 3.7006233503678327,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003689203403541857,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2242407977581024,
      "num_tokens": 73567300.0,
      "step": 32950
    },
    {
      "entropy": 5.307569360733032,
      "epoch": 3.7011849272757904,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036888385523801325,
      "loss": 4.9566,
      "mean_token_accuracy": 0.21813049167394638,
      "num_tokens": 73577387.0,
      "step": 32955
    },
    {
      "entropy": 5.341034460067749,
      "epoch": 3.701746504183748,
      "grad_norm": 1.0,
      "learning_rate": 0.00036884736713254247,
      "loss": 4.9469,
      "mean_token_accuracy": 0.22119967639446259,
      "num_tokens": 73588340.0,
      "step": 32960
    },
    {
      "entropy": 5.310978841781616,
      "epoch": 3.7023080810917053,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00036881087603893515,
      "loss": 4.913,
      "mean_token_accuracy": 0.22563314586877822,
      "num_tokens": 73599650.0,
      "step": 32965
    },
    {
      "entropy": 5.3381401062011715,
      "epoch": 3.702869657999663,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036877438195835307,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22517788410186768,
      "num_tokens": 73611279.0,
      "step": 32970
    },
    {
      "entropy": 5.170384788513184,
      "epoch": 3.7034312349076206,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003687378848919584,
      "loss": 4.8869,
      "mean_token_accuracy": 0.22637786120176315,
      "num_tokens": 73623279.0,
      "step": 32975
    },
    {
      "entropy": 5.237657880783081,
      "epoch": 3.7039928118155783,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036870138484091297,
      "loss": 4.9564,
      "mean_token_accuracy": 0.2169398173689842,
      "num_tokens": 73633756.0,
      "step": 32980
    },
    {
      "entropy": 5.37052960395813,
      "epoch": 3.704554388723536,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036866488180637926,
      "loss": 5.0189,
      "mean_token_accuracy": 0.2113729327917099,
      "num_tokens": 73646033.0,
      "step": 32985
    },
    {
      "entropy": 5.265288591384888,
      "epoch": 3.705115965631493,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003686283757895193,
      "loss": 4.9043,
      "mean_token_accuracy": 0.2197069764137268,
      "num_tokens": 73655833.0,
      "step": 32990
    },
    {
      "entropy": 5.3440916538238525,
      "epoch": 3.705677542539451,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036859186679149555,
      "loss": 4.9746,
      "mean_token_accuracy": 0.2166818603873253,
      "num_tokens": 73667755.0,
      "step": 32995
    },
    {
      "entropy": 5.36148362159729,
      "epoch": 3.7062391194474085,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003685553548134705,
      "loss": 4.9972,
      "mean_token_accuracy": 0.22115977108478546,
      "num_tokens": 73680537.0,
      "step": 33000
    },
    {
      "epoch": 3.7062391194474085,
      "eval_entropy": 5.077532782516499,
      "eval_loss": 5.060890197753906,
      "eval_mean_token_accuracy": 0.222233716512821,
      "eval_num_tokens": 73680537.0,
      "eval_runtime": 23.9223,
      "eval_samples_per_second": 1296.404,
      "eval_steps_per_second": 162.066,
      "step": 33000
    },
    {
      "entropy": 5.231734657287598,
      "epoch": 3.7068006963553657,
      "grad_norm": 1.0,
      "learning_rate": 0.0003685188398566066,
      "loss": 4.8592,
      "mean_token_accuracy": 0.23170552849769593,
      "num_tokens": 73692446.0,
      "step": 33005
    },
    {
      "entropy": 5.248845911026001,
      "epoch": 3.7073622732633233,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003684823219220666,
      "loss": 4.9081,
      "mean_token_accuracy": 0.21779075115919114,
      "num_tokens": 73704059.0,
      "step": 33010
    },
    {
      "entropy": 5.229152393341065,
      "epoch": 3.707923850171281,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003684458010110133,
      "loss": 4.8743,
      "mean_token_accuracy": 0.22356147617101668,
      "num_tokens": 73715234.0,
      "step": 33015
    },
    {
      "entropy": 5.314599180221558,
      "epoch": 3.7084854270792382,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003684092771246094,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21648112684488297,
      "num_tokens": 73726930.0,
      "step": 33020
    },
    {
      "entropy": 5.287427806854248,
      "epoch": 3.709047003987196,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003683727502640178,
      "loss": 4.9032,
      "mean_token_accuracy": 0.21973323076963425,
      "num_tokens": 73738107.0,
      "step": 33025
    },
    {
      "entropy": 5.308217096328735,
      "epoch": 3.7096085808951536,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003683362204304017,
      "loss": 5.0114,
      "mean_token_accuracy": 0.21639705300331116,
      "num_tokens": 73749165.0,
      "step": 33030
    },
    {
      "entropy": 5.282676792144775,
      "epoch": 3.710170157803111,
      "grad_norm": 1.0,
      "learning_rate": 0.0003682996876249241,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23056585490703582,
      "num_tokens": 73760921.0,
      "step": 33035
    },
    {
      "entropy": 5.2220100402832035,
      "epoch": 3.710731734711069,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003682631518487483,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22313523888587952,
      "num_tokens": 73772045.0,
      "step": 33040
    },
    {
      "entropy": 5.232369327545166,
      "epoch": 3.711293311619026,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003682266131030375,
      "loss": 4.8407,
      "mean_token_accuracy": 0.22814140766859053,
      "num_tokens": 73782698.0,
      "step": 33045
    },
    {
      "entropy": 5.159593296051026,
      "epoch": 3.7118548885269838,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036819007138895525,
      "loss": 4.8259,
      "mean_token_accuracy": 0.22877446115016936,
      "num_tokens": 73793758.0,
      "step": 33050
    },
    {
      "entropy": 5.289063596725464,
      "epoch": 3.7124164654349414,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000368153526707665,
      "loss": 4.9389,
      "mean_token_accuracy": 0.21455997824668885,
      "num_tokens": 73805479.0,
      "step": 33055
    },
    {
      "entropy": 5.293049097061157,
      "epoch": 3.7129780423428986,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003681169790603303,
      "loss": 4.9085,
      "mean_token_accuracy": 0.21763805598020552,
      "num_tokens": 73817361.0,
      "step": 33060
    },
    {
      "entropy": 5.263939905166626,
      "epoch": 3.7135396192508563,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036808042844811484,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22155203372240068,
      "num_tokens": 73829651.0,
      "step": 33065
    },
    {
      "entropy": 5.1980735778808596,
      "epoch": 3.714101196158814,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036804387487218254,
      "loss": 4.803,
      "mean_token_accuracy": 0.22877197712659836,
      "num_tokens": 73839608.0,
      "step": 33070
    },
    {
      "entropy": 5.230276441574096,
      "epoch": 3.7146627730667716,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000368007318333697,
      "loss": 4.8807,
      "mean_token_accuracy": 0.2250270128250122,
      "num_tokens": 73850663.0,
      "step": 33075
    },
    {
      "entropy": 5.304261779785156,
      "epoch": 3.7152243499747293,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003679707588338225,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2237917199730873,
      "num_tokens": 73862441.0,
      "step": 33080
    },
    {
      "entropy": 5.286645698547363,
      "epoch": 3.7157859268826865,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000367934196373723,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22947108894586563,
      "num_tokens": 73873549.0,
      "step": 33085
    },
    {
      "entropy": 5.227525138854981,
      "epoch": 3.716347503790644,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00036789763095456264,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22534943222999573,
      "num_tokens": 73884872.0,
      "step": 33090
    },
    {
      "entropy": 5.340947389602661,
      "epoch": 3.716909080698602,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003678610625775056,
      "loss": 5.0679,
      "mean_token_accuracy": 0.21158191710710525,
      "num_tokens": 73896137.0,
      "step": 33095
    },
    {
      "entropy": 5.2032311916351315,
      "epoch": 3.717470657606559,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036782449124371643,
      "loss": 4.7883,
      "mean_token_accuracy": 0.22548956125974656,
      "num_tokens": 73906135.0,
      "step": 33100
    },
    {
      "entropy": 5.290094661712646,
      "epoch": 3.7180322345145167,
      "grad_norm": 0.875,
      "learning_rate": 0.00036778791695435935,
      "loss": 4.904,
      "mean_token_accuracy": 0.22415556758642197,
      "num_tokens": 73916856.0,
      "step": 33105
    },
    {
      "entropy": 5.135684204101563,
      "epoch": 3.7185938114224744,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036775133971059904,
      "loss": 4.821,
      "mean_token_accuracy": 0.23302647918462754,
      "num_tokens": 73928268.0,
      "step": 33110
    },
    {
      "entropy": 5.3202800273895265,
      "epoch": 3.7191553883304316,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003677147595136001,
      "loss": 4.9627,
      "mean_token_accuracy": 0.22269116938114167,
      "num_tokens": 73940327.0,
      "step": 33115
    },
    {
      "entropy": 5.369922637939453,
      "epoch": 3.7197169652383892,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036767817636452723,
      "loss": 4.918,
      "mean_token_accuracy": 0.2277526244521141,
      "num_tokens": 73951488.0,
      "step": 33120
    },
    {
      "entropy": 5.221787548065185,
      "epoch": 3.720278542146347,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036764159026454534,
      "loss": 4.8531,
      "mean_token_accuracy": 0.23145677298307418,
      "num_tokens": 73963479.0,
      "step": 33125
    },
    {
      "entropy": 5.117607307434082,
      "epoch": 3.7208401190543046,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003676050012148193,
      "loss": 4.7814,
      "mean_token_accuracy": 0.22397050708532334,
      "num_tokens": 73974878.0,
      "step": 33130
    },
    {
      "entropy": 5.221049690246582,
      "epoch": 3.7214016959622622,
      "grad_norm": 1.125,
      "learning_rate": 0.000367568409216514,
      "loss": 4.9206,
      "mean_token_accuracy": 0.22548263370990754,
      "num_tokens": 73985983.0,
      "step": 33135
    },
    {
      "entropy": 5.281276655197144,
      "epoch": 3.7219632728702194,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003675318142707947,
      "loss": 4.935,
      "mean_token_accuracy": 0.2192873850464821,
      "num_tokens": 73997141.0,
      "step": 33140
    },
    {
      "entropy": 5.291124677658081,
      "epoch": 3.722524849778177,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003674952163788266,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22873612195253373,
      "num_tokens": 74008262.0,
      "step": 33145
    },
    {
      "entropy": 5.270225906372071,
      "epoch": 3.7230864266861348,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036745861554177486,
      "loss": 4.9022,
      "mean_token_accuracy": 0.22245410531759263,
      "num_tokens": 74020465.0,
      "step": 33150
    },
    {
      "entropy": 5.268958520889282,
      "epoch": 3.723648003594092,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036742201176080496,
      "loss": 4.8887,
      "mean_token_accuracy": 0.2281578525900841,
      "num_tokens": 74031100.0,
      "step": 33155
    },
    {
      "entropy": 5.279617261886597,
      "epoch": 3.7242095805020496,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003673854050370824,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21450054347515107,
      "num_tokens": 74043313.0,
      "step": 33160
    },
    {
      "entropy": 5.2365562438964846,
      "epoch": 3.7247711574100073,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036734879537177276,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22311893701553345,
      "num_tokens": 74054540.0,
      "step": 33165
    },
    {
      "entropy": 5.331911993026734,
      "epoch": 3.725332734317965,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003673121827660415,
      "loss": 5.0569,
      "mean_token_accuracy": 0.21454151570796967,
      "num_tokens": 74065636.0,
      "step": 33170
    },
    {
      "entropy": 5.217788648605347,
      "epoch": 3.7258943112259226,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036727556722105465,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22161523699760438,
      "num_tokens": 74076516.0,
      "step": 33175
    },
    {
      "entropy": 5.2764294147491455,
      "epoch": 3.72645588813388,
      "grad_norm": 1.0,
      "learning_rate": 0.00036723894873797786,
      "loss": 4.9238,
      "mean_token_accuracy": 0.22693295031785965,
      "num_tokens": 74087481.0,
      "step": 33180
    },
    {
      "entropy": 5.3683257579803465,
      "epoch": 3.7270174650418375,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036720232731797726,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22563147246837617,
      "num_tokens": 74098577.0,
      "step": 33185
    },
    {
      "entropy": 5.291546869277954,
      "epoch": 3.727579041949795,
      "grad_norm": 1.0625,
      "learning_rate": 0.00036716570296221873,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21719274967908858,
      "num_tokens": 74108813.0,
      "step": 33190
    },
    {
      "entropy": 5.179608917236328,
      "epoch": 3.7281406188577524,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036712907567186857,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2231182262301445,
      "num_tokens": 74118937.0,
      "step": 33195
    },
    {
      "entropy": 5.279478263854981,
      "epoch": 3.72870219576571,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003670924454480928,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22471064776182176,
      "num_tokens": 74129879.0,
      "step": 33200
    },
    {
      "entropy": 5.2521039009094235,
      "epoch": 3.7292637726736677,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003670558122920579,
      "loss": 4.8016,
      "mean_token_accuracy": 0.22353991270065307,
      "num_tokens": 74140455.0,
      "step": 33205
    },
    {
      "entropy": 5.261266994476318,
      "epoch": 3.729825349581625,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036701917620493026,
      "loss": 4.9456,
      "mean_token_accuracy": 0.21997007131576538,
      "num_tokens": 74152508.0,
      "step": 33210
    },
    {
      "entropy": 5.2712071418762205,
      "epoch": 3.7303869264895826,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003669825371878764,
      "loss": 4.8223,
      "mean_token_accuracy": 0.22912129759788513,
      "num_tokens": 74163700.0,
      "step": 33215
    },
    {
      "entropy": 5.205265665054322,
      "epoch": 3.7309485033975403,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036694589524206275,
      "loss": 4.8905,
      "mean_token_accuracy": 0.2274702250957489,
      "num_tokens": 74174715.0,
      "step": 33220
    },
    {
      "entropy": 5.294836568832397,
      "epoch": 3.731510080305498,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003669092503686562,
      "loss": 4.9565,
      "mean_token_accuracy": 0.2209405079483986,
      "num_tokens": 74186679.0,
      "step": 33225
    },
    {
      "entropy": 5.310052967071533,
      "epoch": 3.7320716572134556,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036687260256882347,
      "loss": 4.884,
      "mean_token_accuracy": 0.22508785575628282,
      "num_tokens": 74196382.0,
      "step": 33230
    },
    {
      "entropy": 5.300515794754029,
      "epoch": 3.732633234121413,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036683595184373147,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22365398406982423,
      "num_tokens": 74207587.0,
      "step": 33235
    },
    {
      "entropy": 5.216281175613403,
      "epoch": 3.7331948110293705,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003667992981945471,
      "loss": 4.8656,
      "mean_token_accuracy": 0.22431963086128234,
      "num_tokens": 74218826.0,
      "step": 33240
    },
    {
      "entropy": 5.34842209815979,
      "epoch": 3.733756387937328,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036676264162243744,
      "loss": 4.9859,
      "mean_token_accuracy": 0.2210303783416748,
      "num_tokens": 74230450.0,
      "step": 33245
    },
    {
      "entropy": 5.294094085693359,
      "epoch": 3.7343179648452853,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003667259821285698,
      "loss": 4.9327,
      "mean_token_accuracy": 0.2164280667901039,
      "num_tokens": 74241438.0,
      "step": 33250
    },
    {
      "entropy": 5.347292280197143,
      "epoch": 3.734879541753243,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003666893197141112,
      "loss": 4.984,
      "mean_token_accuracy": 0.21764660477638245,
      "num_tokens": 74252460.0,
      "step": 33255
    },
    {
      "entropy": 5.286917686462402,
      "epoch": 3.7354411186612007,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003666526543802291,
      "loss": 4.8486,
      "mean_token_accuracy": 0.22693878412246704,
      "num_tokens": 74262984.0,
      "step": 33260
    },
    {
      "entropy": 5.260368776321411,
      "epoch": 3.7360026955691583,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000366615986128091,
      "loss": 4.9908,
      "mean_token_accuracy": 0.21719557791948318,
      "num_tokens": 74274368.0,
      "step": 33265
    },
    {
      "entropy": 5.21182107925415,
      "epoch": 3.736564272477116,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003665793149588644,
      "loss": 4.8554,
      "mean_token_accuracy": 0.23136204928159715,
      "num_tokens": 74286394.0,
      "step": 33270
    },
    {
      "entropy": 5.258127975463867,
      "epoch": 3.737125849385073,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036654264087371686,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2263406291604042,
      "num_tokens": 74296803.0,
      "step": 33275
    },
    {
      "entropy": 5.245683097839356,
      "epoch": 3.737687426293031,
      "grad_norm": 1.0,
      "learning_rate": 0.00036650596387381615,
      "loss": 4.9017,
      "mean_token_accuracy": 0.22865183353424073,
      "num_tokens": 74306471.0,
      "step": 33280
    },
    {
      "entropy": 5.244932317733765,
      "epoch": 3.7382490032009885,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036646928396033,
      "loss": 4.8689,
      "mean_token_accuracy": 0.2305384024977684,
      "num_tokens": 74318875.0,
      "step": 33285
    },
    {
      "entropy": 5.248578643798828,
      "epoch": 3.7388105801089457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036643260113442634,
      "loss": 4.9011,
      "mean_token_accuracy": 0.22054578810930253,
      "num_tokens": 74328464.0,
      "step": 33290
    },
    {
      "entropy": 5.291854381561279,
      "epoch": 3.7393721570169034,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003663959153972732,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22702265828847884,
      "num_tokens": 74339462.0,
      "step": 33295
    },
    {
      "entropy": 5.307493734359741,
      "epoch": 3.739933733924861,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003663592267500388,
      "loss": 4.9403,
      "mean_token_accuracy": 0.22475723475217818,
      "num_tokens": 74350667.0,
      "step": 33300
    },
    {
      "entropy": 5.251964378356933,
      "epoch": 3.7404953108328187,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003663225351938911,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22339988350868226,
      "num_tokens": 74361734.0,
      "step": 33305
    },
    {
      "entropy": 5.221673250198364,
      "epoch": 3.741056887740776,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003662858407299984,
      "loss": 4.9589,
      "mean_token_accuracy": 0.2097532793879509,
      "num_tokens": 74373465.0,
      "step": 33310
    },
    {
      "entropy": 5.27665867805481,
      "epoch": 3.7416184646487336,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036624914335952924,
      "loss": 4.8556,
      "mean_token_accuracy": 0.22373138070106507,
      "num_tokens": 74384853.0,
      "step": 33315
    },
    {
      "entropy": 5.206744718551636,
      "epoch": 3.7421800415566913,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036621244308365195,
      "loss": 4.7479,
      "mean_token_accuracy": 0.22778978198766708,
      "num_tokens": 74394356.0,
      "step": 33320
    },
    {
      "entropy": 5.274547004699707,
      "epoch": 3.742741618464649,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036617573990353503,
      "loss": 4.9471,
      "mean_token_accuracy": 0.22029224187135696,
      "num_tokens": 74406721.0,
      "step": 33325
    },
    {
      "entropy": 5.265752267837525,
      "epoch": 3.743303195372606,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003661390338203472,
      "loss": 4.9211,
      "mean_token_accuracy": 0.22251540273427964,
      "num_tokens": 74418217.0,
      "step": 33330
    },
    {
      "entropy": 5.2587137699127195,
      "epoch": 3.743864772280564,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003661023248352572,
      "loss": 4.8632,
      "mean_token_accuracy": 0.2274929851293564,
      "num_tokens": 74428644.0,
      "step": 33335
    },
    {
      "entropy": 5.208025074005127,
      "epoch": 3.7444263491885215,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003660656129494339,
      "loss": 4.8484,
      "mean_token_accuracy": 0.22903382182121276,
      "num_tokens": 74439575.0,
      "step": 33340
    },
    {
      "entropy": 5.253690576553344,
      "epoch": 3.7449879260964787,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036602889816404613,
      "loss": 4.9228,
      "mean_token_accuracy": 0.22053639888763427,
      "num_tokens": 74450191.0,
      "step": 33345
    },
    {
      "entropy": 5.302103042602539,
      "epoch": 3.7455495030044363,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036599218048026294,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22199006080627443,
      "num_tokens": 74460816.0,
      "step": 33350
    },
    {
      "entropy": 5.246585941314697,
      "epoch": 3.746111079912394,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036595545989925345,
      "loss": 4.8256,
      "mean_token_accuracy": 0.2342224359512329,
      "num_tokens": 74470999.0,
      "step": 33355
    },
    {
      "entropy": 5.2301191806793215,
      "epoch": 3.7466726568203517,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036591873642218676,
      "loss": 4.7202,
      "mean_token_accuracy": 0.23681976348161698,
      "num_tokens": 74480884.0,
      "step": 33360
    },
    {
      "entropy": 5.178241157531739,
      "epoch": 3.7472342337283093,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036588201005023223,
      "loss": 4.8622,
      "mean_token_accuracy": 0.22850542068481444,
      "num_tokens": 74492500.0,
      "step": 33365
    },
    {
      "entropy": 5.226892948150635,
      "epoch": 3.7477958106362665,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003658452807845593,
      "loss": 4.8677,
      "mean_token_accuracy": 0.22540850192308426,
      "num_tokens": 74504635.0,
      "step": 33370
    },
    {
      "entropy": 5.320550584793091,
      "epoch": 3.748357387544224,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036580854862633747,
      "loss": 4.9351,
      "mean_token_accuracy": 0.2201668605208397,
      "num_tokens": 74515743.0,
      "step": 33375
    },
    {
      "entropy": 5.19791202545166,
      "epoch": 3.748918964452182,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003657718135767362,
      "loss": 4.8603,
      "mean_token_accuracy": 0.22290459722280503,
      "num_tokens": 74528089.0,
      "step": 33380
    },
    {
      "entropy": 5.300633478164673,
      "epoch": 3.749480541360139,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003657350756369251,
      "loss": 4.8893,
      "mean_token_accuracy": 0.2264542117714882,
      "num_tokens": 74538543.0,
      "step": 33385
    },
    {
      "entropy": 5.343930006027222,
      "epoch": 3.7500421182680967,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00036569833480807403,
      "loss": 4.9686,
      "mean_token_accuracy": 0.2210914745926857,
      "num_tokens": 74549825.0,
      "step": 33390
    },
    {
      "entropy": 5.228324270248413,
      "epoch": 3.7506036951760544,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003656615910913528,
      "loss": 4.8094,
      "mean_token_accuracy": 0.22784673422574997,
      "num_tokens": 74560304.0,
      "step": 33395
    },
    {
      "entropy": 5.144937515258789,
      "epoch": 3.751165272084012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036562484448793135,
      "loss": 4.7854,
      "mean_token_accuracy": 0.23051050901412964,
      "num_tokens": 74571549.0,
      "step": 33400
    },
    {
      "entropy": 5.318655633926392,
      "epoch": 3.7517268489919693,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003655880949989798,
      "loss": 4.9509,
      "mean_token_accuracy": 0.21888828426599502,
      "num_tokens": 74581865.0,
      "step": 33405
    },
    {
      "entropy": 5.212148952484131,
      "epoch": 3.752288425899927,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00036555134262566806,
      "loss": 4.7577,
      "mean_token_accuracy": 0.22617979049682618,
      "num_tokens": 74592079.0,
      "step": 33410
    },
    {
      "entropy": 5.217684507369995,
      "epoch": 3.7528500028078846,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003655145873691665,
      "loss": 4.9057,
      "mean_token_accuracy": 0.22514787316322327,
      "num_tokens": 74603058.0,
      "step": 33415
    },
    {
      "entropy": 5.280522060394287,
      "epoch": 3.7534115797158423,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036547782923064537,
      "loss": 4.9056,
      "mean_token_accuracy": 0.22269367128610612,
      "num_tokens": 74614008.0,
      "step": 33420
    },
    {
      "entropy": 5.374057817459106,
      "epoch": 3.7539731566237995,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036544106821127505,
      "loss": 4.9964,
      "mean_token_accuracy": 0.21568580120801925,
      "num_tokens": 74625146.0,
      "step": 33425
    },
    {
      "entropy": 5.374421215057373,
      "epoch": 3.754534733531757,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000365404304312226,
      "loss": 4.9989,
      "mean_token_accuracy": 0.2197217673063278,
      "num_tokens": 74636578.0,
      "step": 33430
    },
    {
      "entropy": 5.268278503417969,
      "epoch": 3.755096310439715,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036536753753466897,
      "loss": 4.9156,
      "mean_token_accuracy": 0.2215370789170265,
      "num_tokens": 74648551.0,
      "step": 33435
    },
    {
      "entropy": 5.232323932647705,
      "epoch": 3.755657887347672,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003653307678797744,
      "loss": 4.7505,
      "mean_token_accuracy": 0.23150190263986586,
      "num_tokens": 74658712.0,
      "step": 33440
    },
    {
      "entropy": 5.205546140670776,
      "epoch": 3.7562194642556297,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003652939953487131,
      "loss": 4.8219,
      "mean_token_accuracy": 0.220924311876297,
      "num_tokens": 74670196.0,
      "step": 33445
    },
    {
      "entropy": 5.293343210220337,
      "epoch": 3.7567810411635874,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036525721994265606,
      "loss": 4.939,
      "mean_token_accuracy": 0.22484760880470275,
      "num_tokens": 74680577.0,
      "step": 33450
    },
    {
      "entropy": 5.279461765289307,
      "epoch": 3.757342618071545,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003652204416627741,
      "loss": 4.8034,
      "mean_token_accuracy": 0.23063654601573944,
      "num_tokens": 74692054.0,
      "step": 33455
    },
    {
      "entropy": 5.25291337966919,
      "epoch": 3.7579041949795027,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036518366051023824,
      "loss": 4.8931,
      "mean_token_accuracy": 0.22525404989719391,
      "num_tokens": 74703324.0,
      "step": 33460
    },
    {
      "entropy": 5.200524473190308,
      "epoch": 3.75846577188746,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003651468764862197,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22726139575242996,
      "num_tokens": 74713028.0,
      "step": 33465
    },
    {
      "entropy": 5.198781585693359,
      "epoch": 3.7590273487954176,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036511008959188966,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22398777008056642,
      "num_tokens": 74723964.0,
      "step": 33470
    },
    {
      "entropy": 5.236271715164184,
      "epoch": 3.759588925703375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036507329982841935,
      "loss": 4.871,
      "mean_token_accuracy": 0.22829782217741013,
      "num_tokens": 74734928.0,
      "step": 33475
    },
    {
      "entropy": 5.381882429122925,
      "epoch": 3.7601505026113324,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003650365071969803,
      "loss": 5.0078,
      "mean_token_accuracy": 0.21767152696847916,
      "num_tokens": 74747198.0,
      "step": 33480
    },
    {
      "entropy": 5.2911937713623045,
      "epoch": 3.76071207951929,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003649997116987439,
      "loss": 4.8975,
      "mean_token_accuracy": 0.23086741268634797,
      "num_tokens": 74758180.0,
      "step": 33485
    },
    {
      "entropy": 5.300287199020386,
      "epoch": 3.7612736564272478,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003649629133348818,
      "loss": 4.9129,
      "mean_token_accuracy": 0.22216543406248093,
      "num_tokens": 74769227.0,
      "step": 33490
    },
    {
      "entropy": 5.251453256607055,
      "epoch": 3.7618352333352054,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036492611210656574,
      "loss": 4.8524,
      "mean_token_accuracy": 0.22405951470136642,
      "num_tokens": 74780347.0,
      "step": 33495
    },
    {
      "entropy": 5.214123868942261,
      "epoch": 3.7623968102431626,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003648893080149673,
      "loss": 4.8705,
      "mean_token_accuracy": 0.22545224279165268,
      "num_tokens": 74791902.0,
      "step": 33500
    },
    {
      "entropy": 5.228350639343262,
      "epoch": 3.7629583871511203,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00036485250106125844,
      "loss": 4.9056,
      "mean_token_accuracy": 0.2235523834824562,
      "num_tokens": 74803530.0,
      "step": 33505
    },
    {
      "entropy": 5.255982828140259,
      "epoch": 3.763519964059078,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000364815691246611,
      "loss": 4.853,
      "mean_token_accuracy": 0.23394372463226318,
      "num_tokens": 74815128.0,
      "step": 33510
    },
    {
      "entropy": 5.388701915740967,
      "epoch": 3.7640815409670356,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036477887857219726,
      "loss": 5.0617,
      "mean_token_accuracy": 0.21755849868059157,
      "num_tokens": 74827729.0,
      "step": 33515
    },
    {
      "entropy": 5.2362456798553465,
      "epoch": 3.764643117874993,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036474206303918914,
      "loss": 4.848,
      "mean_token_accuracy": 0.22517725974321365,
      "num_tokens": 74837922.0,
      "step": 33520
    },
    {
      "entropy": 5.240433406829834,
      "epoch": 3.7652046947829505,
      "grad_norm": 0.96875,
      "learning_rate": 0.000364705244648759,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23035919666290283,
      "num_tokens": 74848239.0,
      "step": 33525
    },
    {
      "entropy": 5.290357208251953,
      "epoch": 3.765766271690908,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003646684234020789,
      "loss": 4.9266,
      "mean_token_accuracy": 0.21883775144815446,
      "num_tokens": 74860454.0,
      "step": 33530
    },
    {
      "entropy": 5.208421659469605,
      "epoch": 3.7663278485988654,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036463159930032163,
      "loss": 4.8303,
      "mean_token_accuracy": 0.22171302139759064,
      "num_tokens": 74871849.0,
      "step": 33535
    },
    {
      "entropy": 5.281003522872925,
      "epoch": 3.766889425506823,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036459477234465935,
      "loss": 4.9493,
      "mean_token_accuracy": 0.21433872282505034,
      "num_tokens": 74883877.0,
      "step": 33540
    },
    {
      "entropy": 5.462366199493408,
      "epoch": 3.7674510024147807,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00036455794253626486,
      "loss": 5.138,
      "mean_token_accuracy": 0.2114585518836975,
      "num_tokens": 74897218.0,
      "step": 33545
    },
    {
      "entropy": 5.329246950149536,
      "epoch": 3.7680125793227384,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036452110987631066,
      "loss": 4.9029,
      "mean_token_accuracy": 0.22066891342401504,
      "num_tokens": 74908708.0,
      "step": 33550
    },
    {
      "entropy": 5.256685209274292,
      "epoch": 3.768574156230696,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036448427436596966,
      "loss": 4.8447,
      "mean_token_accuracy": 0.22835680693387986,
      "num_tokens": 74919664.0,
      "step": 33555
    },
    {
      "entropy": 5.199338436126709,
      "epoch": 3.7691357331386532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003644474360064146,
      "loss": 4.8856,
      "mean_token_accuracy": 0.227119979262352,
      "num_tokens": 74930386.0,
      "step": 33560
    },
    {
      "entropy": 5.226723480224609,
      "epoch": 3.769697310046611,
      "grad_norm": 1.0,
      "learning_rate": 0.0003644105947988185,
      "loss": 4.8769,
      "mean_token_accuracy": 0.2244809627532959,
      "num_tokens": 74942835.0,
      "step": 33565
    },
    {
      "entropy": 5.3035801410675045,
      "epoch": 3.7702588869545686,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036437375074435446,
      "loss": 4.8925,
      "mean_token_accuracy": 0.2182585284113884,
      "num_tokens": 74954332.0,
      "step": 33570
    },
    {
      "entropy": 5.3248392105102536,
      "epoch": 3.770820463862526,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003643369038441955,
      "loss": 4.9621,
      "mean_token_accuracy": 0.2210143193602562,
      "num_tokens": 74965530.0,
      "step": 33575
    },
    {
      "entropy": 5.248055362701416,
      "epoch": 3.7713820407704834,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036430005409951476,
      "loss": 4.8988,
      "mean_token_accuracy": 0.22345015853643418,
      "num_tokens": 74976319.0,
      "step": 33580
    },
    {
      "entropy": 5.238651418685913,
      "epoch": 3.771943617678441,
      "grad_norm": 1.0,
      "learning_rate": 0.00036426320151148575,
      "loss": 4.8841,
      "mean_token_accuracy": 0.2211881622672081,
      "num_tokens": 74987948.0,
      "step": 33585
    },
    {
      "entropy": 5.293311643600464,
      "epoch": 3.7725051945863988,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036422634608128184,
      "loss": 4.9093,
      "mean_token_accuracy": 0.22801199108362197,
      "num_tokens": 74998989.0,
      "step": 33590
    },
    {
      "entropy": 5.2387395858764645,
      "epoch": 3.773066771494356,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036418948781007636,
      "loss": 4.8887,
      "mean_token_accuracy": 0.22605094611644744,
      "num_tokens": 75010510.0,
      "step": 33595
    },
    {
      "entropy": 5.266831588745117,
      "epoch": 3.7736283484023136,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00036415262669904305,
      "loss": 4.9474,
      "mean_token_accuracy": 0.21682542264461518,
      "num_tokens": 75020708.0,
      "step": 33600
    },
    {
      "entropy": 5.321080112457276,
      "epoch": 3.7741899253102713,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036411576274935546,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22832269519567489,
      "num_tokens": 75031468.0,
      "step": 33605
    },
    {
      "entropy": 5.213931131362915,
      "epoch": 3.774751502218229,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036407889596218757,
      "loss": 4.7789,
      "mean_token_accuracy": 0.2326907843351364,
      "num_tokens": 75042170.0,
      "step": 33610
    },
    {
      "entropy": 5.161793184280396,
      "epoch": 3.775313079126186,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000364042026338713,
      "loss": 4.8281,
      "mean_token_accuracy": 0.22095779329538345,
      "num_tokens": 75053916.0,
      "step": 33615
    },
    {
      "entropy": 5.226132822036743,
      "epoch": 3.775874656034144,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036400515388010574,
      "loss": 4.8448,
      "mean_token_accuracy": 0.23154441118240357,
      "num_tokens": 75065026.0,
      "step": 33620
    },
    {
      "entropy": 5.220203399658203,
      "epoch": 3.7764362329421015,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003639682785875399,
      "loss": 4.8875,
      "mean_token_accuracy": 0.2179056167602539,
      "num_tokens": 75075617.0,
      "step": 33625
    },
    {
      "entropy": 5.307335615158081,
      "epoch": 3.7769978098500587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036393140046218957,
      "loss": 4.889,
      "mean_token_accuracy": 0.21742044389247894,
      "num_tokens": 75085931.0,
      "step": 33630
    },
    {
      "entropy": 5.344880437850952,
      "epoch": 3.7775593867580164,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036389451950522895,
      "loss": 4.9,
      "mean_token_accuracy": 0.23100185841321946,
      "num_tokens": 75096404.0,
      "step": 33635
    },
    {
      "entropy": 5.313661193847656,
      "epoch": 3.778120963665974,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036385763571783236,
      "loss": 4.9982,
      "mean_token_accuracy": 0.21929064095020295,
      "num_tokens": 75108300.0,
      "step": 33640
    },
    {
      "entropy": 5.257787752151489,
      "epoch": 3.7786825405739317,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003638207491011742,
      "loss": 4.9439,
      "mean_token_accuracy": 0.21763595640659333,
      "num_tokens": 75119799.0,
      "step": 33645
    },
    {
      "entropy": 5.211350536346435,
      "epoch": 3.7792441174818894,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003637838596564289,
      "loss": 4.8149,
      "mean_token_accuracy": 0.22670796364545823,
      "num_tokens": 75130803.0,
      "step": 33650
    },
    {
      "entropy": 5.264693069458008,
      "epoch": 3.7798056943898466,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036374696738477105,
      "loss": 4.8274,
      "mean_token_accuracy": 0.2305898129940033,
      "num_tokens": 75140621.0,
      "step": 33655
    },
    {
      "entropy": 5.244742822647095,
      "epoch": 3.7803672712978043,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003637100722873754,
      "loss": 4.8791,
      "mean_token_accuracy": 0.21747235357761383,
      "num_tokens": 75152446.0,
      "step": 33660
    },
    {
      "entropy": 5.306168746948242,
      "epoch": 3.780928848205762,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003636731743654166,
      "loss": 4.8886,
      "mean_token_accuracy": 0.2231098547577858,
      "num_tokens": 75162936.0,
      "step": 33665
    },
    {
      "entropy": 5.150378179550171,
      "epoch": 3.781490425113719,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00036363627362006955,
      "loss": 4.8363,
      "mean_token_accuracy": 0.2254427492618561,
      "num_tokens": 75174184.0,
      "step": 33670
    },
    {
      "entropy": 5.0941730499267575,
      "epoch": 3.782052002021677,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003635993700525092,
      "loss": 4.7985,
      "mean_token_accuracy": 0.2326538860797882,
      "num_tokens": 75185609.0,
      "step": 33675
    },
    {
      "entropy": 5.321225595474243,
      "epoch": 3.7826135789296345,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036356246366391054,
      "loss": 4.97,
      "mean_token_accuracy": 0.21855993270874025,
      "num_tokens": 75196098.0,
      "step": 33680
    },
    {
      "entropy": 5.382747554779053,
      "epoch": 3.783175155837592,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036352555445544864,
      "loss": 4.945,
      "mean_token_accuracy": 0.22198013961315155,
      "num_tokens": 75206411.0,
      "step": 33685
    },
    {
      "entropy": 5.2462788105010985,
      "epoch": 3.78373673274555,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036348864242829874,
      "loss": 4.78,
      "mean_token_accuracy": 0.22870215326547622,
      "num_tokens": 75217029.0,
      "step": 33690
    },
    {
      "entropy": 5.155340623855591,
      "epoch": 3.784298309653507,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036345172758363623,
      "loss": 4.7986,
      "mean_token_accuracy": 0.2232123374938965,
      "num_tokens": 75228124.0,
      "step": 33695
    },
    {
      "entropy": 5.136536312103272,
      "epoch": 3.7848598865614647,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003634148099226364,
      "loss": 4.8737,
      "mean_token_accuracy": 0.21708013564348222,
      "num_tokens": 75239776.0,
      "step": 33700
    },
    {
      "entropy": 5.217989492416382,
      "epoch": 3.7854214634694223,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036337788944647466,
      "loss": 4.8117,
      "mean_token_accuracy": 0.22585595548152923,
      "num_tokens": 75251354.0,
      "step": 33705
    },
    {
      "entropy": 5.307124614715576,
      "epoch": 3.7859830403773795,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003633409661563267,
      "loss": 4.9364,
      "mean_token_accuracy": 0.21382663547992706,
      "num_tokens": 75263167.0,
      "step": 33710
    },
    {
      "entropy": 5.212621212005615,
      "epoch": 3.786544617285337,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036330404005336807,
      "loss": 4.8455,
      "mean_token_accuracy": 0.23019191920757293,
      "num_tokens": 75274200.0,
      "step": 33715
    },
    {
      "entropy": 5.261620950698853,
      "epoch": 3.787106194193295,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003632671111387746,
      "loss": 4.945,
      "mean_token_accuracy": 0.21928970664739608,
      "num_tokens": 75287060.0,
      "step": 33720
    },
    {
      "entropy": 5.340345048904419,
      "epoch": 3.787667771101252,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036323017941372203,
      "loss": 4.9256,
      "mean_token_accuracy": 0.2285855770111084,
      "num_tokens": 75297862.0,
      "step": 33725
    },
    {
      "entropy": 5.315982294082642,
      "epoch": 3.7882293480092097,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003631932448793864,
      "loss": 4.9412,
      "mean_token_accuracy": 0.2215859517455101,
      "num_tokens": 75308844.0,
      "step": 33730
    },
    {
      "entropy": 5.343476724624634,
      "epoch": 3.7887909249171674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003631563075369436,
      "loss": 5.0018,
      "mean_token_accuracy": 0.21907846629619598,
      "num_tokens": 75321035.0,
      "step": 33735
    },
    {
      "entropy": 5.183072185516357,
      "epoch": 3.789352501825125,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00036311936738756977,
      "loss": 4.7547,
      "mean_token_accuracy": 0.2299270898103714,
      "num_tokens": 75331407.0,
      "step": 33740
    },
    {
      "entropy": 5.211436653137207,
      "epoch": 3.7899140787330827,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003630824244324411,
      "loss": 4.8527,
      "mean_token_accuracy": 0.2300698786973953,
      "num_tokens": 75343082.0,
      "step": 33745
    },
    {
      "entropy": 5.224194478988648,
      "epoch": 3.79047565564104,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003630454786727339,
      "loss": 4.835,
      "mean_token_accuracy": 0.233067986369133,
      "num_tokens": 75353345.0,
      "step": 33750
    },
    {
      "entropy": 5.326954555511475,
      "epoch": 3.7910372325489976,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003630085301096245,
      "loss": 4.9856,
      "mean_token_accuracy": 0.22262730598449706,
      "num_tokens": 75364039.0,
      "step": 33755
    },
    {
      "entropy": 5.364118766784668,
      "epoch": 3.7915988094569553,
      "grad_norm": 1.0,
      "learning_rate": 0.00036297157874428933,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2257040858268738,
      "num_tokens": 75374166.0,
      "step": 33760
    },
    {
      "entropy": 5.253111886978149,
      "epoch": 3.7921603863649125,
      "grad_norm": 1.0,
      "learning_rate": 0.00036293462457790505,
      "loss": 4.8829,
      "mean_token_accuracy": 0.21890429705381392,
      "num_tokens": 75384792.0,
      "step": 33765
    },
    {
      "entropy": 5.284813737869262,
      "epoch": 3.79272196327287,
      "grad_norm": 0.953125,
      "learning_rate": 0.00036289766761164814,
      "loss": 4.9417,
      "mean_token_accuracy": 0.21590101718902588,
      "num_tokens": 75395478.0,
      "step": 33770
    },
    {
      "entropy": 5.310920524597168,
      "epoch": 3.793283540180828,
      "grad_norm": 1.0,
      "learning_rate": 0.00036286070784669543,
      "loss": 4.945,
      "mean_token_accuracy": 0.22271737903356553,
      "num_tokens": 75406293.0,
      "step": 33775
    },
    {
      "entropy": 5.302475023269653,
      "epoch": 3.7938451170887855,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00036282374528422373,
      "loss": 4.9322,
      "mean_token_accuracy": 0.22914220690727233,
      "num_tokens": 75417588.0,
      "step": 33780
    },
    {
      "entropy": 5.163408946990967,
      "epoch": 3.794406693996743,
      "grad_norm": 1.0,
      "learning_rate": 0.00036278677992540993,
      "loss": 4.8185,
      "mean_token_accuracy": 0.22490643411874772,
      "num_tokens": 75428651.0,
      "step": 33785
    },
    {
      "entropy": 5.314674282073975,
      "epoch": 3.7949682709047003,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000362749811771431,
      "loss": 4.9249,
      "mean_token_accuracy": 0.22496708184480668,
      "num_tokens": 75438808.0,
      "step": 33790
    },
    {
      "entropy": 5.225064039230347,
      "epoch": 3.795529847812658,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003627128408234641,
      "loss": 4.7784,
      "mean_token_accuracy": 0.22621652632951736,
      "num_tokens": 75449326.0,
      "step": 33795
    },
    {
      "entropy": 5.223838901519775,
      "epoch": 3.7960914247206157,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003626758670826861,
      "loss": 4.911,
      "mean_token_accuracy": 0.22441623210906983,
      "num_tokens": 75460499.0,
      "step": 33800
    },
    {
      "entropy": 5.258955240249634,
      "epoch": 3.796653001628573,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003626388905502747,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22780285328626632,
      "num_tokens": 75471031.0,
      "step": 33805
    },
    {
      "entropy": 5.25372633934021,
      "epoch": 3.7972145785365305,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000362601911227407,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2224702149629593,
      "num_tokens": 75483608.0,
      "step": 33810
    },
    {
      "entropy": 5.322487640380859,
      "epoch": 3.797776155444488,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003625649291152604,
      "loss": 4.9258,
      "mean_token_accuracy": 0.22626373916864395,
      "num_tokens": 75495221.0,
      "step": 33815
    },
    {
      "entropy": 5.213936614990234,
      "epoch": 3.7983377323524454,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036252794421501257,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23091211169958115,
      "num_tokens": 75507245.0,
      "step": 33820
    },
    {
      "entropy": 5.313355827331543,
      "epoch": 3.798899309260403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036249095652784096,
      "loss": 4.9747,
      "mean_token_accuracy": 0.2214798301458359,
      "num_tokens": 75518525.0,
      "step": 33825
    },
    {
      "entropy": 5.260707139968872,
      "epoch": 3.7994608861683608,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036245396605492344,
      "loss": 4.8671,
      "mean_token_accuracy": 0.2189846530556679,
      "num_tokens": 75529674.0,
      "step": 33830
    },
    {
      "entropy": 5.201648902893067,
      "epoch": 3.8000224630763184,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003624169727974376,
      "loss": 4.746,
      "mean_token_accuracy": 0.23498879075050355,
      "num_tokens": 75539596.0,
      "step": 33835
    },
    {
      "entropy": 5.1575652122497555,
      "epoch": 3.800584039984276,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036237997675656167,
      "loss": 4.831,
      "mean_token_accuracy": 0.23171035349369049,
      "num_tokens": 75551164.0,
      "step": 33840
    },
    {
      "entropy": 5.3106895923614506,
      "epoch": 3.8011456168922333,
      "grad_norm": 0.90625,
      "learning_rate": 0.0003623429779334732,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22234206646680832,
      "num_tokens": 75563758.0,
      "step": 33845
    },
    {
      "entropy": 5.385889387130737,
      "epoch": 3.801707193800191,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00036230597632935053,
      "loss": 5.011,
      "mean_token_accuracy": 0.22336628437042236,
      "num_tokens": 75574332.0,
      "step": 33850
    },
    {
      "entropy": 5.292816638946533,
      "epoch": 3.8022687707081486,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003622689719453717,
      "loss": 5.0033,
      "mean_token_accuracy": 0.21377793103456497,
      "num_tokens": 75586229.0,
      "step": 33855
    },
    {
      "entropy": 5.274424839019775,
      "epoch": 3.802830347616106,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00036223196478271495,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22812028229236603,
      "num_tokens": 75597396.0,
      "step": 33860
    },
    {
      "entropy": 5.262957286834717,
      "epoch": 3.8033919245240635,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003621949548425585,
      "loss": 4.9548,
      "mean_token_accuracy": 0.218230040371418,
      "num_tokens": 75609171.0,
      "step": 33865
    },
    {
      "entropy": 5.362941598892212,
      "epoch": 3.803953501432021,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000362157942126081,
      "loss": 4.931,
      "mean_token_accuracy": 0.2259111449122429,
      "num_tokens": 75619855.0,
      "step": 33870
    },
    {
      "entropy": 5.388857126235962,
      "epoch": 3.804515078339979,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003621209266344608,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2223486453294754,
      "num_tokens": 75631810.0,
      "step": 33875
    },
    {
      "entropy": 5.203442573547363,
      "epoch": 3.8050766552479365,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003620839083688765,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22794607728719712,
      "num_tokens": 75644894.0,
      "step": 33880
    },
    {
      "entropy": 5.273346471786499,
      "epoch": 3.8056382321558937,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003620468873305068,
      "loss": 4.8611,
      "mean_token_accuracy": 0.22466604858636857,
      "num_tokens": 75657357.0,
      "step": 33885
    },
    {
      "entropy": 5.274138069152832,
      "epoch": 3.8061998090638514,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00036200986352053036,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22581869810819627,
      "num_tokens": 75667357.0,
      "step": 33890
    },
    {
      "entropy": 5.180035400390625,
      "epoch": 3.806761385971809,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00036197283694012615,
      "loss": 4.7984,
      "mean_token_accuracy": 0.2302628442645073,
      "num_tokens": 75678758.0,
      "step": 33895
    },
    {
      "entropy": 5.290223932266235,
      "epoch": 3.8073229628797662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003619358075904732,
      "loss": 4.8544,
      "mean_token_accuracy": 0.23072421699762344,
      "num_tokens": 75688392.0,
      "step": 33900
    },
    {
      "entropy": 5.270389699935913,
      "epoch": 3.807884539787724,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00036189877547275023,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22485502809286118,
      "num_tokens": 75699839.0,
      "step": 33905
    },
    {
      "entropy": 5.300531721115112,
      "epoch": 3.8084461166956816,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00036186174058813666,
      "loss": 4.8746,
      "mean_token_accuracy": 0.22490164488554001,
      "num_tokens": 75711877.0,
      "step": 33910
    },
    {
      "entropy": 5.213392066955566,
      "epoch": 3.8090076936036388,
      "grad_norm": 0.984375,
      "learning_rate": 0.00036182470293781154,
      "loss": 4.7982,
      "mean_token_accuracy": 0.22631338238716125,
      "num_tokens": 75721646.0,
      "step": 33915
    },
    {
      "entropy": 5.2864899158477785,
      "epoch": 3.8095692705115964,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003617876625229543,
      "loss": 5.0052,
      "mean_token_accuracy": 0.21170790493488312,
      "num_tokens": 75733291.0,
      "step": 33920
    },
    {
      "entropy": 5.2403546333312985,
      "epoch": 3.810130847419554,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003617506193447442,
      "loss": 4.8879,
      "mean_token_accuracy": 0.2346237853169441,
      "num_tokens": 75744475.0,
      "step": 33925
    },
    {
      "entropy": 5.197023487091064,
      "epoch": 3.8106924243275118,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003617135734043606,
      "loss": 4.869,
      "mean_token_accuracy": 0.2241922438144684,
      "num_tokens": 75755098.0,
      "step": 33930
    },
    {
      "entropy": 5.306526517868042,
      "epoch": 3.8112540012354694,
      "grad_norm": 0.859375,
      "learning_rate": 0.0003616765247029833,
      "loss": 4.8838,
      "mean_token_accuracy": 0.2312107190489769,
      "num_tokens": 75767405.0,
      "step": 33935
    },
    {
      "entropy": 5.281508207321167,
      "epoch": 3.8118155781434266,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036163947324179186,
      "loss": 4.8897,
      "mean_token_accuracy": 0.21799315512180328,
      "num_tokens": 75778614.0,
      "step": 33940
    },
    {
      "entropy": 5.245003604888916,
      "epoch": 3.8123771550513843,
      "grad_norm": 1.0,
      "learning_rate": 0.00036160241902196587,
      "loss": 4.8447,
      "mean_token_accuracy": 0.2345143124461174,
      "num_tokens": 75789087.0,
      "step": 33945
    },
    {
      "entropy": 5.27449746131897,
      "epoch": 3.812938731959342,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036156536204468535,
      "loss": 4.8758,
      "mean_token_accuracy": 0.2228206768631935,
      "num_tokens": 75799431.0,
      "step": 33950
    },
    {
      "entropy": 5.221765470504761,
      "epoch": 3.813500308867299,
      "grad_norm": 1.046875,
      "learning_rate": 0.00036152830231113013,
      "loss": 4.8582,
      "mean_token_accuracy": 0.2215055912733078,
      "num_tokens": 75810149.0,
      "step": 33955
    },
    {
      "entropy": 5.232305192947388,
      "epoch": 3.814061885775257,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003614912398224802,
      "loss": 4.8909,
      "mean_token_accuracy": 0.22316998839378357,
      "num_tokens": 75820072.0,
      "step": 33960
    },
    {
      "entropy": 5.306483125686645,
      "epoch": 3.8146234626832145,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003614541745799156,
      "loss": 4.9232,
      "mean_token_accuracy": 0.21656876057386398,
      "num_tokens": 75831017.0,
      "step": 33965
    },
    {
      "entropy": 5.251373386383056,
      "epoch": 3.815185039591172,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003614171065846166,
      "loss": 4.9101,
      "mean_token_accuracy": 0.22170924246311188,
      "num_tokens": 75842606.0,
      "step": 33970
    },
    {
      "entropy": 5.248541212081909,
      "epoch": 3.81574661649913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003613800358377635,
      "loss": 4.8035,
      "mean_token_accuracy": 0.23051410168409348,
      "num_tokens": 75853820.0,
      "step": 33975
    },
    {
      "entropy": 5.224689769744873,
      "epoch": 3.816308193407087,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003613429623405364,
      "loss": 4.8261,
      "mean_token_accuracy": 0.22492383271455765,
      "num_tokens": 75864959.0,
      "step": 33980
    },
    {
      "entropy": 5.234539031982422,
      "epoch": 3.8168697703150447,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036130588609411597,
      "loss": 4.8102,
      "mean_token_accuracy": 0.229533551633358,
      "num_tokens": 75877247.0,
      "step": 33985
    },
    {
      "entropy": 5.307689905166626,
      "epoch": 3.8174313472230024,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00036126880709968264,
      "loss": 5.0096,
      "mean_token_accuracy": 0.21074576526880265,
      "num_tokens": 75888654.0,
      "step": 33990
    },
    {
      "entropy": 5.279342603683472,
      "epoch": 3.8179929241309596,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036123172535841705,
      "loss": 4.8394,
      "mean_token_accuracy": 0.2232509061694145,
      "num_tokens": 75899953.0,
      "step": 33995
    },
    {
      "entropy": 5.2743579864501955,
      "epoch": 3.8185545010389172,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003611946408714998,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23284826427698135,
      "num_tokens": 75911174.0,
      "step": 34000
    },
    {
      "entropy": 5.2048583984375,
      "epoch": 3.819116077946875,
      "grad_norm": 1.03125,
      "learning_rate": 0.00036115755364011186,
      "loss": 4.8924,
      "mean_token_accuracy": 0.22845025956630707,
      "num_tokens": 75922221.0,
      "step": 34005
    },
    {
      "entropy": 5.141364812850952,
      "epoch": 3.819677654854832,
      "grad_norm": 1.015625,
      "learning_rate": 0.000361120463665434,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22613843083381652,
      "num_tokens": 75932630.0,
      "step": 34010
    },
    {
      "entropy": 5.238550996780395,
      "epoch": 3.82023923176279,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003610833709486471,
      "loss": 4.8652,
      "mean_token_accuracy": 0.22520046234130858,
      "num_tokens": 75944490.0,
      "step": 34015
    },
    {
      "entropy": 5.252916717529297,
      "epoch": 3.8208008086707475,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003610462754909323,
      "loss": 4.8501,
      "mean_token_accuracy": 0.22754893451929092,
      "num_tokens": 75956608.0,
      "step": 34020
    },
    {
      "entropy": 5.288291978836059,
      "epoch": 3.821362385578705,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00036100917729347073,
      "loss": 4.9252,
      "mean_token_accuracy": 0.21860044002532958,
      "num_tokens": 75967516.0,
      "step": 34025
    },
    {
      "entropy": 5.177999353408813,
      "epoch": 3.8219239624866628,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00036097207635744363,
      "loss": 4.7563,
      "mean_token_accuracy": 0.23434255123138428,
      "num_tokens": 75978329.0,
      "step": 34030
    },
    {
      "entropy": 5.310885429382324,
      "epoch": 3.82248553939462,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00036093497268403215,
      "loss": 4.8744,
      "mean_token_accuracy": 0.2325591638684273,
      "num_tokens": 75990170.0,
      "step": 34035
    },
    {
      "entropy": 5.24755825996399,
      "epoch": 3.8230471163025777,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000360897866274418,
      "loss": 4.8385,
      "mean_token_accuracy": 0.22841050773859023,
      "num_tokens": 76002368.0,
      "step": 34040
    },
    {
      "entropy": 5.221774911880493,
      "epoch": 3.8236086932105353,
      "grad_norm": 0.96875,
      "learning_rate": 0.00036086075712978233,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22780305445194243,
      "num_tokens": 76012867.0,
      "step": 34045
    },
    {
      "entropy": 5.2557769298553465,
      "epoch": 3.8241702701184925,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036082364525130684,
      "loss": 4.9349,
      "mean_token_accuracy": 0.22326294481754302,
      "num_tokens": 76023900.0,
      "step": 34050
    },
    {
      "entropy": 5.193145132064819,
      "epoch": 3.82473184702645,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003607865306401732,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23356823921203612,
      "num_tokens": 76034296.0,
      "step": 34055
    },
    {
      "entropy": 5.258613920211792,
      "epoch": 3.825293423934408,
      "grad_norm": 1.015625,
      "learning_rate": 0.00036074941329756327,
      "loss": 4.9158,
      "mean_token_accuracy": 0.22278940975666045,
      "num_tokens": 76045355.0,
      "step": 34060
    },
    {
      "entropy": 5.260082721710205,
      "epoch": 3.8258550008423655,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003607122932246587,
      "loss": 4.7927,
      "mean_token_accuracy": 0.235805144906044,
      "num_tokens": 76056158.0,
      "step": 34065
    },
    {
      "entropy": 5.241888046264648,
      "epoch": 3.826416577750323,
      "grad_norm": 1.078125,
      "learning_rate": 0.00036067517042264147,
      "loss": 4.9084,
      "mean_token_accuracy": 0.22293949425220488,
      "num_tokens": 76066930.0,
      "step": 34070
    },
    {
      "entropy": 5.251004648208618,
      "epoch": 3.8269781546582804,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003606380448926936,
      "loss": 4.8422,
      "mean_token_accuracy": 0.22782463133335112,
      "num_tokens": 76078791.0,
      "step": 34075
    },
    {
      "entropy": 5.323307371139526,
      "epoch": 3.827539731566238,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003606009166359972,
      "loss": 4.9651,
      "mean_token_accuracy": 0.2189207911491394,
      "num_tokens": 76090356.0,
      "step": 34080
    },
    {
      "entropy": 5.1906288146972654,
      "epoch": 3.8281013084741957,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036056378565373434,
      "loss": 4.8246,
      "mean_token_accuracy": 0.22750007212162018,
      "num_tokens": 76101137.0,
      "step": 34085
    },
    {
      "entropy": 5.124757194519043,
      "epoch": 3.828662885382153,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003605266519470874,
      "loss": 4.7364,
      "mean_token_accuracy": 0.23048990666866304,
      "num_tokens": 76111661.0,
      "step": 34090
    },
    {
      "entropy": 5.2937627792358395,
      "epoch": 3.8292244622901106,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003604895155172386,
      "loss": 4.8925,
      "mean_token_accuracy": 0.22342265099287034,
      "num_tokens": 76122670.0,
      "step": 34095
    },
    {
      "entropy": 5.270393943786621,
      "epoch": 3.8297860391980683,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003604523763653706,
      "loss": 4.98,
      "mean_token_accuracy": 0.21761023998260498,
      "num_tokens": 76133899.0,
      "step": 34100
    },
    {
      "entropy": 5.287248325347901,
      "epoch": 3.8303476161060255,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00036041523449266573,
      "loss": 4.9378,
      "mean_token_accuracy": 0.21489761769771576,
      "num_tokens": 76144865.0,
      "step": 34105
    },
    {
      "entropy": 5.349058723449707,
      "epoch": 3.830909193013983,
      "grad_norm": 0.875,
      "learning_rate": 0.00036037808990030667,
      "loss": 4.9143,
      "mean_token_accuracy": 0.22289246320724487,
      "num_tokens": 76157392.0,
      "step": 34110
    },
    {
      "entropy": 5.3009271144866945,
      "epoch": 3.831470769921941,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003603409425894761,
      "loss": 4.9075,
      "mean_token_accuracy": 0.2201978862285614,
      "num_tokens": 76168339.0,
      "step": 34115
    },
    {
      "entropy": 5.1763793468475345,
      "epoch": 3.8320323468298985,
      "grad_norm": 1.0,
      "learning_rate": 0.0003603037925613568,
      "loss": 4.8383,
      "mean_token_accuracy": 0.23037845641374588,
      "num_tokens": 76179377.0,
      "step": 34120
    },
    {
      "entropy": 5.259068012237549,
      "epoch": 3.832593923737856,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003602666398171317,
      "loss": 4.8657,
      "mean_token_accuracy": 0.22355666011571884,
      "num_tokens": 76190030.0,
      "step": 34125
    },
    {
      "entropy": 5.154702138900757,
      "epoch": 3.8331555006458133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00036022948435798373,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2244683638215065,
      "num_tokens": 76201339.0,
      "step": 34130
    },
    {
      "entropy": 5.2714379787445065,
      "epoch": 3.833717077553771,
      "grad_norm": 1.03125,
      "learning_rate": 0.000360192326185096,
      "loss": 4.9188,
      "mean_token_accuracy": 0.22536829113960266,
      "num_tokens": 76211760.0,
      "step": 34135
    },
    {
      "entropy": 5.295661878585816,
      "epoch": 3.8342786544617287,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003601551652996514,
      "loss": 4.9131,
      "mean_token_accuracy": 0.2219465896487236,
      "num_tokens": 76222573.0,
      "step": 34140
    },
    {
      "entropy": 5.315975189208984,
      "epoch": 3.834840231369686,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003601180017028334,
      "loss": 4.9275,
      "mean_token_accuracy": 0.21755583882331847,
      "num_tokens": 76233547.0,
      "step": 34145
    },
    {
      "entropy": 5.251824426651001,
      "epoch": 3.8354018082776435,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003600808353958252,
      "loss": 4.8667,
      "mean_token_accuracy": 0.2234919250011444,
      "num_tokens": 76246067.0,
      "step": 34150
    },
    {
      "entropy": 5.155296564102173,
      "epoch": 3.835963385185601,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003600436663798102,
      "loss": 4.759,
      "mean_token_accuracy": 0.2405823528766632,
      "num_tokens": 76257698.0,
      "step": 34155
    },
    {
      "entropy": 5.144668531417847,
      "epoch": 3.836524962093559,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00036000649465597185,
      "loss": 4.8168,
      "mean_token_accuracy": 0.23024744987487794,
      "num_tokens": 76268533.0,
      "step": 34160
    },
    {
      "entropy": 5.160677671432495,
      "epoch": 3.8370865390015165,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035996932022549377,
      "loss": 4.7852,
      "mean_token_accuracy": 0.23349788337945937,
      "num_tokens": 76278529.0,
      "step": 34165
    },
    {
      "entropy": 5.20997200012207,
      "epoch": 3.8376481159094737,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035993214308955955,
      "loss": 4.9052,
      "mean_token_accuracy": 0.22010547965764998,
      "num_tokens": 76289461.0,
      "step": 34170
    },
    {
      "entropy": 5.249225425720215,
      "epoch": 3.8382096928174314,
      "grad_norm": 1.015625,
      "learning_rate": 0.000359894963249353,
      "loss": 4.8463,
      "mean_token_accuracy": 0.2277579963207245,
      "num_tokens": 76300641.0,
      "step": 34175
    },
    {
      "entropy": 5.288647556304932,
      "epoch": 3.838771269725389,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035985778070605785,
      "loss": 4.9596,
      "mean_token_accuracy": 0.22065404057502747,
      "num_tokens": 76310918.0,
      "step": 34180
    },
    {
      "entropy": 5.228465557098389,
      "epoch": 3.8393328466333463,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003598205954608581,
      "loss": 4.776,
      "mean_token_accuracy": 0.23188881129026412,
      "num_tokens": 76323037.0,
      "step": 34185
    },
    {
      "entropy": 5.212657260894775,
      "epoch": 3.839894423541304,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035978340751493767,
      "loss": 4.8534,
      "mean_token_accuracy": 0.2246886670589447,
      "num_tokens": 76334252.0,
      "step": 34190
    },
    {
      "entropy": 5.273496627807617,
      "epoch": 3.8404560004492616,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035974621686948063,
      "loss": 4.9519,
      "mean_token_accuracy": 0.2222587525844574,
      "num_tokens": 76344928.0,
      "step": 34195
    },
    {
      "entropy": 5.331060791015625,
      "epoch": 3.841017577357219,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003597090235256712,
      "loss": 4.9733,
      "mean_token_accuracy": 0.21731043010950088,
      "num_tokens": 76356529.0,
      "step": 34200
    },
    {
      "entropy": 5.256835842132569,
      "epoch": 3.8415791542651765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003596718274846936,
      "loss": 4.8856,
      "mean_token_accuracy": 0.21842799782752992,
      "num_tokens": 76366842.0,
      "step": 34205
    },
    {
      "entropy": 5.2716224193573,
      "epoch": 3.842140731173134,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003596346287477322,
      "loss": 4.9217,
      "mean_token_accuracy": 0.21614581644535064,
      "num_tokens": 76377109.0,
      "step": 34210
    },
    {
      "entropy": 5.224264621734619,
      "epoch": 3.842702308081092,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003595974273159714,
      "loss": 4.8666,
      "mean_token_accuracy": 0.2221038669347763,
      "num_tokens": 76387621.0,
      "step": 34215
    },
    {
      "entropy": 5.314399766921997,
      "epoch": 3.8432638849890495,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003595602231905957,
      "loss": 4.9807,
      "mean_token_accuracy": 0.21549178808927535,
      "num_tokens": 76398537.0,
      "step": 34220
    },
    {
      "entropy": 5.28865156173706,
      "epoch": 3.8438254618970067,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035952301637278973,
      "loss": 4.8533,
      "mean_token_accuracy": 0.2239707499742508,
      "num_tokens": 76410192.0,
      "step": 34225
    },
    {
      "entropy": 5.219415283203125,
      "epoch": 3.8443870388049644,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035948580686373803,
      "loss": 4.867,
      "mean_token_accuracy": 0.2267143040895462,
      "num_tokens": 76420024.0,
      "step": 34230
    },
    {
      "entropy": 5.245912075042725,
      "epoch": 3.844948615712922,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035944859466462554,
      "loss": 4.9176,
      "mean_token_accuracy": 0.2290056437253952,
      "num_tokens": 76430833.0,
      "step": 34235
    },
    {
      "entropy": 5.2449061393737795,
      "epoch": 3.8455101926208792,
      "grad_norm": 1.015625,
      "learning_rate": 0.000359411379776637,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2232070431113243,
      "num_tokens": 76442800.0,
      "step": 34240
    },
    {
      "entropy": 5.296216154098511,
      "epoch": 3.846071769528837,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035937416220095745,
      "loss": 4.8757,
      "mean_token_accuracy": 0.2233264699578285,
      "num_tokens": 76454988.0,
      "step": 34245
    },
    {
      "entropy": 5.378973770141601,
      "epoch": 3.8466333464367946,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003593369419387718,
      "loss": 4.9823,
      "mean_token_accuracy": 0.21382761746644974,
      "num_tokens": 76466523.0,
      "step": 34250
    },
    {
      "entropy": 5.3375811100006105,
      "epoch": 3.847194923344752,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003592997189912652,
      "loss": 5.0252,
      "mean_token_accuracy": 0.2051542580127716,
      "num_tokens": 76478691.0,
      "step": 34255
    },
    {
      "entropy": 5.282332849502564,
      "epoch": 3.84775650025271,
      "grad_norm": 0.8671875,
      "learning_rate": 0.00035926249335962296,
      "loss": 4.9855,
      "mean_token_accuracy": 0.21726084649562835,
      "num_tokens": 76490544.0,
      "step": 34260
    },
    {
      "entropy": 5.219467401504517,
      "epoch": 3.848318077160667,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035922526504503016,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23196391612291337,
      "num_tokens": 76500269.0,
      "step": 34265
    },
    {
      "entropy": 5.182312154769898,
      "epoch": 3.8488796540686248,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003591880340486723,
      "loss": 4.8092,
      "mean_token_accuracy": 0.22533345222473145,
      "num_tokens": 76511458.0,
      "step": 34270
    },
    {
      "entropy": 5.270624589920044,
      "epoch": 3.8494412309765824,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003591508003717347,
      "loss": 4.9209,
      "mean_token_accuracy": 0.22263258844614028,
      "num_tokens": 76521722.0,
      "step": 34275
    },
    {
      "entropy": 5.343128061294555,
      "epoch": 3.8500028078845396,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035911356401540306,
      "loss": 4.9786,
      "mean_token_accuracy": 0.22438313961029052,
      "num_tokens": 76532935.0,
      "step": 34280
    },
    {
      "entropy": 5.2737054347991945,
      "epoch": 3.8505643847924973,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003590763249808629,
      "loss": 4.8845,
      "mean_token_accuracy": 0.2260259971022606,
      "num_tokens": 76543545.0,
      "step": 34285
    },
    {
      "entropy": 5.17232027053833,
      "epoch": 3.851125961700455,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035903908326929995,
      "loss": 4.8205,
      "mean_token_accuracy": 0.23204179257154464,
      "num_tokens": 76553621.0,
      "step": 34290
    },
    {
      "entropy": 5.2665996074676515,
      "epoch": 3.851687538608412,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035900183888190005,
      "loss": 4.8779,
      "mean_token_accuracy": 0.2323288008570671,
      "num_tokens": 76563154.0,
      "step": 34295
    },
    {
      "entropy": 5.3016801357269285,
      "epoch": 3.85224911551637,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003589645918198489,
      "loss": 4.832,
      "mean_token_accuracy": 0.22276916056871415,
      "num_tokens": 76573930.0,
      "step": 34300
    },
    {
      "entropy": 5.195742225646972,
      "epoch": 3.8528106924243275,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035892734208433264,
      "loss": 4.7992,
      "mean_token_accuracy": 0.2317344143986702,
      "num_tokens": 76585216.0,
      "step": 34305
    },
    {
      "entropy": 5.239481449127197,
      "epoch": 3.853372269332285,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035889008967653725,
      "loss": 4.9069,
      "mean_token_accuracy": 0.22538412660360335,
      "num_tokens": 76596479.0,
      "step": 34310
    },
    {
      "entropy": 5.351177120208741,
      "epoch": 3.853933846240243,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003588528345976489,
      "loss": 4.984,
      "mean_token_accuracy": 0.21782667934894562,
      "num_tokens": 76607123.0,
      "step": 34315
    },
    {
      "entropy": 5.3067626953125,
      "epoch": 3.8544954231482,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003588155768488537,
      "loss": 4.901,
      "mean_token_accuracy": 0.22823788970708847,
      "num_tokens": 76616898.0,
      "step": 34320
    },
    {
      "entropy": 5.163452434539795,
      "epoch": 3.8550570000561577,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035877831643133803,
      "loss": 4.722,
      "mean_token_accuracy": 0.2344128206372261,
      "num_tokens": 76627528.0,
      "step": 34325
    },
    {
      "entropy": 5.216481304168701,
      "epoch": 3.8556185769641154,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003587410533462883,
      "loss": 4.7891,
      "mean_token_accuracy": 0.2269002839922905,
      "num_tokens": 76639173.0,
      "step": 34330
    },
    {
      "entropy": 5.233207893371582,
      "epoch": 3.8561801538720726,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035870378759489104,
      "loss": 4.9166,
      "mean_token_accuracy": 0.2192594140768051,
      "num_tokens": 76648970.0,
      "step": 34335
    },
    {
      "entropy": 5.208154344558716,
      "epoch": 3.8567417307800302,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003586665191783326,
      "loss": 4.8262,
      "mean_token_accuracy": 0.22290070205926896,
      "num_tokens": 76659101.0,
      "step": 34340
    },
    {
      "entropy": 5.272469568252563,
      "epoch": 3.857303307687988,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003586292480977998,
      "loss": 4.8728,
      "mean_token_accuracy": 0.22796701192855834,
      "num_tokens": 76670087.0,
      "step": 34345
    },
    {
      "entropy": 5.307417201995849,
      "epoch": 3.8578648845959456,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035859197435447927,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22427531629800795,
      "num_tokens": 76681116.0,
      "step": 34350
    },
    {
      "entropy": 5.223203563690186,
      "epoch": 3.8584264615039032,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035855469794955787,
      "loss": 4.8614,
      "mean_token_accuracy": 0.21936872154474257,
      "num_tokens": 76692050.0,
      "step": 34355
    },
    {
      "entropy": 5.3585093975067135,
      "epoch": 3.8589880384118604,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00035851741888422254,
      "loss": 4.9608,
      "mean_token_accuracy": 0.21900322288274765,
      "num_tokens": 76704338.0,
      "step": 34360
    },
    {
      "entropy": 5.225214099884033,
      "epoch": 3.859549615319818,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003584801371596601,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22629442512989045,
      "num_tokens": 76715300.0,
      "step": 34365
    },
    {
      "entropy": 5.283959531784058,
      "epoch": 3.8601111922277758,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003584428527770578,
      "loss": 4.8969,
      "mean_token_accuracy": 0.21730591207742692,
      "num_tokens": 76726654.0,
      "step": 34370
    },
    {
      "entropy": 5.136643362045288,
      "epoch": 3.860672769135733,
      "grad_norm": 1.109375,
      "learning_rate": 0.00035840556573760266,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23317297846078872,
      "num_tokens": 76737087.0,
      "step": 34375
    },
    {
      "entropy": 5.16426682472229,
      "epoch": 3.8612343460436906,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003583682760424819,
      "loss": 4.8199,
      "mean_token_accuracy": 0.22774490416049958,
      "num_tokens": 76748006.0,
      "step": 34380
    },
    {
      "entropy": 5.175247097015381,
      "epoch": 3.8617959229516483,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035833098369288303,
      "loss": 4.7647,
      "mean_token_accuracy": 0.23934967964887618,
      "num_tokens": 76758063.0,
      "step": 34385
    },
    {
      "entropy": 5.21448655128479,
      "epoch": 3.8623574998596055,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035829368868999325,
      "loss": 4.8564,
      "mean_token_accuracy": 0.22728202790021895,
      "num_tokens": 76769080.0,
      "step": 34390
    },
    {
      "entropy": 5.217199182510376,
      "epoch": 3.862919076767563,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035825639103500005,
      "loss": 4.8883,
      "mean_token_accuracy": 0.22437025606632233,
      "num_tokens": 76780743.0,
      "step": 34395
    },
    {
      "entropy": 5.21939606666565,
      "epoch": 3.863480653675521,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035821909072909113,
      "loss": 4.87,
      "mean_token_accuracy": 0.22850780189037323,
      "num_tokens": 76792277.0,
      "step": 34400
    },
    {
      "entropy": 5.23780255317688,
      "epoch": 3.8640422305834785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035818178777345417,
      "loss": 4.7786,
      "mean_token_accuracy": 0.2296567052602768,
      "num_tokens": 76802979.0,
      "step": 34405
    },
    {
      "entropy": 5.190300941467285,
      "epoch": 3.864603807491436,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003581444821692767,
      "loss": 4.8147,
      "mean_token_accuracy": 0.23153113424777985,
      "num_tokens": 76815016.0,
      "step": 34410
    },
    {
      "entropy": 5.196734046936035,
      "epoch": 3.8651653843993934,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003581071739177467,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23342572003602982,
      "num_tokens": 76826360.0,
      "step": 34415
    },
    {
      "entropy": 5.262342309951782,
      "epoch": 3.865726961307351,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00035806986302005205,
      "loss": 4.926,
      "mean_token_accuracy": 0.22201489210128783,
      "num_tokens": 76838876.0,
      "step": 34420
    },
    {
      "entropy": 5.198180675506592,
      "epoch": 3.8662885382153087,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003580325494773808,
      "loss": 4.8087,
      "mean_token_accuracy": 0.22811456322669982,
      "num_tokens": 76849156.0,
      "step": 34425
    },
    {
      "entropy": 5.3015343189239506,
      "epoch": 3.866850115123266,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035799523329092093,
      "loss": 4.9797,
      "mean_token_accuracy": 0.22151432931423187,
      "num_tokens": 76859683.0,
      "step": 34430
    },
    {
      "entropy": 5.183727359771728,
      "epoch": 3.8674116920312236,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035795791446186073,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23426515161991118,
      "num_tokens": 76870509.0,
      "step": 34435
    },
    {
      "entropy": 5.257809638977051,
      "epoch": 3.8679732689391813,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003579205929913882,
      "loss": 4.9199,
      "mean_token_accuracy": 0.22595060765743255,
      "num_tokens": 76882181.0,
      "step": 34440
    },
    {
      "entropy": 5.254165267944336,
      "epoch": 3.868534845847139,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003578832688806919,
      "loss": 4.9555,
      "mean_token_accuracy": 0.21740141808986663,
      "num_tokens": 76894709.0,
      "step": 34445
    },
    {
      "entropy": 5.249564170837402,
      "epoch": 3.8690964227550966,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003578459421309602,
      "loss": 4.823,
      "mean_token_accuracy": 0.23397017568349837,
      "num_tokens": 76905974.0,
      "step": 34450
    },
    {
      "entropy": 5.372653961181641,
      "epoch": 3.869657999663054,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003578086127433816,
      "loss": 4.9779,
      "mean_token_accuracy": 0.22210419774055482,
      "num_tokens": 76918009.0,
      "step": 34455
    },
    {
      "entropy": 5.269751358032226,
      "epoch": 3.8702195765710115,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035777128071914454,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22865015715360643,
      "num_tokens": 76928750.0,
      "step": 34460
    },
    {
      "entropy": 5.289512300491333,
      "epoch": 3.870781153478969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003577339460594378,
      "loss": 4.9835,
      "mean_token_accuracy": 0.2250794216990471,
      "num_tokens": 76939989.0,
      "step": 34465
    },
    {
      "entropy": 5.149450778961182,
      "epoch": 3.8713427303869263,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003576966087654502,
      "loss": 4.7793,
      "mean_token_accuracy": 0.23391834646463394,
      "num_tokens": 76950026.0,
      "step": 34470
    },
    {
      "entropy": 5.309295988082885,
      "epoch": 3.871904307294884,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035765926883837054,
      "loss": 4.9863,
      "mean_token_accuracy": 0.22157077491283417,
      "num_tokens": 76961198.0,
      "step": 34475
    },
    {
      "entropy": 5.273694181442261,
      "epoch": 3.8724658842028417,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003576219262793877,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23183956295251845,
      "num_tokens": 76973012.0,
      "step": 34480
    },
    {
      "entropy": 5.276099729537964,
      "epoch": 3.873027461110799,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035758458108969056,
      "loss": 4.8606,
      "mean_token_accuracy": 0.22686750888824464,
      "num_tokens": 76983613.0,
      "step": 34485
    },
    {
      "entropy": 5.335975790023804,
      "epoch": 3.8735890380187565,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003575472332704684,
      "loss": 4.9995,
      "mean_token_accuracy": 0.2154989242553711,
      "num_tokens": 76995471.0,
      "step": 34490
    },
    {
      "entropy": 5.2211065769195555,
      "epoch": 3.874150614926714,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035750988282291037,
      "loss": 4.786,
      "mean_token_accuracy": 0.23038092404603958,
      "num_tokens": 77005777.0,
      "step": 34495
    },
    {
      "entropy": 5.293227052688598,
      "epoch": 3.874712191834672,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00035747252974820565,
      "loss": 4.9502,
      "mean_token_accuracy": 0.21967346221208572,
      "num_tokens": 77017099.0,
      "step": 34500
    },
    {
      "entropy": 5.249154281616211,
      "epoch": 3.8752737687426295,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035743517404754354,
      "loss": 4.9442,
      "mean_token_accuracy": 0.2240229144692421,
      "num_tokens": 77028350.0,
      "step": 34505
    },
    {
      "entropy": 5.254945182800293,
      "epoch": 3.8758353456505867,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003573978157221136,
      "loss": 4.8509,
      "mean_token_accuracy": 0.23019952476024627,
      "num_tokens": 77039355.0,
      "step": 34510
    },
    {
      "entropy": 5.36250114440918,
      "epoch": 3.8763969225585444,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035736045477310517,
      "loss": 4.9785,
      "mean_token_accuracy": 0.21560013443231582,
      "num_tokens": 77049966.0,
      "step": 34515
    },
    {
      "entropy": 5.2727197170257565,
      "epoch": 3.876958499466502,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035732309120170786,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22553040087223053,
      "num_tokens": 77061102.0,
      "step": 34520
    },
    {
      "entropy": 5.233837223052978,
      "epoch": 3.8775200763744593,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035728572500911155,
      "loss": 4.9148,
      "mean_token_accuracy": 0.22419565469026564,
      "num_tokens": 77071516.0,
      "step": 34525
    },
    {
      "entropy": 5.278199958801269,
      "epoch": 3.878081653282417,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035724835619650573,
      "loss": 4.9638,
      "mean_token_accuracy": 0.21902329176664354,
      "num_tokens": 77081924.0,
      "step": 34530
    },
    {
      "entropy": 5.305172348022461,
      "epoch": 3.8786432301903746,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003572109847650804,
      "loss": 4.8439,
      "mean_token_accuracy": 0.2241001084446907,
      "num_tokens": 77093064.0,
      "step": 34535
    },
    {
      "entropy": 5.292956256866455,
      "epoch": 3.8792048070983323,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035717361071602546,
      "loss": 4.9069,
      "mean_token_accuracy": 0.2260842278599739,
      "num_tokens": 77104428.0,
      "step": 34540
    },
    {
      "entropy": 5.232142066955566,
      "epoch": 3.87976638400629,
      "grad_norm": 1.09375,
      "learning_rate": 0.00035713623405053083,
      "loss": 4.8465,
      "mean_token_accuracy": 0.2363671451807022,
      "num_tokens": 77114949.0,
      "step": 34545
    },
    {
      "entropy": 5.293532800674439,
      "epoch": 3.880327960914247,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035709885476978666,
      "loss": 4.9622,
      "mean_token_accuracy": 0.2198263317346573,
      "num_tokens": 77125178.0,
      "step": 34550
    },
    {
      "entropy": 5.185492038726807,
      "epoch": 3.880889537822205,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035706147287498303,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2311899781227112,
      "num_tokens": 77135967.0,
      "step": 34555
    },
    {
      "entropy": 5.297052717208862,
      "epoch": 3.8814511147301625,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035702408836731043,
      "loss": 4.9604,
      "mean_token_accuracy": 0.22661980390548705,
      "num_tokens": 77148052.0,
      "step": 34560
    },
    {
      "entropy": 5.283688497543335,
      "epoch": 3.8820126916381197,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000356986701247959,
      "loss": 4.8928,
      "mean_token_accuracy": 0.22823770344257355,
      "num_tokens": 77159008.0,
      "step": 34565
    },
    {
      "entropy": 5.2080999374389645,
      "epoch": 3.8825742685460773,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003569493115181192,
      "loss": 4.846,
      "mean_token_accuracy": 0.2225597769021988,
      "num_tokens": 77169846.0,
      "step": 34570
    },
    {
      "entropy": 5.221019124984741,
      "epoch": 3.883135845454035,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003569119191789815,
      "loss": 4.894,
      "mean_token_accuracy": 0.22916221916675567,
      "num_tokens": 77182881.0,
      "step": 34575
    },
    {
      "entropy": 5.254027891159057,
      "epoch": 3.8836974223619922,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035687452423173655,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22588847875595092,
      "num_tokens": 77193852.0,
      "step": 34580
    },
    {
      "entropy": 5.28880500793457,
      "epoch": 3.88425899926995,
      "grad_norm": 0.9453125,
      "learning_rate": 0.000356837126677575,
      "loss": 4.9176,
      "mean_token_accuracy": 0.22256182879209518,
      "num_tokens": 77205141.0,
      "step": 34585
    },
    {
      "entropy": 5.311837148666382,
      "epoch": 3.8848205761779075,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003567997265176876,
      "loss": 4.9268,
      "mean_token_accuracy": 0.21520335376262664,
      "num_tokens": 77217521.0,
      "step": 34590
    },
    {
      "entropy": 5.3606384754180905,
      "epoch": 3.885382153085865,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035676232375326513,
      "loss": 4.9981,
      "mean_token_accuracy": 0.2125537857413292,
      "num_tokens": 77227468.0,
      "step": 34595
    },
    {
      "entropy": 5.1964113235473635,
      "epoch": 3.885943729993823,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035672491838549867,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22546356171369553,
      "num_tokens": 77237908.0,
      "step": 34600
    },
    {
      "entropy": 5.169530773162842,
      "epoch": 3.88650530690178,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000356687510415579,
      "loss": 4.7908,
      "mean_token_accuracy": 0.22993137836456298,
      "num_tokens": 77248621.0,
      "step": 34605
    },
    {
      "entropy": 5.313137149810791,
      "epoch": 3.8870668838097377,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003566500998446973,
      "loss": 4.9318,
      "mean_token_accuracy": 0.22077716290950775,
      "num_tokens": 77260261.0,
      "step": 34610
    },
    {
      "entropy": 5.3024329662323,
      "epoch": 3.8876284607176954,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035661268667404485,
      "loss": 4.8487,
      "mean_token_accuracy": 0.22355639636516572,
      "num_tokens": 77270985.0,
      "step": 34615
    },
    {
      "entropy": 5.249892902374268,
      "epoch": 3.8881900376256526,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003565752709048127,
      "loss": 4.9137,
      "mean_token_accuracy": 0.22862082570791245,
      "num_tokens": 77281876.0,
      "step": 34620
    },
    {
      "entropy": 5.177705478668213,
      "epoch": 3.8887516145336103,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003565378525381924,
      "loss": 4.7826,
      "mean_token_accuracy": 0.2286886215209961,
      "num_tokens": 77293141.0,
      "step": 34625
    },
    {
      "entropy": 5.272109031677246,
      "epoch": 3.889313191441568,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003565004315753752,
      "loss": 4.8757,
      "mean_token_accuracy": 0.23591670989990235,
      "num_tokens": 77305723.0,
      "step": 34630
    },
    {
      "entropy": 5.17573356628418,
      "epoch": 3.8898747683495256,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035646300801755266,
      "loss": 4.8108,
      "mean_token_accuracy": 0.23453576862812042,
      "num_tokens": 77317736.0,
      "step": 34635
    },
    {
      "entropy": 5.240825748443603,
      "epoch": 3.8904363452574833,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035642558186591635,
      "loss": 4.8796,
      "mean_token_accuracy": 0.2219565451145172,
      "num_tokens": 77328293.0,
      "step": 34640
    },
    {
      "entropy": 5.263805723190307,
      "epoch": 3.8909979221654405,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003563881531216579,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2220342457294464,
      "num_tokens": 77339371.0,
      "step": 34645
    },
    {
      "entropy": 5.307437324523926,
      "epoch": 3.891559499073398,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003563507217859691,
      "loss": 4.9329,
      "mean_token_accuracy": 0.2146347790956497,
      "num_tokens": 77350234.0,
      "step": 34650
    },
    {
      "entropy": 5.267419767379761,
      "epoch": 3.892121075981356,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003563132878600419,
      "loss": 4.9042,
      "mean_token_accuracy": 0.22700024098157884,
      "num_tokens": 77361144.0,
      "step": 34655
    },
    {
      "entropy": 5.269934415817261,
      "epoch": 3.892682652889313,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003562758513450678,
      "loss": 4.8988,
      "mean_token_accuracy": 0.2297837495803833,
      "num_tokens": 77372090.0,
      "step": 34660
    },
    {
      "entropy": 5.301858043670654,
      "epoch": 3.8932442297972707,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003562384122422393,
      "loss": 4.9749,
      "mean_token_accuracy": 0.2199007034301758,
      "num_tokens": 77382974.0,
      "step": 34665
    },
    {
      "entropy": 5.268221855163574,
      "epoch": 3.8938058067052284,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003562009705527481,
      "loss": 4.9116,
      "mean_token_accuracy": 0.2225977212190628,
      "num_tokens": 77395971.0,
      "step": 34670
    },
    {
      "entropy": 5.3070519924163815,
      "epoch": 3.8943673836131856,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035616352627778665,
      "loss": 4.8789,
      "mean_token_accuracy": 0.21904073655605316,
      "num_tokens": 77405776.0,
      "step": 34675
    },
    {
      "entropy": 5.298217248916626,
      "epoch": 3.8949289605211432,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035612607941854695,
      "loss": 4.9187,
      "mean_token_accuracy": 0.22043498754501342,
      "num_tokens": 77417668.0,
      "step": 34680
    },
    {
      "entropy": 5.247859239578247,
      "epoch": 3.895490537429101,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003560886299762214,
      "loss": 4.8614,
      "mean_token_accuracy": 0.22655385732650757,
      "num_tokens": 77428581.0,
      "step": 34685
    },
    {
      "entropy": 5.293102836608886,
      "epoch": 3.8960521143370586,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003560511779520025,
      "loss": 4.9433,
      "mean_token_accuracy": 0.22395906299352647,
      "num_tokens": 77441587.0,
      "step": 34690
    },
    {
      "entropy": 5.287802362442017,
      "epoch": 3.896613691245016,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003560137233470826,
      "loss": 4.854,
      "mean_token_accuracy": 0.2281571224331856,
      "num_tokens": 77452579.0,
      "step": 34695
    },
    {
      "entropy": 5.167351007461548,
      "epoch": 3.8971752681529734,
      "grad_norm": 1.0,
      "learning_rate": 0.00035597626616265434,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22356649339199067,
      "num_tokens": 77464465.0,
      "step": 34700
    },
    {
      "entropy": 5.191482830047607,
      "epoch": 3.897736845060931,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00035593880639991037,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22402188926935196,
      "num_tokens": 77474944.0,
      "step": 34705
    },
    {
      "entropy": 5.200306415557861,
      "epoch": 3.8982984219688888,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035590134406004336,
      "loss": 4.7343,
      "mean_token_accuracy": 0.23508620858192444,
      "num_tokens": 77485389.0,
      "step": 34710
    },
    {
      "entropy": 5.227944087982178,
      "epoch": 3.898859998876846,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003558638791442463,
      "loss": 4.8228,
      "mean_token_accuracy": 0.2274376004934311,
      "num_tokens": 77496834.0,
      "step": 34715
    },
    {
      "entropy": 5.144000816345215,
      "epoch": 3.8994215757848036,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003558264116537119,
      "loss": 4.7932,
      "mean_token_accuracy": 0.22749113589525222,
      "num_tokens": 77508123.0,
      "step": 34720
    },
    {
      "entropy": 5.241891384124756,
      "epoch": 3.8999831526927613,
      "grad_norm": 0.91015625,
      "learning_rate": 0.0003557889415896332,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22035575956106185,
      "num_tokens": 77519599.0,
      "step": 34725
    },
    {
      "entropy": 5.260749816894531,
      "epoch": 3.900544729600719,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035575146895320327,
      "loss": 4.863,
      "mean_token_accuracy": 0.22491923421621324,
      "num_tokens": 77532241.0,
      "step": 34730
    },
    {
      "entropy": 5.30978307723999,
      "epoch": 3.9011063065086766,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003557139937456152,
      "loss": 4.9347,
      "mean_token_accuracy": 0.22650652080774308,
      "num_tokens": 77542702.0,
      "step": 34735
    },
    {
      "entropy": 5.1969400405883786,
      "epoch": 3.901667883416634,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00035567651596806236,
      "loss": 4.9216,
      "mean_token_accuracy": 0.21674414724111557,
      "num_tokens": 77554235.0,
      "step": 34740
    },
    {
      "entropy": 5.224600219726563,
      "epoch": 3.9022294603245915,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035563903562173786,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23101520389318467,
      "num_tokens": 77565704.0,
      "step": 34745
    },
    {
      "entropy": 5.225348472595215,
      "epoch": 3.902791037232549,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035560155270783527,
      "loss": 4.8396,
      "mean_token_accuracy": 0.22277108132839202,
      "num_tokens": 77576524.0,
      "step": 34750
    },
    {
      "entropy": 5.233689403533935,
      "epoch": 3.9033526141405064,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003555640672275479,
      "loss": 4.8678,
      "mean_token_accuracy": 0.2312922164797783,
      "num_tokens": 77587292.0,
      "step": 34755
    },
    {
      "entropy": 5.2572225570678714,
      "epoch": 3.903914191048464,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003555265791820695,
      "loss": 4.8946,
      "mean_token_accuracy": 0.21854602992534639,
      "num_tokens": 77597958.0,
      "step": 34760
    },
    {
      "entropy": 5.3528361320495605,
      "epoch": 3.9044757679564217,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003554890885725935,
      "loss": 4.9869,
      "mean_token_accuracy": 0.22217700779438018,
      "num_tokens": 77609264.0,
      "step": 34765
    },
    {
      "entropy": 5.3847975730896,
      "epoch": 3.905037344864379,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003554515954003137,
      "loss": 4.9524,
      "mean_token_accuracy": 0.22345019727945328,
      "num_tokens": 77621224.0,
      "step": 34770
    },
    {
      "entropy": 5.262592124938965,
      "epoch": 3.9055989217723366,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003554140996664239,
      "loss": 4.927,
      "mean_token_accuracy": 0.22102377712726592,
      "num_tokens": 77632009.0,
      "step": 34775
    },
    {
      "entropy": 5.203677034378051,
      "epoch": 3.9061604986802942,
      "grad_norm": 1.0,
      "learning_rate": 0.000355376601372118,
      "loss": 4.877,
      "mean_token_accuracy": 0.224220871925354,
      "num_tokens": 77643094.0,
      "step": 34780
    },
    {
      "entropy": 5.272225379943848,
      "epoch": 3.906722075588252,
      "grad_norm": 1.0,
      "learning_rate": 0.0003553391005185899,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2315453454852104,
      "num_tokens": 77654049.0,
      "step": 34785
    },
    {
      "entropy": 5.24551362991333,
      "epoch": 3.9072836524962096,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003553015971070338,
      "loss": 4.833,
      "mean_token_accuracy": 0.23042436987161635,
      "num_tokens": 77664820.0,
      "step": 34790
    },
    {
      "entropy": 5.322029399871826,
      "epoch": 3.907845229404167,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003552640911386436,
      "loss": 4.9776,
      "mean_token_accuracy": 0.21945967376232148,
      "num_tokens": 77675107.0,
      "step": 34795
    },
    {
      "entropy": 5.210684108734131,
      "epoch": 3.9084068063121244,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035522658261461366,
      "loss": 4.8441,
      "mean_token_accuracy": 0.2232692837715149,
      "num_tokens": 77685487.0,
      "step": 34800
    },
    {
      "entropy": 5.26816029548645,
      "epoch": 3.908968383220082,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003551890715361382,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22667221277952193,
      "num_tokens": 77697391.0,
      "step": 34805
    },
    {
      "entropy": 5.269564628601074,
      "epoch": 3.9095299601280393,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003551515579044116,
      "loss": 4.8579,
      "mean_token_accuracy": 0.22697050422430037,
      "num_tokens": 77707270.0,
      "step": 34810
    },
    {
      "entropy": 5.141001224517822,
      "epoch": 3.910091537035997,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035511404172062827,
      "loss": 4.7726,
      "mean_token_accuracy": 0.23352750688791274,
      "num_tokens": 77717164.0,
      "step": 34815
    },
    {
      "entropy": 5.345995235443115,
      "epoch": 3.9106531139439547,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003550765229859828,
      "loss": 4.935,
      "mean_token_accuracy": 0.22459143549203872,
      "num_tokens": 77728142.0,
      "step": 34820
    },
    {
      "entropy": 5.306021213531494,
      "epoch": 3.9112146908519123,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003550390017016698,
      "loss": 4.9424,
      "mean_token_accuracy": 0.23303065001964568,
      "num_tokens": 77740162.0,
      "step": 34825
    },
    {
      "entropy": 5.227339744567871,
      "epoch": 3.91177626775987,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035500147786888387,
      "loss": 4.8418,
      "mean_token_accuracy": 0.2358763724565506,
      "num_tokens": 77751574.0,
      "step": 34830
    },
    {
      "entropy": 5.283274698257446,
      "epoch": 3.912337844667827,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035496395148881994,
      "loss": 4.9475,
      "mean_token_accuracy": 0.2185026451945305,
      "num_tokens": 77762919.0,
      "step": 34835
    },
    {
      "entropy": 5.2328277111053465,
      "epoch": 3.912899421575785,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035492642256267277,
      "loss": 4.8852,
      "mean_token_accuracy": 0.22656481415033342,
      "num_tokens": 77774578.0,
      "step": 34840
    },
    {
      "entropy": 5.290378999710083,
      "epoch": 3.9134609984837425,
      "grad_norm": 1.03125,
      "learning_rate": 0.00035488889109163725,
      "loss": 4.9127,
      "mean_token_accuracy": 0.22004263699054719,
      "num_tokens": 77784872.0,
      "step": 34845
    },
    {
      "entropy": 5.249824380874633,
      "epoch": 3.9140225753916997,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003548513570769085,
      "loss": 4.8463,
      "mean_token_accuracy": 0.22568560838699342,
      "num_tokens": 77796159.0,
      "step": 34850
    },
    {
      "entropy": 5.2018718242645265,
      "epoch": 3.9145841522996574,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003548138205196816,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23166735768318175,
      "num_tokens": 77805779.0,
      "step": 34855
    },
    {
      "entropy": 5.266987228393555,
      "epoch": 3.915145729207615,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003547762814211517,
      "loss": 4.9445,
      "mean_token_accuracy": 0.22481747418642045,
      "num_tokens": 77818102.0,
      "step": 34860
    },
    {
      "entropy": 5.271075439453125,
      "epoch": 3.9157073061155723,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000354738739782514,
      "loss": 4.936,
      "mean_token_accuracy": 0.21762678623199463,
      "num_tokens": 77830107.0,
      "step": 34865
    },
    {
      "entropy": 5.318042993545532,
      "epoch": 3.91626888302353,
      "grad_norm": 1.109375,
      "learning_rate": 0.000354701195604964,
      "loss": 4.9508,
      "mean_token_accuracy": 0.22376483529806138,
      "num_tokens": 77841388.0,
      "step": 34870
    },
    {
      "entropy": 5.289633703231812,
      "epoch": 3.9168304599314876,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003546636488896969,
      "loss": 4.9448,
      "mean_token_accuracy": 0.22115990817546843,
      "num_tokens": 77853418.0,
      "step": 34875
    },
    {
      "entropy": 5.3171813011169435,
      "epoch": 3.9173920368394453,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003546260996379083,
      "loss": 4.9001,
      "mean_token_accuracy": 0.22329031974077224,
      "num_tokens": 77865019.0,
      "step": 34880
    },
    {
      "entropy": 5.247244548797608,
      "epoch": 3.917953613747403,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000354588547850794,
      "loss": 4.8248,
      "mean_token_accuracy": 0.22881938368082047,
      "num_tokens": 77876546.0,
      "step": 34885
    },
    {
      "entropy": 5.313500452041626,
      "epoch": 3.91851519065536,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003545509935295493,
      "loss": 4.9585,
      "mean_token_accuracy": 0.22437463104724883,
      "num_tokens": 77888081.0,
      "step": 34890
    },
    {
      "entropy": 5.176113176345825,
      "epoch": 3.919076767563318,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003545134366753701,
      "loss": 4.7648,
      "mean_token_accuracy": 0.23324240744113922,
      "num_tokens": 77900106.0,
      "step": 34895
    },
    {
      "entropy": 5.2259715557098385,
      "epoch": 3.9196383444712755,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003544758772894523,
      "loss": 4.8196,
      "mean_token_accuracy": 0.2314447745680809,
      "num_tokens": 77910452.0,
      "step": 34900
    },
    {
      "entropy": 5.317700290679932,
      "epoch": 3.9201999213792327,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00035443831537299177,
      "loss": 4.9336,
      "mean_token_accuracy": 0.22499697059392929,
      "num_tokens": 77921826.0,
      "step": 34905
    },
    {
      "entropy": 5.391689109802246,
      "epoch": 3.9207614982871903,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003544007509271844,
      "loss": 4.9859,
      "mean_token_accuracy": 0.21957776844501495,
      "num_tokens": 77933909.0,
      "step": 34910
    },
    {
      "entropy": 5.3422609806060795,
      "epoch": 3.921323075195148,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035436318395322636,
      "loss": 4.9411,
      "mean_token_accuracy": 0.22274529486894606,
      "num_tokens": 77945232.0,
      "step": 34915
    },
    {
      "entropy": 5.236718225479126,
      "epoch": 3.9218846521031057,
      "grad_norm": 1.0,
      "learning_rate": 0.0003543256144523137,
      "loss": 4.9007,
      "mean_token_accuracy": 0.22434057146310807,
      "num_tokens": 77957305.0,
      "step": 34920
    },
    {
      "entropy": 5.1929014205932615,
      "epoch": 3.9224462290110633,
      "grad_norm": 1.0,
      "learning_rate": 0.0003542880424256428,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2295893669128418,
      "num_tokens": 77968036.0,
      "step": 34925
    },
    {
      "entropy": 5.23861289024353,
      "epoch": 3.9230078059190205,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003542504678744099,
      "loss": 4.8241,
      "mean_token_accuracy": 0.22720644176006316,
      "num_tokens": 77978826.0,
      "step": 34930
    },
    {
      "entropy": 5.234014844894409,
      "epoch": 3.923569382826978,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003542128907998113,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22582610696554184,
      "num_tokens": 77989820.0,
      "step": 34935
    },
    {
      "entropy": 5.239054441452026,
      "epoch": 3.924130959734936,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003541753112030435,
      "loss": 4.8895,
      "mean_token_accuracy": 0.22748371660709382,
      "num_tokens": 78000114.0,
      "step": 34940
    },
    {
      "entropy": 5.201059103012085,
      "epoch": 3.924692536642893,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035413772908530317,
      "loss": 4.8348,
      "mean_token_accuracy": 0.22840785831212998,
      "num_tokens": 78010680.0,
      "step": 34945
    },
    {
      "entropy": 5.256719970703125,
      "epoch": 3.9252541135508507,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003541001444477869,
      "loss": 4.8494,
      "mean_token_accuracy": 0.22396430373191833,
      "num_tokens": 78022172.0,
      "step": 34950
    },
    {
      "entropy": 5.285445928573608,
      "epoch": 3.9258156904588084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035406255729169124,
      "loss": 4.8721,
      "mean_token_accuracy": 0.223747093975544,
      "num_tokens": 78033535.0,
      "step": 34955
    },
    {
      "entropy": 5.262641477584839,
      "epoch": 3.9263772673667656,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003540249676182131,
      "loss": 4.8762,
      "mean_token_accuracy": 0.2197497606277466,
      "num_tokens": 78044555.0,
      "step": 34960
    },
    {
      "entropy": 5.247398805618286,
      "epoch": 3.9269388442747233,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035398737542854936,
      "loss": 4.8238,
      "mean_token_accuracy": 0.23019593060016633,
      "num_tokens": 78056078.0,
      "step": 34965
    },
    {
      "entropy": 5.266637229919434,
      "epoch": 3.927500421182681,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003539497807238969,
      "loss": 4.8848,
      "mean_token_accuracy": 0.2257019802927971,
      "num_tokens": 78068324.0,
      "step": 34970
    },
    {
      "entropy": 5.178621387481689,
      "epoch": 3.9280619980906386,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035391218350545287,
      "loss": 4.802,
      "mean_token_accuracy": 0.22598838657140732,
      "num_tokens": 78080013.0,
      "step": 34975
    },
    {
      "entropy": 5.187977981567383,
      "epoch": 3.9286235749985963,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00035387458377441434,
      "loss": 4.8879,
      "mean_token_accuracy": 0.22399178743362427,
      "num_tokens": 78091494.0,
      "step": 34980
    },
    {
      "entropy": 5.313629007339477,
      "epoch": 3.9291851519065535,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003538369815319784,
      "loss": 4.8252,
      "mean_token_accuracy": 0.2238859936594963,
      "num_tokens": 78102037.0,
      "step": 34985
    },
    {
      "entropy": 5.28602089881897,
      "epoch": 3.929746728814511,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035379937677934245,
      "loss": 4.9278,
      "mean_token_accuracy": 0.2203660562634468,
      "num_tokens": 78112651.0,
      "step": 34990
    },
    {
      "entropy": 5.252895784378052,
      "epoch": 3.930308305722469,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003537617695177037,
      "loss": 4.9267,
      "mean_token_accuracy": 0.22347216606140136,
      "num_tokens": 78123273.0,
      "step": 34995
    },
    {
      "entropy": 5.282237529754639,
      "epoch": 3.930869882630426,
      "grad_norm": 1.125,
      "learning_rate": 0.00035372415974825974,
      "loss": 4.9485,
      "mean_token_accuracy": 0.21576579809188842,
      "num_tokens": 78133868.0,
      "step": 35000
    },
    {
      "entropy": 5.331671047210693,
      "epoch": 3.9314314595383837,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035368654747220795,
      "loss": 4.9245,
      "mean_token_accuracy": 0.21587336510419847,
      "num_tokens": 78145001.0,
      "step": 35005
    },
    {
      "entropy": 5.31006383895874,
      "epoch": 3.9319930364463413,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00035364893269074606,
      "loss": 4.8712,
      "mean_token_accuracy": 0.22370219528675078,
      "num_tokens": 78155699.0,
      "step": 35010
    },
    {
      "entropy": 5.158105564117432,
      "epoch": 3.932554613354299,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035361131540507157,
      "loss": 4.702,
      "mean_token_accuracy": 0.23247910141944886,
      "num_tokens": 78165941.0,
      "step": 35015
    },
    {
      "entropy": 5.207322454452514,
      "epoch": 3.9331161902622567,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003535736956163823,
      "loss": 4.888,
      "mean_token_accuracy": 0.23352362513542174,
      "num_tokens": 78177049.0,
      "step": 35020
    },
    {
      "entropy": 5.303995132446289,
      "epoch": 3.933677767170214,
      "grad_norm": 1.0,
      "learning_rate": 0.00035353607332587613,
      "loss": 4.9953,
      "mean_token_accuracy": 0.21975812464952468,
      "num_tokens": 78188834.0,
      "step": 35025
    },
    {
      "entropy": 5.342876243591308,
      "epoch": 3.9342393440781716,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003534984485347509,
      "loss": 4.9183,
      "mean_token_accuracy": 0.23026802986860276,
      "num_tokens": 78199847.0,
      "step": 35030
    },
    {
      "entropy": 5.112040853500366,
      "epoch": 3.934800920986129,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003534608212442046,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23212774246931075,
      "num_tokens": 78210747.0,
      "step": 35035
    },
    {
      "entropy": 5.202830743789673,
      "epoch": 3.9353624978940864,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035342319145543533,
      "loss": 4.9096,
      "mean_token_accuracy": 0.2263314187526703,
      "num_tokens": 78221711.0,
      "step": 35040
    },
    {
      "entropy": 5.381248188018799,
      "epoch": 3.935924074802044,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035338555916964126,
      "loss": 5.0446,
      "mean_token_accuracy": 0.21339091956615447,
      "num_tokens": 78233957.0,
      "step": 35045
    },
    {
      "entropy": 5.2505419731140135,
      "epoch": 3.9364856517100018,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035334792438802056,
      "loss": 4.9113,
      "mean_token_accuracy": 0.22240829467773438,
      "num_tokens": 78245403.0,
      "step": 35050
    },
    {
      "entropy": 5.291314554214478,
      "epoch": 3.937047228617959,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003533102871117716,
      "loss": 4.8539,
      "mean_token_accuracy": 0.22307586073875427,
      "num_tokens": 78256062.0,
      "step": 35055
    },
    {
      "entropy": 5.306577539443969,
      "epoch": 3.9376088055259166,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003532726473420927,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22252924144268035,
      "num_tokens": 78265895.0,
      "step": 35060
    },
    {
      "entropy": 5.25228066444397,
      "epoch": 3.9381703824338743,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003532350050801823,
      "loss": 4.8723,
      "mean_token_accuracy": 0.22203131169080734,
      "num_tokens": 78276752.0,
      "step": 35065
    },
    {
      "entropy": 5.212960433959961,
      "epoch": 3.938731959341832,
      "grad_norm": 1.078125,
      "learning_rate": 0.00035319736032723905,
      "loss": 4.8932,
      "mean_token_accuracy": 0.2287276268005371,
      "num_tokens": 78288357.0,
      "step": 35070
    },
    {
      "entropy": 5.257140493392944,
      "epoch": 3.9392935362497896,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003531597130844616,
      "loss": 4.8478,
      "mean_token_accuracy": 0.23096616119146346,
      "num_tokens": 78299558.0,
      "step": 35075
    },
    {
      "entropy": 5.362671709060669,
      "epoch": 3.939855113157747,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035312206335304843,
      "loss": 4.8763,
      "mean_token_accuracy": 0.22549673318862914,
      "num_tokens": 78311443.0,
      "step": 35080
    },
    {
      "entropy": 5.29826397895813,
      "epoch": 3.9404166900657045,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00035308441113419855,
      "loss": 4.9603,
      "mean_token_accuracy": 0.2199905037879944,
      "num_tokens": 78322575.0,
      "step": 35085
    },
    {
      "entropy": 5.239525842666626,
      "epoch": 3.940978266973662,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00035304675642911074,
      "loss": 4.8687,
      "mean_token_accuracy": 0.2244923770427704,
      "num_tokens": 78333583.0,
      "step": 35090
    },
    {
      "entropy": 5.24848895072937,
      "epoch": 3.9415398438816194,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003530090992389838,
      "loss": 4.8556,
      "mean_token_accuracy": 0.21781325787305833,
      "num_tokens": 78343339.0,
      "step": 35095
    },
    {
      "entropy": 5.264852380752563,
      "epoch": 3.942101420789577,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000352971439565017,
      "loss": 4.8234,
      "mean_token_accuracy": 0.22624384015798568,
      "num_tokens": 78354613.0,
      "step": 35100
    },
    {
      "entropy": 5.226109027862549,
      "epoch": 3.9426629976975347,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035293377740840934,
      "loss": 4.86,
      "mean_token_accuracy": 0.2265414133667946,
      "num_tokens": 78365105.0,
      "step": 35105
    },
    {
      "entropy": 5.193156385421753,
      "epoch": 3.9432245746054924,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035289611277036,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23303317576646804,
      "num_tokens": 78375311.0,
      "step": 35110
    },
    {
      "entropy": 5.2025360584259035,
      "epoch": 3.94378615151345,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003528584456520681,
      "loss": 4.8082,
      "mean_token_accuracy": 0.23405503034591674,
      "num_tokens": 78386653.0,
      "step": 35115
    },
    {
      "entropy": 5.128850221633911,
      "epoch": 3.9443477284214072,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035282077605473317,
      "loss": 4.8144,
      "mean_token_accuracy": 0.22905107736587524,
      "num_tokens": 78397296.0,
      "step": 35120
    },
    {
      "entropy": 5.240899324417114,
      "epoch": 3.944909305329365,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003527831039795545,
      "loss": 4.8123,
      "mean_token_accuracy": 0.23395518958568573,
      "num_tokens": 78408825.0,
      "step": 35125
    },
    {
      "entropy": 5.186050510406494,
      "epoch": 3.9454708822373226,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003527454294277316,
      "loss": 4.8236,
      "mean_token_accuracy": 0.23327605277299882,
      "num_tokens": 78419895.0,
      "step": 35130
    },
    {
      "entropy": 5.210085916519165,
      "epoch": 3.94603245914528,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003527077524004642,
      "loss": 4.7252,
      "mean_token_accuracy": 0.2307958722114563,
      "num_tokens": 78430111.0,
      "step": 35135
    },
    {
      "entropy": 5.230750703811646,
      "epoch": 3.9465940360532374,
      "grad_norm": 1.046875,
      "learning_rate": 0.00035267007289895176,
      "loss": 4.9151,
      "mean_token_accuracy": 0.22289850264787675,
      "num_tokens": 78441778.0,
      "step": 35140
    },
    {
      "entropy": 5.2363639831542965,
      "epoch": 3.947155612961195,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00035263239092439417,
      "loss": 4.8011,
      "mean_token_accuracy": 0.22971824258565904,
      "num_tokens": 78451930.0,
      "step": 35145
    },
    {
      "entropy": 5.213479709625244,
      "epoch": 3.9477171898691523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00035259470647799104,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23016582280397416,
      "num_tokens": 78463002.0,
      "step": 35150
    },
    {
      "entropy": 5.171776533126831,
      "epoch": 3.94827876677711,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00035255701956094236,
      "loss": 4.8704,
      "mean_token_accuracy": 0.22462131679058076,
      "num_tokens": 78473573.0,
      "step": 35155
    },
    {
      "entropy": 5.266355085372925,
      "epoch": 3.9488403436850676,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00035251933017444814,
      "loss": 4.8928,
      "mean_token_accuracy": 0.21832499355077745,
      "num_tokens": 78485336.0,
      "step": 35160
    },
    {
      "entropy": 5.3003088474273685,
      "epoch": 3.9494019205930253,
      "grad_norm": 0.953125,
      "learning_rate": 0.00035248163831970837,
      "loss": 4.8613,
      "mean_token_accuracy": 0.22739288210868835,
      "num_tokens": 78497801.0,
      "step": 35165
    },
    {
      "entropy": 5.215439128875732,
      "epoch": 3.949963497500983,
      "grad_norm": 1.015625,
      "learning_rate": 0.00035244394399792304,
      "loss": 4.8721,
      "mean_token_accuracy": 0.2266494408249855,
      "num_tokens": 78509288.0,
      "step": 35170
    },
    {
      "entropy": 5.22067346572876,
      "epoch": 3.95052507440894,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003524062472102927,
      "loss": 4.7887,
      "mean_token_accuracy": 0.2291015386581421,
      "num_tokens": 78519871.0,
      "step": 35175
    },
    {
      "entropy": 5.2226818084716795,
      "epoch": 3.951086651316898,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003523685479580174,
      "loss": 4.8957,
      "mean_token_accuracy": 0.2245445266366005,
      "num_tokens": 78530287.0,
      "step": 35180
    },
    {
      "entropy": 5.242900609970093,
      "epoch": 3.9516482282248555,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003523308462422974,
      "loss": 4.8529,
      "mean_token_accuracy": 0.22251611799001694,
      "num_tokens": 78542341.0,
      "step": 35185
    },
    {
      "entropy": 5.197686576843262,
      "epoch": 3.9522098051328127,
      "grad_norm": 1.125,
      "learning_rate": 0.0003522931420643334,
      "loss": 4.7667,
      "mean_token_accuracy": 0.23406655788421632,
      "num_tokens": 78552533.0,
      "step": 35190
    },
    {
      "entropy": 5.210479784011841,
      "epoch": 3.9527713820407704,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003522554354253257,
      "loss": 4.8352,
      "mean_token_accuracy": 0.22702159434556962,
      "num_tokens": 78564007.0,
      "step": 35195
    },
    {
      "entropy": 5.220238876342774,
      "epoch": 3.953332958948728,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035221772632647503,
      "loss": 4.8942,
      "mean_token_accuracy": 0.2186807230114937,
      "num_tokens": 78574015.0,
      "step": 35200
    },
    {
      "entropy": 5.341465282440185,
      "epoch": 3.9538945358566857,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035218001476898194,
      "loss": 4.9489,
      "mean_token_accuracy": 0.22227212339639663,
      "num_tokens": 78586230.0,
      "step": 35205
    },
    {
      "entropy": 5.259969758987427,
      "epoch": 3.9544561127646434,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003521423007540472,
      "loss": 4.8979,
      "mean_token_accuracy": 0.22653911411762237,
      "num_tokens": 78596819.0,
      "step": 35210
    },
    {
      "entropy": 5.239494132995605,
      "epoch": 3.9550176896726006,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003521045842828718,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22379339933395387,
      "num_tokens": 78608788.0,
      "step": 35215
    },
    {
      "entropy": 5.247535324096679,
      "epoch": 3.9555792665805583,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003520668653566565,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22605569064617156,
      "num_tokens": 78619124.0,
      "step": 35220
    },
    {
      "entropy": 5.360646390914917,
      "epoch": 3.956140843488516,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003520291439766024,
      "loss": 4.9512,
      "mean_token_accuracy": 0.22764596939086915,
      "num_tokens": 78630799.0,
      "step": 35225
    },
    {
      "entropy": 5.148532056808472,
      "epoch": 3.956702420396473,
      "grad_norm": 0.90625,
      "learning_rate": 0.00035199142014391034,
      "loss": 4.7855,
      "mean_token_accuracy": 0.23173086047172547,
      "num_tokens": 78641664.0,
      "step": 35230
    },
    {
      "entropy": 5.257238578796387,
      "epoch": 3.957263997304431,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00035195369385978165,
      "loss": 4.8637,
      "mean_token_accuracy": 0.2303938090801239,
      "num_tokens": 78652025.0,
      "step": 35235
    },
    {
      "entropy": 5.315252208709717,
      "epoch": 3.9578255742123885,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035191596512541747,
      "loss": 4.8716,
      "mean_token_accuracy": 0.23460220843553542,
      "num_tokens": 78663539.0,
      "step": 35240
    },
    {
      "entropy": 5.23100790977478,
      "epoch": 3.9583871511203457,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003518782339420192,
      "loss": 4.8145,
      "mean_token_accuracy": 0.22181666344404222,
      "num_tokens": 78674398.0,
      "step": 35245
    },
    {
      "entropy": 5.202570867538452,
      "epoch": 3.9589487280283033,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00035184050031078815,
      "loss": 4.908,
      "mean_token_accuracy": 0.21731058806180953,
      "num_tokens": 78685945.0,
      "step": 35250
    },
    {
      "entropy": 5.184304809570312,
      "epoch": 3.959510304936261,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035180276423292566,
      "loss": 4.8332,
      "mean_token_accuracy": 0.2222583919763565,
      "num_tokens": 78696664.0,
      "step": 35255
    },
    {
      "entropy": 5.232047176361084,
      "epoch": 3.9600718818442187,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003517650257096334,
      "loss": 4.7498,
      "mean_token_accuracy": 0.237975013256073,
      "num_tokens": 78706812.0,
      "step": 35260
    },
    {
      "entropy": 5.190837669372558,
      "epoch": 3.9606334587521763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00035172728474211306,
      "loss": 4.8188,
      "mean_token_accuracy": 0.23395814150571823,
      "num_tokens": 78716657.0,
      "step": 35265
    },
    {
      "entropy": 5.246962261199951,
      "epoch": 3.9611950356601335,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035168954133156613,
      "loss": 4.8933,
      "mean_token_accuracy": 0.22073846459388732,
      "num_tokens": 78728313.0,
      "step": 35270
    },
    {
      "entropy": 5.257748317718506,
      "epoch": 3.961756612568091,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003516517954791944,
      "loss": 4.8739,
      "mean_token_accuracy": 0.2277812570333481,
      "num_tokens": 78739888.0,
      "step": 35275
    },
    {
      "entropy": 5.234655809402466,
      "epoch": 3.962318189476049,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035161404718619986,
      "loss": 4.8574,
      "mean_token_accuracy": 0.22829244583845137,
      "num_tokens": 78750624.0,
      "step": 35280
    },
    {
      "entropy": 5.251072883605957,
      "epoch": 3.962879766384006,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003515762964537844,
      "loss": 4.8324,
      "mean_token_accuracy": 0.2258531242609024,
      "num_tokens": 78762506.0,
      "step": 35285
    },
    {
      "entropy": 5.256919908523559,
      "epoch": 3.9634413432919637,
      "grad_norm": 1.125,
      "learning_rate": 0.0003515385432831499,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22235438525676726,
      "num_tokens": 78772565.0,
      "step": 35290
    },
    {
      "entropy": 5.1920207977294925,
      "epoch": 3.9640029201999214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003515007876754984,
      "loss": 4.7867,
      "mean_token_accuracy": 0.2231651246547699,
      "num_tokens": 78784107.0,
      "step": 35295
    },
    {
      "entropy": 5.294389009475708,
      "epoch": 3.964564497107879,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003514630296320323,
      "loss": 4.928,
      "mean_token_accuracy": 0.22356996089220046,
      "num_tokens": 78796182.0,
      "step": 35300
    },
    {
      "entropy": 5.292371797561645,
      "epoch": 3.9651260740158367,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003514252691539536,
      "loss": 4.9174,
      "mean_token_accuracy": 0.22386641055345535,
      "num_tokens": 78806620.0,
      "step": 35305
    },
    {
      "entropy": 5.208851432800293,
      "epoch": 3.965687650923794,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00035138750624246466,
      "loss": 4.7898,
      "mean_token_accuracy": 0.2358657658100128,
      "num_tokens": 78817034.0,
      "step": 35310
    },
    {
      "entropy": 5.2406778812408445,
      "epoch": 3.9662492278317516,
      "grad_norm": 1.2578125,
      "learning_rate": 0.000351349740898768,
      "loss": 4.8683,
      "mean_token_accuracy": 0.22493952959775926,
      "num_tokens": 78828974.0,
      "step": 35315
    },
    {
      "entropy": 5.294817686080933,
      "epoch": 3.9668108047397093,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003513119731240659,
      "loss": 4.9008,
      "mean_token_accuracy": 0.2271907091140747,
      "num_tokens": 78840514.0,
      "step": 35320
    },
    {
      "entropy": 5.241962909698486,
      "epoch": 3.9673723816476665,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00035127420291956104,
      "loss": 4.8983,
      "mean_token_accuracy": 0.23109278976917266,
      "num_tokens": 78850839.0,
      "step": 35325
    },
    {
      "entropy": 5.222212028503418,
      "epoch": 3.967933958555624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003512364302864559,
      "loss": 4.9005,
      "mean_token_accuracy": 0.2267572373151779,
      "num_tokens": 78862931.0,
      "step": 35330
    },
    {
      "entropy": 5.267194938659668,
      "epoch": 3.968495535463582,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003511986552259533,
      "loss": 4.8274,
      "mean_token_accuracy": 0.22273052781820296,
      "num_tokens": 78872324.0,
      "step": 35335
    },
    {
      "entropy": 5.212417793273926,
      "epoch": 3.969057112371539,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00035116087773925594,
      "loss": 4.8072,
      "mean_token_accuracy": 0.22678662985563278,
      "num_tokens": 78884075.0,
      "step": 35340
    },
    {
      "entropy": 5.140532445907593,
      "epoch": 3.9696186892794967,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003511230978275667,
      "loss": 4.7645,
      "mean_token_accuracy": 0.2324545592069626,
      "num_tokens": 78894804.0,
      "step": 35345
    },
    {
      "entropy": 5.322742176055908,
      "epoch": 3.9701802661874543,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00035108531549208846,
      "loss": 4.928,
      "mean_token_accuracy": 0.21577457934617997,
      "num_tokens": 78906069.0,
      "step": 35350
    },
    {
      "entropy": 5.4165199279785154,
      "epoch": 3.970741843095412,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00035104753073402433,
      "loss": 5.007,
      "mean_token_accuracy": 0.22035336047410964,
      "num_tokens": 78918490.0,
      "step": 35355
    },
    {
      "entropy": 5.2810094356536865,
      "epoch": 3.9713034200033697,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003510097435545774,
      "loss": 4.8893,
      "mean_token_accuracy": 0.21920208334922792,
      "num_tokens": 78929790.0,
      "step": 35360
    },
    {
      "entropy": 5.216127014160156,
      "epoch": 3.971864996911327,
      "grad_norm": 0.96875,
      "learning_rate": 0.00035097195395495057,
      "loss": 4.8391,
      "mean_token_accuracy": 0.23486475497484208,
      "num_tokens": 78941600.0,
      "step": 35365
    },
    {
      "entropy": 5.309979343414307,
      "epoch": 3.9724265738192845,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003509341619363474,
      "loss": 5.0074,
      "mean_token_accuracy": 0.224652498960495,
      "num_tokens": 78953764.0,
      "step": 35370
    },
    {
      "entropy": 5.306994104385376,
      "epoch": 3.972988150727242,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00035089636749997095,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2263655349612236,
      "num_tokens": 78964781.0,
      "step": 35375
    },
    {
      "entropy": 5.334341669082642,
      "epoch": 3.9735497276351994,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00035085857064702487,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22049473077058793,
      "num_tokens": 78976325.0,
      "step": 35380
    },
    {
      "entropy": 5.2001227855682375,
      "epoch": 3.974111304543157,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003508207713787123,
      "loss": 4.7641,
      "mean_token_accuracy": 0.23134999722242355,
      "num_tokens": 78987174.0,
      "step": 35385
    },
    {
      "entropy": 5.256938028335571,
      "epoch": 3.9746728814511147,
      "grad_norm": 0.9375,
      "learning_rate": 0.00035078296969623716,
      "loss": 4.8624,
      "mean_token_accuracy": 0.23081450611352922,
      "num_tokens": 78999195.0,
      "step": 35390
    },
    {
      "entropy": 5.14507155418396,
      "epoch": 3.9752344583590724,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00035074516560080273,
      "loss": 4.7123,
      "mean_token_accuracy": 0.2333172708749771,
      "num_tokens": 79010981.0,
      "step": 35395
    },
    {
      "entropy": 5.217225551605225,
      "epoch": 3.97579603526703,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003507073590936129,
      "loss": 4.8649,
      "mean_token_accuracy": 0.22925853133201599,
      "num_tokens": 79022059.0,
      "step": 35400
    },
    {
      "entropy": 5.4146575927734375,
      "epoch": 3.9763576121749873,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035066955017587135,
      "loss": 5.0409,
      "mean_token_accuracy": 0.22105298340320587,
      "num_tokens": 79033261.0,
      "step": 35405
    },
    {
      "entropy": 5.252393054962158,
      "epoch": 3.976919189082945,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000350631738848782,
      "loss": 4.8693,
      "mean_token_accuracy": 0.22528122663497924,
      "num_tokens": 79044518.0,
      "step": 35410
    },
    {
      "entropy": 5.2391585350036625,
      "epoch": 3.9774807659909026,
      "grad_norm": 1.125,
      "learning_rate": 0.0003505939251135488,
      "loss": 4.8715,
      "mean_token_accuracy": 0.22346564531326293,
      "num_tokens": 79055202.0,
      "step": 35415
    },
    {
      "entropy": 5.287415647506714,
      "epoch": 3.97804234289886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003505561089713756,
      "loss": 4.8983,
      "mean_token_accuracy": 0.2219119757413864,
      "num_tokens": 79066448.0,
      "step": 35420
    },
    {
      "entropy": 5.251841831207275,
      "epoch": 3.9786039198068175,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003505182904234667,
      "loss": 4.8726,
      "mean_token_accuracy": 0.2254887714982033,
      "num_tokens": 79077745.0,
      "step": 35425
    },
    {
      "entropy": 5.367795848846436,
      "epoch": 3.979165496714775,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003504804694710261,
      "loss": 4.9585,
      "mean_token_accuracy": 0.21798774003982543,
      "num_tokens": 79088011.0,
      "step": 35430
    },
    {
      "entropy": 5.382599306106568,
      "epoch": 3.9797270736227324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003504426461152581,
      "loss": 4.9779,
      "mean_token_accuracy": 0.2115369349718094,
      "num_tokens": 79098718.0,
      "step": 35435
    },
    {
      "entropy": 5.2000774383544925,
      "epoch": 3.98028865053069,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003504048203573669,
      "loss": 4.837,
      "mean_token_accuracy": 0.22225389182567595,
      "num_tokens": 79108216.0,
      "step": 35440
    },
    {
      "entropy": 5.219032907485962,
      "epoch": 3.9808502274386477,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003503669921985571,
      "loss": 4.8356,
      "mean_token_accuracy": 0.22393537759780885,
      "num_tokens": 79119709.0,
      "step": 35445
    },
    {
      "entropy": 5.3230000019073485,
      "epoch": 3.9814118043466054,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000350329161640033,
      "loss": 4.9564,
      "mean_token_accuracy": 0.22318258732557297,
      "num_tokens": 79131824.0,
      "step": 35450
    },
    {
      "entropy": 5.291489982604981,
      "epoch": 3.981973381254563,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00035029132868299914,
      "loss": 4.8735,
      "mean_token_accuracy": 0.22042271941900254,
      "num_tokens": 79142855.0,
      "step": 35455
    },
    {
      "entropy": 5.324219131469727,
      "epoch": 3.9825349581625202,
      "grad_norm": 1.0625,
      "learning_rate": 0.00035025349332866017,
      "loss": 4.9954,
      "mean_token_accuracy": 0.21772187650203706,
      "num_tokens": 79153759.0,
      "step": 35460
    },
    {
      "entropy": 5.275825881958008,
      "epoch": 3.983096535070478,
      "grad_norm": 1.0,
      "learning_rate": 0.00035021565557822085,
      "loss": 4.8655,
      "mean_token_accuracy": 0.22454322427511214,
      "num_tokens": 79163586.0,
      "step": 35465
    },
    {
      "entropy": 5.317389583587646,
      "epoch": 3.9836581119784356,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003501778154328858,
      "loss": 4.9534,
      "mean_token_accuracy": 0.213994362950325,
      "num_tokens": 79175067.0,
      "step": 35470
    },
    {
      "entropy": 5.163973951339722,
      "epoch": 3.9842196888863928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003501399728938599,
      "loss": 4.7515,
      "mean_token_accuracy": 0.2364716574549675,
      "num_tokens": 79186039.0,
      "step": 35475
    },
    {
      "entropy": 5.2452904224395756,
      "epoch": 3.9847812657943504,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00035010212796234827,
      "loss": 4.9247,
      "mean_token_accuracy": 0.22300320714712143,
      "num_tokens": 79198571.0,
      "step": 35480
    },
    {
      "entropy": 5.170207166671753,
      "epoch": 3.985342842702308,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003500642806395557,
      "loss": 4.7977,
      "mean_token_accuracy": 0.22997900992631912,
      "num_tokens": 79210253.0,
      "step": 35485
    },
    {
      "entropy": 5.232416105270386,
      "epoch": 3.9859044196102658,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003500264309266873,
      "loss": 4.8414,
      "mean_token_accuracy": 0.22842978090047836,
      "num_tokens": 79221611.0,
      "step": 35490
    },
    {
      "entropy": 5.290653085708618,
      "epoch": 3.9864659965182234,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034998857882494823,
      "loss": 4.9566,
      "mean_token_accuracy": 0.2205450251698494,
      "num_tokens": 79233270.0,
      "step": 35495
    },
    {
      "entropy": 5.2692177295684814,
      "epoch": 3.9870275734261806,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034995072433554377,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2157391682267189,
      "num_tokens": 79245338.0,
      "step": 35500
    },
    {
      "entropy": 5.348285341262818,
      "epoch": 3.9875891503341383,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003499128674596792,
      "loss": 4.9341,
      "mean_token_accuracy": 0.22320929020643235,
      "num_tokens": 79255558.0,
      "step": 35505
    },
    {
      "entropy": 5.327332353591919,
      "epoch": 3.988150727242096,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034987500819855976,
      "loss": 4.9378,
      "mean_token_accuracy": 0.2269112393260002,
      "num_tokens": 79266075.0,
      "step": 35510
    },
    {
      "entropy": 5.303242921829224,
      "epoch": 3.988712304150053,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034983714655339115,
      "loss": 4.936,
      "mean_token_accuracy": 0.21971335709095002,
      "num_tokens": 79277667.0,
      "step": 35515
    },
    {
      "entropy": 5.197019147872925,
      "epoch": 3.989273881058011,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003497992825253787,
      "loss": 4.7995,
      "mean_token_accuracy": 0.2294368639588356,
      "num_tokens": 79288755.0,
      "step": 35520
    },
    {
      "entropy": 5.28092737197876,
      "epoch": 3.9898354579659685,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034976141611572817,
      "loss": 4.9124,
      "mean_token_accuracy": 0.2242048665881157,
      "num_tokens": 79299048.0,
      "step": 35525
    },
    {
      "entropy": 5.263582372665406,
      "epoch": 3.9903970348739257,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034972354732564506,
      "loss": 4.8563,
      "mean_token_accuracy": 0.2285652741789818,
      "num_tokens": 79310847.0,
      "step": 35530
    },
    {
      "entropy": 5.306619644165039,
      "epoch": 3.9909586117818834,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034968567615633516,
      "loss": 4.9167,
      "mean_token_accuracy": 0.2178127273917198,
      "num_tokens": 79322496.0,
      "step": 35535
    },
    {
      "entropy": 5.253956937789917,
      "epoch": 3.991520188689841,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034964780260900445,
      "loss": 4.8981,
      "mean_token_accuracy": 0.2219819098711014,
      "num_tokens": 79334216.0,
      "step": 35540
    },
    {
      "entropy": 5.212043046951294,
      "epoch": 3.9920817655977987,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034960992668485875,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23252913355827332,
      "num_tokens": 79344320.0,
      "step": 35545
    },
    {
      "entropy": 5.333400297164917,
      "epoch": 3.9926433425057564,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034957204838510406,
      "loss": 5.0558,
      "mean_token_accuracy": 0.213144713640213,
      "num_tokens": 79355775.0,
      "step": 35550
    },
    {
      "entropy": 5.211504650115967,
      "epoch": 3.9932049194137136,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034953416771094645,
      "loss": 4.7759,
      "mean_token_accuracy": 0.2285192847251892,
      "num_tokens": 79366866.0,
      "step": 35555
    },
    {
      "entropy": 5.257034873962402,
      "epoch": 3.9937664963216712,
      "grad_norm": 1.015625,
      "learning_rate": 0.000349496284663592,
      "loss": 4.8987,
      "mean_token_accuracy": 0.21694509088993072,
      "num_tokens": 79378599.0,
      "step": 35560
    },
    {
      "entropy": 5.208088397979736,
      "epoch": 3.994328073229629,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034945839924424696,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23138913810253142,
      "num_tokens": 79389283.0,
      "step": 35565
    },
    {
      "entropy": 5.27702374458313,
      "epoch": 3.994889650137586,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003494205114541176,
      "loss": 4.923,
      "mean_token_accuracy": 0.215842404961586,
      "num_tokens": 79400858.0,
      "step": 35570
    },
    {
      "entropy": 5.299483442306519,
      "epoch": 3.995451227045544,
      "grad_norm": 1.125,
      "learning_rate": 0.00034938262129441026,
      "loss": 4.94,
      "mean_token_accuracy": 0.22186098694801332,
      "num_tokens": 79411122.0,
      "step": 35575
    },
    {
      "entropy": 5.241509294509887,
      "epoch": 3.9960128039535014,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003493447287663315,
      "loss": 4.7361,
      "mean_token_accuracy": 0.23652749955654145,
      "num_tokens": 79420153.0,
      "step": 35580
    },
    {
      "entropy": 5.355779409408569,
      "epoch": 3.996574380861459,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003493068338710877,
      "loss": 5.0154,
      "mean_token_accuracy": 0.22299708873033525,
      "num_tokens": 79432964.0,
      "step": 35585
    },
    {
      "entropy": 5.207520151138306,
      "epoch": 3.9971359577694168,
      "grad_norm": 0.90234375,
      "learning_rate": 0.0003492689366098855,
      "loss": 4.8088,
      "mean_token_accuracy": 0.23187992721796036,
      "num_tokens": 79445191.0,
      "step": 35590
    },
    {
      "entropy": 5.303087282180786,
      "epoch": 3.997697534677374,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034923103698393147,
      "loss": 4.8419,
      "mean_token_accuracy": 0.23010018318891526,
      "num_tokens": 79455553.0,
      "step": 35595
    },
    {
      "entropy": 5.30351185798645,
      "epoch": 3.9982591115853316,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003491931349944325,
      "loss": 4.9161,
      "mean_token_accuracy": 0.21937607675790788,
      "num_tokens": 79466781.0,
      "step": 35600
    },
    {
      "entropy": 5.207737064361572,
      "epoch": 3.9988206884932893,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034915523064259533,
      "loss": 4.8775,
      "mean_token_accuracy": 0.22995718121528624,
      "num_tokens": 79478656.0,
      "step": 35605
    },
    {
      "entropy": 5.216588354110717,
      "epoch": 3.9993822654012465,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003491173239296268,
      "loss": 4.8559,
      "mean_token_accuracy": 0.2248761311173439,
      "num_tokens": 79489763.0,
      "step": 35610
    },
    {
      "entropy": 5.2901664733886715,
      "epoch": 3.999943842309204,
      "grad_norm": 0.94921875,
      "learning_rate": 0.000349079414856734,
      "loss": 4.9703,
      "mean_token_accuracy": 0.21946855187416076,
      "num_tokens": 79501828.0,
      "step": 35615
    },
    {
      "entropy": 5.282316790686713,
      "epoch": 4.000449261526366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003490415034251239,
      "loss": 4.8199,
      "mean_token_accuracy": 0.22771555682023367,
      "num_tokens": 79510316.0,
      "step": 35620
    },
    {
      "entropy": 5.203341865539551,
      "epoch": 4.001010838434324,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003490035896360036,
      "loss": 4.7594,
      "mean_token_accuracy": 0.23440032154321672,
      "num_tokens": 79521555.0,
      "step": 35625
    },
    {
      "entropy": 5.32935209274292,
      "epoch": 4.001572415342281,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034896567349058027,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22841309756040573,
      "num_tokens": 79533147.0,
      "step": 35630
    },
    {
      "entropy": 5.315447902679443,
      "epoch": 4.002133992250239,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034892775499006126,
      "loss": 4.9,
      "mean_token_accuracy": 0.22554416507482528,
      "num_tokens": 79544695.0,
      "step": 35635
    },
    {
      "entropy": 5.160246753692627,
      "epoch": 4.002695569158196,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003488898341356538,
      "loss": 4.6616,
      "mean_token_accuracy": 0.2436058357357979,
      "num_tokens": 79555740.0,
      "step": 35640
    },
    {
      "entropy": 5.237882423400879,
      "epoch": 4.0032571460661535,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00034885191092856545,
      "loss": 4.7594,
      "mean_token_accuracy": 0.23162752389907837,
      "num_tokens": 79565648.0,
      "step": 35645
    },
    {
      "entropy": 5.285501623153687,
      "epoch": 4.003818722974112,
      "grad_norm": 1.0,
      "learning_rate": 0.0003488139853700036,
      "loss": 4.849,
      "mean_token_accuracy": 0.23397212475538254,
      "num_tokens": 79577107.0,
      "step": 35650
    },
    {
      "entropy": 5.268370771408081,
      "epoch": 4.004380299882069,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034877605746117576,
      "loss": 4.8044,
      "mean_token_accuracy": 0.232949361205101,
      "num_tokens": 79586689.0,
      "step": 35655
    },
    {
      "entropy": 5.300032997131348,
      "epoch": 4.004941876790026,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003487381272032897,
      "loss": 4.8769,
      "mean_token_accuracy": 0.22474633306264877,
      "num_tokens": 79597517.0,
      "step": 35660
    },
    {
      "entropy": 5.2767493724823,
      "epoch": 4.005503453697984,
      "grad_norm": 0.92578125,
      "learning_rate": 0.000348700194597553,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2230292171239853,
      "num_tokens": 79609516.0,
      "step": 35665
    },
    {
      "entropy": 5.217194128036499,
      "epoch": 4.006065030605941,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034866225964517354,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23270778208971024,
      "num_tokens": 79620706.0,
      "step": 35670
    },
    {
      "entropy": 5.204157590866089,
      "epoch": 4.006626607513899,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003486243223473592,
      "loss": 4.7055,
      "mean_token_accuracy": 0.2351480320096016,
      "num_tokens": 79630897.0,
      "step": 35675
    },
    {
      "entropy": 5.225091409683228,
      "epoch": 4.007188184421857,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034858638270531786,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23155344724655152,
      "num_tokens": 79642426.0,
      "step": 35680
    },
    {
      "entropy": 5.287176704406738,
      "epoch": 4.007749761329814,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00034854844072025755,
      "loss": 4.7879,
      "mean_token_accuracy": 0.23225488066673278,
      "num_tokens": 79653491.0,
      "step": 35685
    },
    {
      "entropy": 5.202410078048706,
      "epoch": 4.008311338237772,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034851049639338637,
      "loss": 4.7351,
      "mean_token_accuracy": 0.2326071634888649,
      "num_tokens": 79665329.0,
      "step": 35690
    },
    {
      "entropy": 5.196435499191284,
      "epoch": 4.008872915145729,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003484725497259125,
      "loss": 4.7258,
      "mean_token_accuracy": 0.2288929432630539,
      "num_tokens": 79676552.0,
      "step": 35695
    },
    {
      "entropy": 5.193324422836303,
      "epoch": 4.009434492053686,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00034843460071904407,
      "loss": 4.7882,
      "mean_token_accuracy": 0.23069617748260499,
      "num_tokens": 79687359.0,
      "step": 35700
    },
    {
      "entropy": 5.2149067401885985,
      "epoch": 4.0099960689616445,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034839664937398964,
      "loss": 4.6914,
      "mean_token_accuracy": 0.2383899673819542,
      "num_tokens": 79698728.0,
      "step": 35705
    },
    {
      "entropy": 5.272552347183227,
      "epoch": 4.010557645869602,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003483586956919572,
      "loss": 4.8996,
      "mean_token_accuracy": 0.2214653253555298,
      "num_tokens": 79709018.0,
      "step": 35710
    },
    {
      "entropy": 5.277088212966919,
      "epoch": 4.011119222777559,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003483207396741556,
      "loss": 4.8304,
      "mean_token_accuracy": 0.22708527594804764,
      "num_tokens": 79720166.0,
      "step": 35715
    },
    {
      "entropy": 5.192399024963379,
      "epoch": 4.011680799685517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003482827813217932,
      "loss": 4.6517,
      "mean_token_accuracy": 0.24293474853038788,
      "num_tokens": 79730467.0,
      "step": 35720
    },
    {
      "entropy": 5.241680002212524,
      "epoch": 4.012242376593474,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034824482063607856,
      "loss": 4.8248,
      "mean_token_accuracy": 0.23434983640909196,
      "num_tokens": 79740570.0,
      "step": 35725
    },
    {
      "entropy": 5.238378381729126,
      "epoch": 4.012803953501432,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003482068576182204,
      "loss": 4.8235,
      "mean_token_accuracy": 0.22982082068920134,
      "num_tokens": 79751856.0,
      "step": 35730
    },
    {
      "entropy": 5.273518228530884,
      "epoch": 4.01336553040939,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034816889226942755,
      "loss": 4.7863,
      "mean_token_accuracy": 0.23807796686887742,
      "num_tokens": 79763267.0,
      "step": 35735
    },
    {
      "entropy": 5.31389799118042,
      "epoch": 4.013927107317347,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003481309245909088,
      "loss": 4.911,
      "mean_token_accuracy": 0.22986797243356705,
      "num_tokens": 79775202.0,
      "step": 35740
    },
    {
      "entropy": 5.243584060668946,
      "epoch": 4.014488684225305,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003480929545838731,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22809989154338836,
      "num_tokens": 79787087.0,
      "step": 35745
    },
    {
      "entropy": 5.352618598937989,
      "epoch": 4.015050261133262,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003480549822495293,
      "loss": 4.8898,
      "mean_token_accuracy": 0.22845367193222046,
      "num_tokens": 79799014.0,
      "step": 35750
    },
    {
      "entropy": 5.231399059295654,
      "epoch": 4.015611838041219,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034801700758908653,
      "loss": 4.7413,
      "mean_token_accuracy": 0.2393426775932312,
      "num_tokens": 79810455.0,
      "step": 35755
    },
    {
      "entropy": 5.267898178100586,
      "epoch": 4.0161734149491775,
      "grad_norm": 1.0,
      "learning_rate": 0.00034797903060375396,
      "loss": 4.8031,
      "mean_token_accuracy": 0.23209430277347565,
      "num_tokens": 79821614.0,
      "step": 35760
    },
    {
      "entropy": 5.270129013061523,
      "epoch": 4.016734991857135,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003479410512947407,
      "loss": 4.9304,
      "mean_token_accuracy": 0.22500271946191788,
      "num_tokens": 79832418.0,
      "step": 35765
    },
    {
      "entropy": 5.311985731124878,
      "epoch": 4.017296568765093,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034790306966325606,
      "loss": 4.9062,
      "mean_token_accuracy": 0.22597259134054185,
      "num_tokens": 79844234.0,
      "step": 35770
    },
    {
      "entropy": 5.230355167388916,
      "epoch": 4.01785814567305,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003478650857105095,
      "loss": 4.7444,
      "mean_token_accuracy": 0.2348562464118004,
      "num_tokens": 79855284.0,
      "step": 35775
    },
    {
      "entropy": 5.19921908378601,
      "epoch": 4.018419722581007,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003478270994377102,
      "loss": 4.7325,
      "mean_token_accuracy": 0.2372565895318985,
      "num_tokens": 79866169.0,
      "step": 35780
    },
    {
      "entropy": 5.278383255004883,
      "epoch": 4.018981299488965,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034778911084606793,
      "loss": 4.9106,
      "mean_token_accuracy": 0.21811489909887313,
      "num_tokens": 79878044.0,
      "step": 35785
    },
    {
      "entropy": 5.206542444229126,
      "epoch": 4.0195428763969225,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003477511199367922,
      "loss": 4.745,
      "mean_token_accuracy": 0.233352991938591,
      "num_tokens": 79888859.0,
      "step": 35790
    },
    {
      "entropy": 5.285934257507324,
      "epoch": 4.02010445330488,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034771312671109256,
      "loss": 4.768,
      "mean_token_accuracy": 0.23011804223060608,
      "num_tokens": 79899886.0,
      "step": 35795
    },
    {
      "entropy": 5.295814847946167,
      "epoch": 4.020666030212838,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003476751311701787,
      "loss": 4.8417,
      "mean_token_accuracy": 0.2359776809811592,
      "num_tokens": 79911752.0,
      "step": 35800
    },
    {
      "entropy": 5.236070060729981,
      "epoch": 4.021227607120795,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003476371333152605,
      "loss": 4.775,
      "mean_token_accuracy": 0.22307239174842836,
      "num_tokens": 79923076.0,
      "step": 35805
    },
    {
      "entropy": 5.236436891555786,
      "epoch": 4.021789184028753,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034759913314754776,
      "loss": 4.7683,
      "mean_token_accuracy": 0.2305786445736885,
      "num_tokens": 79933866.0,
      "step": 35810
    },
    {
      "entropy": 5.219929552078247,
      "epoch": 4.02235076093671,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034756113066825045,
      "loss": 4.8132,
      "mean_token_accuracy": 0.2338604062795639,
      "num_tokens": 79946046.0,
      "step": 35815
    },
    {
      "entropy": 5.220564270019532,
      "epoch": 4.022912337844668,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003475231258785787,
      "loss": 4.7639,
      "mean_token_accuracy": 0.2337728962302208,
      "num_tokens": 79957887.0,
      "step": 35820
    },
    {
      "entropy": 5.161539793014526,
      "epoch": 4.023473914752626,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003474851187797424,
      "loss": 4.6878,
      "mean_token_accuracy": 0.23692340403795242,
      "num_tokens": 79968101.0,
      "step": 35825
    },
    {
      "entropy": 5.200352478027344,
      "epoch": 4.024035491660583,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034744710937295185,
      "loss": 4.8053,
      "mean_token_accuracy": 0.23064144253730773,
      "num_tokens": 79978176.0,
      "step": 35830
    },
    {
      "entropy": 5.26711220741272,
      "epoch": 4.02459706856854,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034740909765941725,
      "loss": 4.7903,
      "mean_token_accuracy": 0.22888897955417634,
      "num_tokens": 79990324.0,
      "step": 35835
    },
    {
      "entropy": 5.286613798141479,
      "epoch": 4.025158645476498,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003473710836403489,
      "loss": 4.8119,
      "mean_token_accuracy": 0.2315091833472252,
      "num_tokens": 80001898.0,
      "step": 35840
    },
    {
      "entropy": 5.186161136627197,
      "epoch": 4.0257202223844555,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003473330673169572,
      "loss": 4.8054,
      "mean_token_accuracy": 0.22800993472337722,
      "num_tokens": 80012072.0,
      "step": 35845
    },
    {
      "entropy": 5.170717048645019,
      "epoch": 4.026281799292413,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003472950486904525,
      "loss": 4.7341,
      "mean_token_accuracy": 0.23027085810899733,
      "num_tokens": 80022718.0,
      "step": 35850
    },
    {
      "entropy": 5.304425239562988,
      "epoch": 4.026843376200371,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003472570277620455,
      "loss": 4.8381,
      "mean_token_accuracy": 0.2274532914161682,
      "num_tokens": 80033899.0,
      "step": 35855
    },
    {
      "entropy": 5.303130865097046,
      "epoch": 4.027404953108328,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003472190045329466,
      "loss": 4.8264,
      "mean_token_accuracy": 0.22963470369577407,
      "num_tokens": 80045398.0,
      "step": 35860
    },
    {
      "entropy": 5.3062866687774655,
      "epoch": 4.027966530016286,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003471809790043667,
      "loss": 4.8774,
      "mean_token_accuracy": 0.2265894040465355,
      "num_tokens": 80055186.0,
      "step": 35865
    },
    {
      "entropy": 5.252239322662353,
      "epoch": 4.028528106924243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034714295117751635,
      "loss": 4.841,
      "mean_token_accuracy": 0.23403607457876205,
      "num_tokens": 80065545.0,
      "step": 35870
    },
    {
      "entropy": 5.2780509948730465,
      "epoch": 4.029089683832201,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034710492105360653,
      "loss": 4.8857,
      "mean_token_accuracy": 0.22729650288820266,
      "num_tokens": 80077451.0,
      "step": 35875
    },
    {
      "entropy": 5.2223405838012695,
      "epoch": 4.029651260740159,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003470668886338481,
      "loss": 4.7295,
      "mean_token_accuracy": 0.23736569434404373,
      "num_tokens": 80088215.0,
      "step": 35880
    },
    {
      "entropy": 5.148496007919311,
      "epoch": 4.030212837648116,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034702885391945184,
      "loss": 4.6639,
      "mean_token_accuracy": 0.23701633214950563,
      "num_tokens": 80098393.0,
      "step": 35885
    },
    {
      "entropy": 5.1996372699737545,
      "epoch": 4.030774414556073,
      "grad_norm": 1.0,
      "learning_rate": 0.000346990816911629,
      "loss": 4.7961,
      "mean_token_accuracy": 0.22510915100574494,
      "num_tokens": 80110183.0,
      "step": 35890
    },
    {
      "entropy": 5.187803411483765,
      "epoch": 4.031335991464031,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003469527776115906,
      "loss": 4.7668,
      "mean_token_accuracy": 0.22652364373207093,
      "num_tokens": 80120954.0,
      "step": 35895
    },
    {
      "entropy": 5.207278680801392,
      "epoch": 4.031897568371988,
      "grad_norm": 1.0,
      "learning_rate": 0.00034691473602054793,
      "loss": 4.6947,
      "mean_token_accuracy": 0.23990667015314102,
      "num_tokens": 80131607.0,
      "step": 35900
    },
    {
      "entropy": 5.1893877506256105,
      "epoch": 4.032459145279946,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003468766921397121,
      "loss": 4.7526,
      "mean_token_accuracy": 0.23271261900663376,
      "num_tokens": 80141700.0,
      "step": 35905
    },
    {
      "entropy": 5.234491157531738,
      "epoch": 4.033020722187904,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003468386459702946,
      "loss": 4.7925,
      "mean_token_accuracy": 0.2313240185379982,
      "num_tokens": 80151776.0,
      "step": 35910
    },
    {
      "entropy": 5.2782660007476805,
      "epoch": 4.033582299095861,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034680059751350664,
      "loss": 4.8257,
      "mean_token_accuracy": 0.23026061803102493,
      "num_tokens": 80162838.0,
      "step": 35915
    },
    {
      "entropy": 5.250958871841431,
      "epoch": 4.034143876003819,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00034676254677055984,
      "loss": 4.8373,
      "mean_token_accuracy": 0.2322317346930504,
      "num_tokens": 80174104.0,
      "step": 35920
    },
    {
      "entropy": 5.271726989746094,
      "epoch": 4.034705452911776,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00034672449374266575,
      "loss": 4.8768,
      "mean_token_accuracy": 0.22544723898172378,
      "num_tokens": 80185015.0,
      "step": 35925
    },
    {
      "entropy": 5.281509923934936,
      "epoch": 4.0352670298197335,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003466864384310359,
      "loss": 4.7509,
      "mean_token_accuracy": 0.236931411921978,
      "num_tokens": 80196893.0,
      "step": 35930
    },
    {
      "entropy": 5.3120277404785154,
      "epoch": 4.035828606727692,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034664838083688223,
      "loss": 4.8382,
      "mean_token_accuracy": 0.23405566811561584,
      "num_tokens": 80209910.0,
      "step": 35935
    },
    {
      "entropy": 5.184444189071655,
      "epoch": 4.036390183635649,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034661032096141615,
      "loss": 4.8082,
      "mean_token_accuracy": 0.22899967581033706,
      "num_tokens": 80219780.0,
      "step": 35940
    },
    {
      "entropy": 5.284019851684571,
      "epoch": 4.036951760543606,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003465722588058497,
      "loss": 4.8584,
      "mean_token_accuracy": 0.2226549819111824,
      "num_tokens": 80231316.0,
      "step": 35945
    },
    {
      "entropy": 5.30410475730896,
      "epoch": 4.037513337451564,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034653419437139475,
      "loss": 4.8402,
      "mean_token_accuracy": 0.23149073123931885,
      "num_tokens": 80242622.0,
      "step": 35950
    },
    {
      "entropy": 5.289302682876587,
      "epoch": 4.038074914359521,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003464961276592634,
      "loss": 4.8259,
      "mean_token_accuracy": 0.23873451799154283,
      "num_tokens": 80255357.0,
      "step": 35955
    },
    {
      "entropy": 5.328776550292969,
      "epoch": 4.0386364912674795,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034645805867066756,
      "loss": 4.834,
      "mean_token_accuracy": 0.23079999834299086,
      "num_tokens": 80267430.0,
      "step": 35960
    },
    {
      "entropy": 5.195296573638916,
      "epoch": 4.039198068175437,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034641998740681933,
      "loss": 4.7382,
      "mean_token_accuracy": 0.23498826026916503,
      "num_tokens": 80277820.0,
      "step": 35965
    },
    {
      "entropy": 5.297460699081421,
      "epoch": 4.039759645083394,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034638191386893115,
      "loss": 4.9357,
      "mean_token_accuracy": 0.21885848492383958,
      "num_tokens": 80289559.0,
      "step": 35970
    },
    {
      "entropy": 5.288376903533935,
      "epoch": 4.040321221991352,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034634383805821504,
      "loss": 4.8727,
      "mean_token_accuracy": 0.22656624615192414,
      "num_tokens": 80300452.0,
      "step": 35975
    },
    {
      "entropy": 5.287605619430542,
      "epoch": 4.040882798899309,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034630575997588353,
      "loss": 4.819,
      "mean_token_accuracy": 0.23154717385768891,
      "num_tokens": 80311403.0,
      "step": 35980
    },
    {
      "entropy": 5.2544310092926025,
      "epoch": 4.0414443758072665,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000346267679623149,
      "loss": 4.7443,
      "mean_token_accuracy": 0.2317111760377884,
      "num_tokens": 80323245.0,
      "step": 35985
    },
    {
      "entropy": 5.221246194839478,
      "epoch": 4.042005952715225,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034622959700122383,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22804774791002275,
      "num_tokens": 80333666.0,
      "step": 35990
    },
    {
      "entropy": 5.255380344390869,
      "epoch": 4.042567529623182,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003461915121113206,
      "loss": 4.7394,
      "mean_token_accuracy": 0.23794298022985458,
      "num_tokens": 80345291.0,
      "step": 35995
    },
    {
      "entropy": 5.17693510055542,
      "epoch": 4.04312910653114,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003461534249546521,
      "loss": 4.7267,
      "mean_token_accuracy": 0.23672664165496826,
      "num_tokens": 80355612.0,
      "step": 36000
    },
    {
      "epoch": 4.04312910653114,
      "eval_entropy": 5.044904067339834,
      "eval_loss": 5.028697490692139,
      "eval_mean_token_accuracy": 0.22616141656140307,
      "eval_num_tokens": 80355612.0,
      "eval_runtime": 23.9309,
      "eval_samples_per_second": 1295.941,
      "eval_steps_per_second": 162.008,
      "step": 36000
    },
    {
      "entropy": 5.178870296478271,
      "epoch": 4.043690683439097,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000346115335532431,
      "loss": 4.7608,
      "mean_token_accuracy": 0.23645397126674653,
      "num_tokens": 80365952.0,
      "step": 36005
    },
    {
      "entropy": 5.252665185928345,
      "epoch": 4.044252260347054,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034607724384587006,
      "loss": 4.8286,
      "mean_token_accuracy": 0.22864236831665039,
      "num_tokens": 80376962.0,
      "step": 36010
    },
    {
      "entropy": 5.3610655784606935,
      "epoch": 4.044813837255012,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034603914989618197,
      "loss": 4.9004,
      "mean_token_accuracy": 0.22806253284215927,
      "num_tokens": 80388671.0,
      "step": 36015
    },
    {
      "entropy": 5.292159366607666,
      "epoch": 4.04537541416297,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003460010536845799,
      "loss": 4.8305,
      "mean_token_accuracy": 0.233014814555645,
      "num_tokens": 80399215.0,
      "step": 36020
    },
    {
      "entropy": 5.2313354969024655,
      "epoch": 4.045936991070927,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003459629552122767,
      "loss": 4.8083,
      "mean_token_accuracy": 0.22751581966876983,
      "num_tokens": 80410590.0,
      "step": 36025
    },
    {
      "entropy": 5.160434341430664,
      "epoch": 4.046498567978885,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003459248544804855,
      "loss": 4.7137,
      "mean_token_accuracy": 0.23941366821527482,
      "num_tokens": 80422085.0,
      "step": 36030
    },
    {
      "entropy": 5.171689701080322,
      "epoch": 4.047060144886842,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003458867514904194,
      "loss": 4.7091,
      "mean_token_accuracy": 0.23736728727817535,
      "num_tokens": 80434389.0,
      "step": 36035
    },
    {
      "entropy": 5.2290280818939205,
      "epoch": 4.047621721794799,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003458486462432916,
      "loss": 4.8236,
      "mean_token_accuracy": 0.2327451765537262,
      "num_tokens": 80445073.0,
      "step": 36040
    },
    {
      "entropy": 5.256468152999878,
      "epoch": 4.0481832987027575,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034581053874031554,
      "loss": 4.7461,
      "mean_token_accuracy": 0.23281451165676117,
      "num_tokens": 80455206.0,
      "step": 36045
    },
    {
      "entropy": 5.266667985916138,
      "epoch": 4.048744875610715,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003457724289827044,
      "loss": 4.8055,
      "mean_token_accuracy": 0.24206395000219344,
      "num_tokens": 80467707.0,
      "step": 36050
    },
    {
      "entropy": 5.243747806549072,
      "epoch": 4.049306452518673,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003457343169716716,
      "loss": 4.8047,
      "mean_token_accuracy": 0.23304833471775055,
      "num_tokens": 80478493.0,
      "step": 36055
    },
    {
      "entropy": 5.163288402557373,
      "epoch": 4.04986802942663,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034569620270843087,
      "loss": 4.817,
      "mean_token_accuracy": 0.2334905058145523,
      "num_tokens": 80490057.0,
      "step": 36060
    },
    {
      "entropy": 5.263451814651489,
      "epoch": 4.050429606334587,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034565808619419554,
      "loss": 4.8243,
      "mean_token_accuracy": 0.22469039261341095,
      "num_tokens": 80500590.0,
      "step": 36065
    },
    {
      "entropy": 5.28191385269165,
      "epoch": 4.050991183242545,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003456199674301793,
      "loss": 4.7822,
      "mean_token_accuracy": 0.2317795693874359,
      "num_tokens": 80510836.0,
      "step": 36070
    },
    {
      "entropy": 5.23626675605774,
      "epoch": 4.051552760150503,
      "grad_norm": 0.98828125,
      "learning_rate": 0.000345581846417596,
      "loss": 4.8406,
      "mean_token_accuracy": 0.22672017067670822,
      "num_tokens": 80521770.0,
      "step": 36075
    },
    {
      "entropy": 5.319570589065552,
      "epoch": 4.05211433705846,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034554372315765927,
      "loss": 4.8073,
      "mean_token_accuracy": 0.23655263632535933,
      "num_tokens": 80532331.0,
      "step": 36080
    },
    {
      "entropy": 5.290404891967773,
      "epoch": 4.052675913966418,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003455055976515831,
      "loss": 4.7904,
      "mean_token_accuracy": 0.22857017666101456,
      "num_tokens": 80544858.0,
      "step": 36085
    },
    {
      "entropy": 5.232823705673217,
      "epoch": 4.053237490874375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00034546746990058125,
      "loss": 4.8965,
      "mean_token_accuracy": 0.22409212440252305,
      "num_tokens": 80556493.0,
      "step": 36090
    },
    {
      "entropy": 5.27481780052185,
      "epoch": 4.053799067782332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034542933990586795,
      "loss": 4.848,
      "mean_token_accuracy": 0.23082523196935653,
      "num_tokens": 80567416.0,
      "step": 36095
    },
    {
      "entropy": 5.293190002441406,
      "epoch": 4.0543606446902904,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034539120766865713,
      "loss": 4.825,
      "mean_token_accuracy": 0.22995976060628892,
      "num_tokens": 80578175.0,
      "step": 36100
    },
    {
      "entropy": 5.198750638961792,
      "epoch": 4.054922221598248,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003453530731901628,
      "loss": 4.7177,
      "mean_token_accuracy": 0.24227286577224733,
      "num_tokens": 80588755.0,
      "step": 36105
    },
    {
      "entropy": 5.217647695541382,
      "epoch": 4.055483798506206,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003453149364715995,
      "loss": 4.8435,
      "mean_token_accuracy": 0.2278007373213768,
      "num_tokens": 80599734.0,
      "step": 36110
    },
    {
      "entropy": 5.134049797058106,
      "epoch": 4.056045375414163,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00034527679751418126,
      "loss": 4.7118,
      "mean_token_accuracy": 0.24387237429618835,
      "num_tokens": 80610765.0,
      "step": 36115
    },
    {
      "entropy": 5.270928430557251,
      "epoch": 4.05660695232212,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003452386563191226,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23799055069684982,
      "num_tokens": 80622680.0,
      "step": 36120
    },
    {
      "entropy": 5.27770962715149,
      "epoch": 4.057168529230078,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034520051288763784,
      "loss": 4.8421,
      "mean_token_accuracy": 0.22457377463579178,
      "num_tokens": 80633115.0,
      "step": 36125
    },
    {
      "entropy": 5.287224960327149,
      "epoch": 4.0577301061380355,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003451623672209415,
      "loss": 4.7585,
      "mean_token_accuracy": 0.23127024620771408,
      "num_tokens": 80643750.0,
      "step": 36130
    },
    {
      "entropy": 5.25613374710083,
      "epoch": 4.058291683045993,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034512421932024825,
      "loss": 4.8562,
      "mean_token_accuracy": 0.2238221138715744,
      "num_tokens": 80654392.0,
      "step": 36135
    },
    {
      "entropy": 5.186011409759521,
      "epoch": 4.058853259953951,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003450860691867726,
      "loss": 4.8013,
      "mean_token_accuracy": 0.22900995910167693,
      "num_tokens": 80665509.0,
      "step": 36140
    },
    {
      "entropy": 5.217418146133423,
      "epoch": 4.059414836861908,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034504791682172934,
      "loss": 4.7921,
      "mean_token_accuracy": 0.23031475096940995,
      "num_tokens": 80676895.0,
      "step": 36145
    },
    {
      "entropy": 5.253609943389892,
      "epoch": 4.059976413769866,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003450097622263333,
      "loss": 4.794,
      "mean_token_accuracy": 0.23130337297916412,
      "num_tokens": 80688321.0,
      "step": 36150
    },
    {
      "entropy": 5.196014881134033,
      "epoch": 4.060537990677823,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034497160540179924,
      "loss": 4.6933,
      "mean_token_accuracy": 0.24514185637235641,
      "num_tokens": 80699534.0,
      "step": 36155
    },
    {
      "entropy": 5.132897996902466,
      "epoch": 4.061099567585781,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034493344634934205,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24634879529476167,
      "num_tokens": 80710999.0,
      "step": 36160
    },
    {
      "entropy": 5.14208493232727,
      "epoch": 4.061661144493739,
      "grad_norm": 1.0,
      "learning_rate": 0.0003448952850701769,
      "loss": 4.7778,
      "mean_token_accuracy": 0.23083793073892594,
      "num_tokens": 80722943.0,
      "step": 36165
    },
    {
      "entropy": 5.207236242294312,
      "epoch": 4.062222721401696,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034485712156551873,
      "loss": 4.8344,
      "mean_token_accuracy": 0.2255697578191757,
      "num_tokens": 80733142.0,
      "step": 36170
    },
    {
      "entropy": 5.236158132553101,
      "epoch": 4.062784298309653,
      "grad_norm": 0.9375,
      "learning_rate": 0.00034481895583658267,
      "loss": 4.781,
      "mean_token_accuracy": 0.23044755905866623,
      "num_tokens": 80745228.0,
      "step": 36175
    },
    {
      "entropy": 5.294999027252198,
      "epoch": 4.063345875217611,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034478078788458406,
      "loss": 4.8126,
      "mean_token_accuracy": 0.23642753958702087,
      "num_tokens": 80756866.0,
      "step": 36180
    },
    {
      "entropy": 5.1575878143310545,
      "epoch": 4.0639074521255685,
      "grad_norm": 1.078125,
      "learning_rate": 0.000344742617710738,
      "loss": 4.7089,
      "mean_token_accuracy": 0.24439531266689302,
      "num_tokens": 80767541.0,
      "step": 36185
    },
    {
      "entropy": 5.209853458404541,
      "epoch": 4.064469029033527,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034470444531626003,
      "loss": 4.8011,
      "mean_token_accuracy": 0.234771066904068,
      "num_tokens": 80779577.0,
      "step": 36190
    },
    {
      "entropy": 5.279819536209106,
      "epoch": 4.065030605941484,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034466627070236545,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23598128259181977,
      "num_tokens": 80790569.0,
      "step": 36195
    },
    {
      "entropy": 5.306984996795654,
      "epoch": 4.065592182849441,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003446280938702698,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22526146322488785,
      "num_tokens": 80800887.0,
      "step": 36200
    },
    {
      "entropy": 5.287614631652832,
      "epoch": 4.066153759757399,
      "grad_norm": 1.0,
      "learning_rate": 0.00034458991482118865,
      "loss": 4.7748,
      "mean_token_accuracy": 0.2327641323208809,
      "num_tokens": 80811967.0,
      "step": 36205
    },
    {
      "entropy": 5.184700393676758,
      "epoch": 4.066715336665356,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003445517335563376,
      "loss": 4.7477,
      "mean_token_accuracy": 0.2396802455186844,
      "num_tokens": 80824770.0,
      "step": 36210
    },
    {
      "entropy": 5.177638816833496,
      "epoch": 4.067276913573314,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034451355007693237,
      "loss": 4.8096,
      "mean_token_accuracy": 0.23204611241817474,
      "num_tokens": 80834913.0,
      "step": 36215
    },
    {
      "entropy": 5.235136413574219,
      "epoch": 4.067838490481272,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003444753643841888,
      "loss": 4.7828,
      "mean_token_accuracy": 0.22837306708097457,
      "num_tokens": 80846555.0,
      "step": 36220
    },
    {
      "entropy": 5.357314777374268,
      "epoch": 4.068400067389229,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00034443717647932265,
      "loss": 4.9138,
      "mean_token_accuracy": 0.22603025734424592,
      "num_tokens": 80857774.0,
      "step": 36225
    },
    {
      "entropy": 5.331984519958496,
      "epoch": 4.068961644297186,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003443989863635498,
      "loss": 4.891,
      "mean_token_accuracy": 0.22672188878059388,
      "num_tokens": 80868618.0,
      "step": 36230
    },
    {
      "entropy": 5.248871803283691,
      "epoch": 4.069523221205144,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00034436079403808647,
      "loss": 4.9036,
      "mean_token_accuracy": 0.22160889953374863,
      "num_tokens": 80880319.0,
      "step": 36235
    },
    {
      "entropy": 5.256568098068238,
      "epoch": 4.070084798113101,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034432259950414846,
      "loss": 4.85,
      "mean_token_accuracy": 0.227320072054863,
      "num_tokens": 80891194.0,
      "step": 36240
    },
    {
      "entropy": 5.21748685836792,
      "epoch": 4.0706463750210595,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000344284402762952,
      "loss": 4.6826,
      "mean_token_accuracy": 0.23753790855407714,
      "num_tokens": 80902322.0,
      "step": 36245
    },
    {
      "entropy": 5.271781587600708,
      "epoch": 4.071207951929017,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003442462038157132,
      "loss": 4.8605,
      "mean_token_accuracy": 0.22802126109600068,
      "num_tokens": 80913324.0,
      "step": 36250
    },
    {
      "entropy": 5.122277355194091,
      "epoch": 4.071769528836974,
      "grad_norm": 1.0,
      "learning_rate": 0.0003442080026636485,
      "loss": 4.7421,
      "mean_token_accuracy": 0.2375991702079773,
      "num_tokens": 80923360.0,
      "step": 36255
    },
    {
      "entropy": 5.253201627731324,
      "epoch": 4.072331105744932,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003441697993079742,
      "loss": 4.8631,
      "mean_token_accuracy": 0.22454130947589873,
      "num_tokens": 80935359.0,
      "step": 36260
    },
    {
      "entropy": 5.187283849716186,
      "epoch": 4.072892682652889,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003441315937499066,
      "loss": 4.7418,
      "mean_token_accuracy": 0.23373289704322814,
      "num_tokens": 80945885.0,
      "step": 36265
    },
    {
      "entropy": 5.297006511688233,
      "epoch": 4.0734542595608465,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00034409338599066224,
      "loss": 4.8118,
      "mean_token_accuracy": 0.23304886668920516,
      "num_tokens": 80955700.0,
      "step": 36270
    },
    {
      "entropy": 5.202056360244751,
      "epoch": 4.074015836468805,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003440551760314577,
      "loss": 4.7778,
      "mean_token_accuracy": 0.24013102650642396,
      "num_tokens": 80967130.0,
      "step": 36275
    },
    {
      "entropy": 5.2495228290557865,
      "epoch": 4.074577413376762,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034401696387350954,
      "loss": 4.8286,
      "mean_token_accuracy": 0.2263338789343834,
      "num_tokens": 80979677.0,
      "step": 36280
    },
    {
      "entropy": 5.262002992630005,
      "epoch": 4.075138990284719,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003439787495180345,
      "loss": 4.8168,
      "mean_token_accuracy": 0.227815705537796,
      "num_tokens": 80990563.0,
      "step": 36285
    },
    {
      "entropy": 5.215271520614624,
      "epoch": 4.075700567192677,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00034394053296624936,
      "loss": 4.7905,
      "mean_token_accuracy": 0.2395736202597618,
      "num_tokens": 81001731.0,
      "step": 36290
    },
    {
      "entropy": 5.2944470882415775,
      "epoch": 4.076262144100634,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003439023142193709,
      "loss": 4.8794,
      "mean_token_accuracy": 0.23111642450094222,
      "num_tokens": 81013072.0,
      "step": 36295
    },
    {
      "entropy": 5.219962787628174,
      "epoch": 4.0768237210085925,
      "grad_norm": 0.953125,
      "learning_rate": 0.000343864093278616,
      "loss": 4.8038,
      "mean_token_accuracy": 0.23661992996931075,
      "num_tokens": 81023628.0,
      "step": 36300
    },
    {
      "entropy": 5.168609714508056,
      "epoch": 4.07738529791655,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003438258701452017,
      "loss": 4.7008,
      "mean_token_accuracy": 0.23961131721735002,
      "num_tokens": 81034856.0,
      "step": 36305
    },
    {
      "entropy": 5.27223801612854,
      "epoch": 4.077946874824507,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003437876448203449,
      "loss": 4.838,
      "mean_token_accuracy": 0.2345771610736847,
      "num_tokens": 81046998.0,
      "step": 36310
    },
    {
      "entropy": 5.215396738052368,
      "epoch": 4.078508451732465,
      "grad_norm": 1.0,
      "learning_rate": 0.0003437494173052629,
      "loss": 4.7982,
      "mean_token_accuracy": 0.2329712375998497,
      "num_tokens": 81058292.0,
      "step": 36315
    },
    {
      "entropy": 5.245857572555542,
      "epoch": 4.079070028640422,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003437111876011728,
      "loss": 4.7617,
      "mean_token_accuracy": 0.23823844641447067,
      "num_tokens": 81069907.0,
      "step": 36320
    },
    {
      "entropy": 5.236102771759033,
      "epoch": 4.0796316055483794,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034367295570929176,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22302936613559723,
      "num_tokens": 81080700.0,
      "step": 36325
    },
    {
      "entropy": 5.220110988616943,
      "epoch": 4.0801931824563376,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003436347216308373,
      "loss": 4.7912,
      "mean_token_accuracy": 0.2318558305501938,
      "num_tokens": 81092484.0,
      "step": 36330
    },
    {
      "entropy": 5.237440299987793,
      "epoch": 4.080754759364295,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003435964853670265,
      "loss": 4.793,
      "mean_token_accuracy": 0.2340324804186821,
      "num_tokens": 81103992.0,
      "step": 36335
    },
    {
      "entropy": 5.259654712677002,
      "epoch": 4.081316336272253,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034355824691907714,
      "loss": 4.8306,
      "mean_token_accuracy": 0.22583164274692535,
      "num_tokens": 81114995.0,
      "step": 36340
    },
    {
      "entropy": 5.138813161849976,
      "epoch": 4.08187791318021,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00034352000628820655,
      "loss": 4.6577,
      "mean_token_accuracy": 0.2418954238295555,
      "num_tokens": 81125627.0,
      "step": 36345
    },
    {
      "entropy": 5.249180793762207,
      "epoch": 4.082439490088167,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003434817634756324,
      "loss": 4.913,
      "mean_token_accuracy": 0.22356138825416566,
      "num_tokens": 81136957.0,
      "step": 36350
    },
    {
      "entropy": 5.234932136535645,
      "epoch": 4.083001066996125,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003434435184825724,
      "loss": 4.7997,
      "mean_token_accuracy": 0.2351285085082054,
      "num_tokens": 81148840.0,
      "step": 36355
    },
    {
      "entropy": 5.308856868743897,
      "epoch": 4.083562643904083,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00034340527131024415,
      "loss": 4.8782,
      "mean_token_accuracy": 0.2282453805208206,
      "num_tokens": 81160385.0,
      "step": 36360
    },
    {
      "entropy": 5.2124895572662355,
      "epoch": 4.08412422081204,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003433670219598656,
      "loss": 4.7763,
      "mean_token_accuracy": 0.2313439130783081,
      "num_tokens": 81171334.0,
      "step": 36365
    },
    {
      "entropy": 5.221629953384399,
      "epoch": 4.084685797719998,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003433287704326545,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2357679307460785,
      "num_tokens": 81182547.0,
      "step": 36370
    },
    {
      "entropy": 5.1978068351745605,
      "epoch": 4.085247374627955,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003432905167298288,
      "loss": 4.7462,
      "mean_token_accuracy": 0.23437293022871017,
      "num_tokens": 81193427.0,
      "step": 36375
    },
    {
      "entropy": 5.176890850067139,
      "epoch": 4.085808951535913,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034325226085260666,
      "loss": 4.7206,
      "mean_token_accuracy": 0.2397024393081665,
      "num_tokens": 81204522.0,
      "step": 36380
    },
    {
      "entropy": 5.284089756011963,
      "epoch": 4.0863705284438705,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000343214002802206,
      "loss": 4.8983,
      "mean_token_accuracy": 0.22694264650344848,
      "num_tokens": 81215209.0,
      "step": 36385
    },
    {
      "entropy": 5.258699226379394,
      "epoch": 4.086932105351828,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00034317574257984514,
      "loss": 4.8694,
      "mean_token_accuracy": 0.22594716697931289,
      "num_tokens": 81226083.0,
      "step": 36390
    },
    {
      "entropy": 5.177792310714722,
      "epoch": 4.087493682259786,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000343137480186742,
      "loss": 4.7387,
      "mean_token_accuracy": 0.2345718652009964,
      "num_tokens": 81236531.0,
      "step": 36395
    },
    {
      "entropy": 5.223849821090698,
      "epoch": 4.088055259167743,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00034309921562411527,
      "loss": 4.8151,
      "mean_token_accuracy": 0.22624652981758117,
      "num_tokens": 81246856.0,
      "step": 36400
    },
    {
      "entropy": 5.286435794830322,
      "epoch": 4.0886168360757,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00034306094889318295,
      "loss": 4.867,
      "mean_token_accuracy": 0.2242755487561226,
      "num_tokens": 81259352.0,
      "step": 36405
    },
    {
      "entropy": 5.308837795257569,
      "epoch": 4.089178412983658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003430226799951637,
      "loss": 4.8834,
      "mean_token_accuracy": 0.22611220926046371,
      "num_tokens": 81270020.0,
      "step": 36410
    },
    {
      "entropy": 5.233649873733521,
      "epoch": 4.089739989891616,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034298440893127584,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22971402257680892,
      "num_tokens": 81281625.0,
      "step": 36415
    },
    {
      "entropy": 5.1465682029724125,
      "epoch": 4.090301566799573,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034294613570273817,
      "loss": 4.6647,
      "mean_token_accuracy": 0.2437828376889229,
      "num_tokens": 81292340.0,
      "step": 36420
    },
    {
      "entropy": 5.207716798782348,
      "epoch": 4.090863143707531,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034290786031076917,
      "loss": 4.763,
      "mean_token_accuracy": 0.23086587935686112,
      "num_tokens": 81304494.0,
      "step": 36425
    },
    {
      "entropy": 5.223622894287109,
      "epoch": 4.091424720615488,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003428695827565875,
      "loss": 4.7935,
      "mean_token_accuracy": 0.23983915150165558,
      "num_tokens": 81315365.0,
      "step": 36430
    },
    {
      "entropy": 5.291869211196899,
      "epoch": 4.091986297523446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034283130304141204,
      "loss": 4.8125,
      "mean_token_accuracy": 0.22953922897577286,
      "num_tokens": 81326970.0,
      "step": 36435
    },
    {
      "entropy": 5.229167318344116,
      "epoch": 4.092547874431403,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003427930211664615,
      "loss": 4.7614,
      "mean_token_accuracy": 0.2373092532157898,
      "num_tokens": 81337117.0,
      "step": 36440
    },
    {
      "entropy": 5.248345279693604,
      "epoch": 4.093109451339361,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000342754737132955,
      "loss": 4.8457,
      "mean_token_accuracy": 0.23066572099924088,
      "num_tokens": 81347984.0,
      "step": 36445
    },
    {
      "entropy": 5.2780626773834225,
      "epoch": 4.093671028247319,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034271645094211125,
      "loss": 4.8183,
      "mean_token_accuracy": 0.2397150233387947,
      "num_tokens": 81358685.0,
      "step": 36450
    },
    {
      "entropy": 5.234919452667237,
      "epoch": 4.094232605155276,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003426781625951495,
      "loss": 4.7498,
      "mean_token_accuracy": 0.22942193895578383,
      "num_tokens": 81368724.0,
      "step": 36455
    },
    {
      "entropy": 5.310825538635254,
      "epoch": 4.094794182063233,
      "grad_norm": 1.109375,
      "learning_rate": 0.00034263987209328886,
      "loss": 4.9172,
      "mean_token_accuracy": 0.22894667834043503,
      "num_tokens": 81380570.0,
      "step": 36460
    },
    {
      "entropy": 5.242436695098877,
      "epoch": 4.095355758971191,
      "grad_norm": 1.125,
      "learning_rate": 0.0003426015794377483,
      "loss": 4.8308,
      "mean_token_accuracy": 0.23072360008955,
      "num_tokens": 81391606.0,
      "step": 36465
    },
    {
      "entropy": 5.300761365890503,
      "epoch": 4.0959173358791485,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003425632846297474,
      "loss": 4.8474,
      "mean_token_accuracy": 0.2272714078426361,
      "num_tokens": 81402439.0,
      "step": 36470
    },
    {
      "entropy": 5.23991379737854,
      "epoch": 4.096478912787107,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034252498767050525,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23117845207452775,
      "num_tokens": 81413500.0,
      "step": 36475
    },
    {
      "entropy": 5.174816656112671,
      "epoch": 4.097040489695064,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003424866885612413,
      "loss": 4.7227,
      "mean_token_accuracy": 0.23168815821409225,
      "num_tokens": 81424656.0,
      "step": 36480
    },
    {
      "entropy": 5.3291933059692385,
      "epoch": 4.097602066603021,
      "grad_norm": 1.078125,
      "learning_rate": 0.000342448387303175,
      "loss": 4.8917,
      "mean_token_accuracy": 0.22649042159318925,
      "num_tokens": 81436597.0,
      "step": 36485
    },
    {
      "entropy": 5.304773569107056,
      "epoch": 4.098163643510979,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034241008389752594,
      "loss": 4.8545,
      "mean_token_accuracy": 0.22635784596204758,
      "num_tokens": 81448955.0,
      "step": 36490
    },
    {
      "entropy": 5.233062028884888,
      "epoch": 4.098725220418936,
      "grad_norm": 1.6796875,
      "learning_rate": 0.0003423717783455136,
      "loss": 4.7952,
      "mean_token_accuracy": 0.23695776760578155,
      "num_tokens": 81460445.0,
      "step": 36495
    },
    {
      "entropy": 5.247979211807251,
      "epoch": 4.099286797326894,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00034233347064835775,
      "loss": 4.7868,
      "mean_token_accuracy": 0.22856681644916535,
      "num_tokens": 81472103.0,
      "step": 36500
    },
    {
      "entropy": 5.21203293800354,
      "epoch": 4.099848374234852,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003422951608072781,
      "loss": 4.7515,
      "mean_token_accuracy": 0.23205338418483734,
      "num_tokens": 81483051.0,
      "step": 36505
    },
    {
      "entropy": 5.188779258728028,
      "epoch": 4.100409951142809,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003422568488234944,
      "loss": 4.7312,
      "mean_token_accuracy": 0.2350623607635498,
      "num_tokens": 81493755.0,
      "step": 36510
    },
    {
      "entropy": 5.151408338546753,
      "epoch": 4.100971528050766,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034221853469822654,
      "loss": 4.7078,
      "mean_token_accuracy": 0.2331218719482422,
      "num_tokens": 81505738.0,
      "step": 36515
    },
    {
      "entropy": 5.26241774559021,
      "epoch": 4.101533104958724,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034218021843269444,
      "loss": 4.828,
      "mean_token_accuracy": 0.22761632949113847,
      "num_tokens": 81516410.0,
      "step": 36520
    },
    {
      "entropy": 5.254907178878784,
      "epoch": 4.1020946818666815,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003421419000281182,
      "loss": 4.839,
      "mean_token_accuracy": 0.22683467268943786,
      "num_tokens": 81527516.0,
      "step": 36525
    },
    {
      "entropy": 5.287481212615967,
      "epoch": 4.10265625877464,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003421035794857178,
      "loss": 4.8676,
      "mean_token_accuracy": 0.22504886984825134,
      "num_tokens": 81538226.0,
      "step": 36530
    },
    {
      "entropy": 5.261243915557861,
      "epoch": 4.103217835682597,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003420652568067134,
      "loss": 4.8342,
      "mean_token_accuracy": 0.2260866269469261,
      "num_tokens": 81549699.0,
      "step": 36535
    },
    {
      "entropy": 5.288008689880371,
      "epoch": 4.103779412590554,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034202693199232516,
      "loss": 4.9372,
      "mean_token_accuracy": 0.22004078179597855,
      "num_tokens": 81560404.0,
      "step": 36540
    },
    {
      "entropy": 5.276943874359131,
      "epoch": 4.104340989498512,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034198860504377343,
      "loss": 4.8594,
      "mean_token_accuracy": 0.22820966839790344,
      "num_tokens": 81571758.0,
      "step": 36545
    },
    {
      "entropy": 5.228875827789307,
      "epoch": 4.104902566406469,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034195027596227857,
      "loss": 4.7128,
      "mean_token_accuracy": 0.23115812242031097,
      "num_tokens": 81583555.0,
      "step": 36550
    },
    {
      "entropy": 5.163924741744995,
      "epoch": 4.1054641433144266,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003419119447490609,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2375527322292328,
      "num_tokens": 81594901.0,
      "step": 36555
    },
    {
      "entropy": 5.233249616622925,
      "epoch": 4.106025720222385,
      "grad_norm": 0.94140625,
      "learning_rate": 0.000341873611405341,
      "loss": 4.7717,
      "mean_token_accuracy": 0.24062797278165818,
      "num_tokens": 81606229.0,
      "step": 36560
    },
    {
      "entropy": 5.361802291870117,
      "epoch": 4.106587297130342,
      "grad_norm": 1.28125,
      "learning_rate": 0.0003418352759323394,
      "loss": 4.8962,
      "mean_token_accuracy": 0.23080866038799286,
      "num_tokens": 81616543.0,
      "step": 36565
    },
    {
      "entropy": 5.201629018783569,
      "epoch": 4.1071488740383,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00034179693833127667,
      "loss": 4.8277,
      "mean_token_accuracy": 0.2374178111553192,
      "num_tokens": 81627517.0,
      "step": 36570
    },
    {
      "entropy": 5.2097001552581785,
      "epoch": 4.107710450946257,
      "grad_norm": 1.09375,
      "learning_rate": 0.00034175859860337347,
      "loss": 4.7972,
      "mean_token_accuracy": 0.22980870604515075,
      "num_tokens": 81637235.0,
      "step": 36575
    },
    {
      "entropy": 5.2311121940612795,
      "epoch": 4.108272027854214,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034172025674985076,
      "loss": 4.7554,
      "mean_token_accuracy": 0.23074508607387542,
      "num_tokens": 81648539.0,
      "step": 36580
    },
    {
      "entropy": 5.213091802597046,
      "epoch": 4.1088336047621725,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034168191277192904,
      "loss": 4.7367,
      "mean_token_accuracy": 0.2313696026802063,
      "num_tokens": 81659981.0,
      "step": 36585
    },
    {
      "entropy": 5.222721433639526,
      "epoch": 4.10939518167013,
      "grad_norm": 1.015625,
      "learning_rate": 0.00034164356667082933,
      "loss": 4.7564,
      "mean_token_accuracy": 0.2348942056298256,
      "num_tokens": 81669602.0,
      "step": 36590
    },
    {
      "entropy": 5.19576587677002,
      "epoch": 4.109956758578087,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003416052184477727,
      "loss": 4.8309,
      "mean_token_accuracy": 0.23022445887327195,
      "num_tokens": 81679995.0,
      "step": 36595
    },
    {
      "entropy": 5.207393074035645,
      "epoch": 4.110518335486045,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003415668681039801,
      "loss": 4.7684,
      "mean_token_accuracy": 0.23129979521036148,
      "num_tokens": 81691295.0,
      "step": 36600
    },
    {
      "entropy": 5.277207612991333,
      "epoch": 4.111079912394002,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00034152851564067254,
      "loss": 4.7957,
      "mean_token_accuracy": 0.2329239636659622,
      "num_tokens": 81700966.0,
      "step": 36605
    },
    {
      "entropy": 5.313080787658691,
      "epoch": 4.1116414893019595,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003414901610590712,
      "loss": 4.9038,
      "mean_token_accuracy": 0.22361451685428618,
      "num_tokens": 81713469.0,
      "step": 36610
    },
    {
      "entropy": 5.233200311660767,
      "epoch": 4.112203066209918,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003414518043603974,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23150193691253662,
      "num_tokens": 81724748.0,
      "step": 36615
    },
    {
      "entropy": 5.281564378738404,
      "epoch": 4.112764643117875,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034141344554587243,
      "loss": 4.9089,
      "mean_token_accuracy": 0.22517352998256684,
      "num_tokens": 81735694.0,
      "step": 36620
    },
    {
      "entropy": 5.324088764190674,
      "epoch": 4.113326220025833,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003413750846167175,
      "loss": 4.8136,
      "mean_token_accuracy": 0.23032353520393373,
      "num_tokens": 81748251.0,
      "step": 36625
    },
    {
      "entropy": 5.1918658256530765,
      "epoch": 4.11388779693379,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003413367215741542,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23680594563484192,
      "num_tokens": 81760750.0,
      "step": 36630
    },
    {
      "entropy": 5.260195398330689,
      "epoch": 4.114449373841747,
      "grad_norm": 0.984375,
      "learning_rate": 0.00034129835641940394,
      "loss": 4.9023,
      "mean_token_accuracy": 0.22830342054367064,
      "num_tokens": 81772100.0,
      "step": 36635
    },
    {
      "entropy": 5.2614563465118405,
      "epoch": 4.1150109507497055,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00034125998915368827,
      "loss": 4.809,
      "mean_token_accuracy": 0.23020980805158614,
      "num_tokens": 81783795.0,
      "step": 36640
    },
    {
      "entropy": 5.216738748550415,
      "epoch": 4.115572527657663,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00034122161977822885,
      "loss": 4.8194,
      "mean_token_accuracy": 0.2289694458246231,
      "num_tokens": 81794713.0,
      "step": 36645
    },
    {
      "entropy": 5.254062509536743,
      "epoch": 4.11613410456562,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003411832482942474,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2270542025566101,
      "num_tokens": 81806493.0,
      "step": 36650
    },
    {
      "entropy": 5.219977283477784,
      "epoch": 4.116695681473578,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00034114487470296575,
      "loss": 4.6801,
      "mean_token_accuracy": 0.23908528983592986,
      "num_tokens": 81818232.0,
      "step": 36655
    },
    {
      "entropy": 5.166375827789307,
      "epoch": 4.117257258381535,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034110649900560554,
      "loss": 4.7608,
      "mean_token_accuracy": 0.23773505985736848,
      "num_tokens": 81829439.0,
      "step": 36660
    },
    {
      "entropy": 5.268791627883911,
      "epoch": 4.117818835289493,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003410681212033887,
      "loss": 4.7928,
      "mean_token_accuracy": 0.23284949213266373,
      "num_tokens": 81839585.0,
      "step": 36665
    },
    {
      "entropy": 5.256207990646362,
      "epoch": 4.1183804121974505,
      "grad_norm": 0.953125,
      "learning_rate": 0.00034102974129753726,
      "loss": 4.8637,
      "mean_token_accuracy": 0.23091269433498382,
      "num_tokens": 81850648.0,
      "step": 36670
    },
    {
      "entropy": 5.226524734497071,
      "epoch": 4.118941989105408,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003409913592892733,
      "loss": 4.7948,
      "mean_token_accuracy": 0.22771715521812438,
      "num_tokens": 81862597.0,
      "step": 36675
    },
    {
      "entropy": 5.283872365951538,
      "epoch": 4.119503566013366,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003409529751798189,
      "loss": 4.8352,
      "mean_token_accuracy": 0.2337930232286453,
      "num_tokens": 81874540.0,
      "step": 36680
    },
    {
      "entropy": 5.184275484085083,
      "epoch": 4.120065142921323,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003409145889703961,
      "loss": 4.7375,
      "mean_token_accuracy": 0.23735374361276626,
      "num_tokens": 81884205.0,
      "step": 36685
    },
    {
      "entropy": 5.143006992340088,
      "epoch": 4.12062671982928,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00034087620066222733,
      "loss": 4.7135,
      "mean_token_accuracy": 0.23453155010938645,
      "num_tokens": 81895552.0,
      "step": 36690
    },
    {
      "entropy": 5.310031318664551,
      "epoch": 4.121188296737238,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003408378102565347,
      "loss": 4.9338,
      "mean_token_accuracy": 0.21632379293441772,
      "num_tokens": 81908169.0,
      "step": 36695
    },
    {
      "entropy": 5.259902763366699,
      "epoch": 4.121749873645196,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003407994177545407,
      "loss": 4.814,
      "mean_token_accuracy": 0.2317250281572342,
      "num_tokens": 81919465.0,
      "step": 36700
    },
    {
      "entropy": 5.245536994934082,
      "epoch": 4.122311450553153,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003407610231574677,
      "loss": 4.8822,
      "mean_token_accuracy": 0.22537608295679093,
      "num_tokens": 81930558.0,
      "step": 36705
    },
    {
      "entropy": 5.269503879547119,
      "epoch": 4.122873027461111,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003407226264665382,
      "loss": 4.816,
      "mean_token_accuracy": 0.2330642119050026,
      "num_tokens": 81941325.0,
      "step": 36710
    },
    {
      "entropy": 5.2418413162231445,
      "epoch": 4.123434604369068,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00034068422768297486,
      "loss": 4.7494,
      "mean_token_accuracy": 0.2288086786866188,
      "num_tokens": 81953225.0,
      "step": 36715
    },
    {
      "entropy": 5.268327856063843,
      "epoch": 4.123996181277026,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003406458268080001,
      "loss": 4.8552,
      "mean_token_accuracy": 0.2276195764541626,
      "num_tokens": 81964365.0,
      "step": 36720
    },
    {
      "entropy": 5.158691024780273,
      "epoch": 4.1245577581849835,
      "grad_norm": 1.046875,
      "learning_rate": 0.00034060742384283695,
      "loss": 4.816,
      "mean_token_accuracy": 0.22673982232809067,
      "num_tokens": 81974648.0,
      "step": 36725
    },
    {
      "entropy": 5.199850082397461,
      "epoch": 4.125119335092941,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003405690187887078,
      "loss": 4.7421,
      "mean_token_accuracy": 0.24164658635854722,
      "num_tokens": 81985502.0,
      "step": 36730
    },
    {
      "entropy": 5.3522380828857425,
      "epoch": 4.125680912000899,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034053061164683574,
      "loss": 4.8582,
      "mean_token_accuracy": 0.22537535130977632,
      "num_tokens": 81998279.0,
      "step": 36735
    },
    {
      "entropy": 5.188260269165039,
      "epoch": 4.126242488908856,
      "grad_norm": 0.96875,
      "learning_rate": 0.00034049220241844357,
      "loss": 4.7699,
      "mean_token_accuracy": 0.23267971277236937,
      "num_tokens": 82009911.0,
      "step": 36740
    },
    {
      "entropy": 5.122441244125366,
      "epoch": 4.126804065816813,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003404537911047543,
      "loss": 4.7045,
      "mean_token_accuracy": 0.24051164835691452,
      "num_tokens": 82020219.0,
      "step": 36745
    },
    {
      "entropy": 5.184546756744385,
      "epoch": 4.127365642724771,
      "grad_norm": 1.03125,
      "learning_rate": 0.00034041537770699095,
      "loss": 4.6553,
      "mean_token_accuracy": 0.2396721586585045,
      "num_tokens": 82030997.0,
      "step": 36750
    },
    {
      "entropy": 5.256243658065796,
      "epoch": 4.127927219632729,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003403769622263765,
      "loss": 4.8126,
      "mean_token_accuracy": 0.23231249451637268,
      "num_tokens": 82043384.0,
      "step": 36755
    },
    {
      "entropy": 5.189081430435181,
      "epoch": 4.128488796540687,
      "grad_norm": 1.0625,
      "learning_rate": 0.00034033854466413425,
      "loss": 4.73,
      "mean_token_accuracy": 0.23643741458654405,
      "num_tokens": 82053346.0,
      "step": 36760
    },
    {
      "entropy": 5.178538656234741,
      "epoch": 4.129050373448644,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003403001250214875,
      "loss": 4.7321,
      "mean_token_accuracy": 0.23333856761455535,
      "num_tokens": 82063628.0,
      "step": 36765
    },
    {
      "entropy": 5.217950010299683,
      "epoch": 4.129611950356601,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003402617032996593,
      "loss": 4.7717,
      "mean_token_accuracy": 0.24202469736337662,
      "num_tokens": 82074097.0,
      "step": 36770
    },
    {
      "entropy": 5.2229907512664795,
      "epoch": 4.130173527264559,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003402232794998733,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22149972915649413,
      "num_tokens": 82085159.0,
      "step": 36775
    },
    {
      "entropy": 5.242319488525391,
      "epoch": 4.130735104172516,
      "grad_norm": 1.078125,
      "learning_rate": 0.00034018485362335266,
      "loss": 4.831,
      "mean_token_accuracy": 0.2317983016371727,
      "num_tokens": 82096299.0,
      "step": 36780
    },
    {
      "entropy": 5.204729127883911,
      "epoch": 4.131296681080474,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003401464256713211,
      "loss": 4.7504,
      "mean_token_accuracy": 0.23697038143873214,
      "num_tokens": 82107111.0,
      "step": 36785
    },
    {
      "entropy": 5.251073408126831,
      "epoch": 4.131858257988432,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00034010799564500197,
      "loss": 4.8018,
      "mean_token_accuracy": 0.23352738320827485,
      "num_tokens": 82117889.0,
      "step": 36790
    },
    {
      "entropy": 5.2034893989562985,
      "epoch": 4.132419834896389,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003400695635456191,
      "loss": 4.7508,
      "mean_token_accuracy": 0.23703891187906265,
      "num_tokens": 82128380.0,
      "step": 36795
    },
    {
      "entropy": 5.1676068782806395,
      "epoch": 4.132981411804346,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003400311293743961,
      "loss": 4.6891,
      "mean_token_accuracy": 0.24008104503154754,
      "num_tokens": 82138368.0,
      "step": 36800
    },
    {
      "entropy": 5.247323513031006,
      "epoch": 4.133542988712304,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003399926931325567,
      "loss": 4.905,
      "mean_token_accuracy": 0.22820907086133957,
      "num_tokens": 82148449.0,
      "step": 36805
    },
    {
      "entropy": 5.114484262466431,
      "epoch": 4.1341045656202615,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033995425482132464,
      "loss": 4.694,
      "mean_token_accuracy": 0.23821786195039749,
      "num_tokens": 82160937.0,
      "step": 36810
    },
    {
      "entropy": 5.187360858917236,
      "epoch": 4.13466614252822,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033991581444192405,
      "loss": 4.7048,
      "mean_token_accuracy": 0.23776309490203856,
      "num_tokens": 82171546.0,
      "step": 36815
    },
    {
      "entropy": 5.231944370269775,
      "epoch": 4.135227719436177,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003398773719955787,
      "loss": 4.7732,
      "mean_token_accuracy": 0.24037902653217316,
      "num_tokens": 82182613.0,
      "step": 36820
    },
    {
      "entropy": 5.122110652923584,
      "epoch": 4.135789296344134,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003398389274835126,
      "loss": 4.7121,
      "mean_token_accuracy": 0.2397312581539154,
      "num_tokens": 82193757.0,
      "step": 36825
    },
    {
      "entropy": 5.192779636383056,
      "epoch": 4.136350873252092,
      "grad_norm": 1.0,
      "learning_rate": 0.00033980048090694997,
      "loss": 4.8074,
      "mean_token_accuracy": 0.22571048736572266,
      "num_tokens": 82205601.0,
      "step": 36830
    },
    {
      "entropy": 5.248617696762085,
      "epoch": 4.136912450160049,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003397620322671149,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23272761851549148,
      "num_tokens": 82217022.0,
      "step": 36835
    },
    {
      "entropy": 5.190395784378052,
      "epoch": 4.137474027068007,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033972358156523163,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2354823112487793,
      "num_tokens": 82227682.0,
      "step": 36840
    },
    {
      "entropy": 5.174015522003174,
      "epoch": 4.138035603975965,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003396851288025243,
      "loss": 4.7273,
      "mean_token_accuracy": 0.2323189228773117,
      "num_tokens": 82238754.0,
      "step": 36845
    },
    {
      "entropy": 5.325496482849121,
      "epoch": 4.138597180883922,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003396466739802175,
      "loss": 4.8805,
      "mean_token_accuracy": 0.22450461536645888,
      "num_tokens": 82249536.0,
      "step": 36850
    },
    {
      "entropy": 5.252026796340942,
      "epoch": 4.13915875779188,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003396082170995355,
      "loss": 4.7646,
      "mean_token_accuracy": 0.23489072769880295,
      "num_tokens": 82260374.0,
      "step": 36855
    },
    {
      "entropy": 5.1977174282073975,
      "epoch": 4.139720334699837,
      "grad_norm": 1.0,
      "learning_rate": 0.0003395697581617028,
      "loss": 4.7685,
      "mean_token_accuracy": 0.23160944283008575,
      "num_tokens": 82271801.0,
      "step": 36860
    },
    {
      "entropy": 5.286045360565185,
      "epoch": 4.1402819116077945,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033953129716794396,
      "loss": 4.9227,
      "mean_token_accuracy": 0.22472193092107773,
      "num_tokens": 82283149.0,
      "step": 36865
    },
    {
      "entropy": 5.246952724456787,
      "epoch": 4.140843488515753,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003394928341194836,
      "loss": 4.7052,
      "mean_token_accuracy": 0.2377084657549858,
      "num_tokens": 82293003.0,
      "step": 36870
    },
    {
      "entropy": 5.224469232559204,
      "epoch": 4.14140506542371,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003394543690175464,
      "loss": 4.798,
      "mean_token_accuracy": 0.2345065087080002,
      "num_tokens": 82303734.0,
      "step": 36875
    },
    {
      "entropy": 5.141237020492554,
      "epoch": 4.141966642331667,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033941590186335703,
      "loss": 4.7703,
      "mean_token_accuracy": 0.23283277153968812,
      "num_tokens": 82314569.0,
      "step": 36880
    },
    {
      "entropy": 5.231340742111206,
      "epoch": 4.142528219239625,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003393774326581404,
      "loss": 4.7953,
      "mean_token_accuracy": 0.2350004270672798,
      "num_tokens": 82325174.0,
      "step": 36885
    },
    {
      "entropy": 5.2707170963287355,
      "epoch": 4.143089796147582,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033933896140312135,
      "loss": 4.8738,
      "mean_token_accuracy": 0.22915714532136916,
      "num_tokens": 82337448.0,
      "step": 36890
    },
    {
      "entropy": 5.275085592269898,
      "epoch": 4.1436513730555395,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033930048809952484,
      "loss": 4.8682,
      "mean_token_accuracy": 0.22709718644618987,
      "num_tokens": 82348705.0,
      "step": 36895
    },
    {
      "entropy": 5.310284900665283,
      "epoch": 4.144212949963498,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033926201274857575,
      "loss": 4.8604,
      "mean_token_accuracy": 0.2312955603003502,
      "num_tokens": 82360439.0,
      "step": 36900
    },
    {
      "entropy": 5.225576543807984,
      "epoch": 4.144774526871455,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033922353535149934,
      "loss": 4.7797,
      "mean_token_accuracy": 0.23361286669969558,
      "num_tokens": 82371346.0,
      "step": 36905
    },
    {
      "entropy": 5.151901435852051,
      "epoch": 4.145336103779413,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003391850559095207,
      "loss": 4.7071,
      "mean_token_accuracy": 0.23985598981380463,
      "num_tokens": 82382221.0,
      "step": 36910
    },
    {
      "entropy": 5.255051469802856,
      "epoch": 4.14589768068737,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003391465744238649,
      "loss": 4.8521,
      "mean_token_accuracy": 0.22397183626890182,
      "num_tokens": 82392290.0,
      "step": 36915
    },
    {
      "entropy": 5.279963493347168,
      "epoch": 4.146459257595327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003391080908957573,
      "loss": 4.8558,
      "mean_token_accuracy": 0.2251773476600647,
      "num_tokens": 82403513.0,
      "step": 36920
    },
    {
      "entropy": 5.212308788299561,
      "epoch": 4.1470208345032855,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033906960532642327,
      "loss": 4.7266,
      "mean_token_accuracy": 0.2398473232984543,
      "num_tokens": 82414537.0,
      "step": 36925
    },
    {
      "entropy": 5.208391714096069,
      "epoch": 4.147582411411243,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003390311177170881,
      "loss": 4.7871,
      "mean_token_accuracy": 0.23613473325967788,
      "num_tokens": 82425330.0,
      "step": 36930
    },
    {
      "entropy": 5.0979828357696535,
      "epoch": 4.1481439883192,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003389926280689773,
      "loss": 4.7467,
      "mean_token_accuracy": 0.22928581088781358,
      "num_tokens": 82435703.0,
      "step": 36935
    },
    {
      "entropy": 5.290266418457032,
      "epoch": 4.148705565227158,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033895413638331646,
      "loss": 4.8737,
      "mean_token_accuracy": 0.23298297375440596,
      "num_tokens": 82447302.0,
      "step": 36940
    },
    {
      "entropy": 5.261217737197876,
      "epoch": 4.149267142135115,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003389156426613311,
      "loss": 4.8305,
      "mean_token_accuracy": 0.22594361305236815,
      "num_tokens": 82458379.0,
      "step": 36945
    },
    {
      "entropy": 5.2348308086395265,
      "epoch": 4.149828719043073,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033887714690424687,
      "loss": 4.7755,
      "mean_token_accuracy": 0.23496630191802978,
      "num_tokens": 82469728.0,
      "step": 36950
    },
    {
      "entropy": 5.2892098903656,
      "epoch": 4.150390295951031,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033883864911328954,
      "loss": 4.8901,
      "mean_token_accuracy": 0.2242341473698616,
      "num_tokens": 82482216.0,
      "step": 36955
    },
    {
      "entropy": 5.281640434265137,
      "epoch": 4.150951872858988,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033880014928968483,
      "loss": 4.8283,
      "mean_token_accuracy": 0.2319189041852951,
      "num_tokens": 82494511.0,
      "step": 36960
    },
    {
      "entropy": 5.252456331253052,
      "epoch": 4.151513449766946,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033876164743465865,
      "loss": 4.838,
      "mean_token_accuracy": 0.23309731483459473,
      "num_tokens": 82505982.0,
      "step": 36965
    },
    {
      "entropy": 5.187915563583374,
      "epoch": 4.152075026674903,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003387231435494368,
      "loss": 4.7363,
      "mean_token_accuracy": 0.234170301258564,
      "num_tokens": 82516682.0,
      "step": 36970
    },
    {
      "entropy": 5.244225931167603,
      "epoch": 4.15263660358286,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033868463763524545,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2287895306944847,
      "num_tokens": 82527742.0,
      "step": 36975
    },
    {
      "entropy": 5.210109996795654,
      "epoch": 4.1531981804908185,
      "grad_norm": 1.0,
      "learning_rate": 0.00033864612969331046,
      "loss": 4.7613,
      "mean_token_accuracy": 0.23451885282993318,
      "num_tokens": 82538130.0,
      "step": 36980
    },
    {
      "entropy": 5.183070802688599,
      "epoch": 4.153759757398776,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003386076197248581,
      "loss": 4.7539,
      "mean_token_accuracy": 0.23214388489723206,
      "num_tokens": 82549695.0,
      "step": 36985
    },
    {
      "entropy": 5.265810966491699,
      "epoch": 4.154321334306733,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003385691077311144,
      "loss": 4.8356,
      "mean_token_accuracy": 0.228856398165226,
      "num_tokens": 82560778.0,
      "step": 36990
    },
    {
      "entropy": 5.186842966079712,
      "epoch": 4.154882911214691,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033853059371330565,
      "loss": 4.7685,
      "mean_token_accuracy": 0.23036761283874513,
      "num_tokens": 82572021.0,
      "step": 36995
    },
    {
      "entropy": 5.312889480590821,
      "epoch": 4.155444488122648,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003384920776726582,
      "loss": 4.8235,
      "mean_token_accuracy": 0.23259421139955522,
      "num_tokens": 82583173.0,
      "step": 37000
    },
    {
      "entropy": 5.2373157978057865,
      "epoch": 4.156006065030606,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033845355961039823,
      "loss": 4.7459,
      "mean_token_accuracy": 0.2359125569462776,
      "num_tokens": 82594279.0,
      "step": 37005
    },
    {
      "entropy": 5.274840450286865,
      "epoch": 4.1565676419385635,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033841503952775255,
      "loss": 4.9086,
      "mean_token_accuracy": 0.22715212851762773,
      "num_tokens": 82605038.0,
      "step": 37010
    },
    {
      "entropy": 5.328705787658691,
      "epoch": 4.157129218846521,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033837651742594734,
      "loss": 4.912,
      "mean_token_accuracy": 0.22676890790462495,
      "num_tokens": 82616627.0,
      "step": 37015
    },
    {
      "entropy": 5.1986610889434814,
      "epoch": 4.157690795754479,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033833799330620917,
      "loss": 4.8197,
      "mean_token_accuracy": 0.22507200092077256,
      "num_tokens": 82628057.0,
      "step": 37020
    },
    {
      "entropy": 5.214141511917115,
      "epoch": 4.158252372662436,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003382994671697647,
      "loss": 4.7654,
      "mean_token_accuracy": 0.22805217653512955,
      "num_tokens": 82639890.0,
      "step": 37025
    },
    {
      "entropy": 5.198433876037598,
      "epoch": 4.158813949570393,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033826093901784065,
      "loss": 4.755,
      "mean_token_accuracy": 0.22542982995510102,
      "num_tokens": 82651770.0,
      "step": 37030
    },
    {
      "entropy": 5.214186573028565,
      "epoch": 4.159375526478351,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033822240885166384,
      "loss": 4.6865,
      "mean_token_accuracy": 0.23858654946088792,
      "num_tokens": 82662159.0,
      "step": 37035
    },
    {
      "entropy": 5.237015867233277,
      "epoch": 4.159937103386309,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033818387667246095,
      "loss": 4.7453,
      "mean_token_accuracy": 0.23323326110839843,
      "num_tokens": 82674348.0,
      "step": 37040
    },
    {
      "entropy": 5.216327238082886,
      "epoch": 4.160498680294267,
      "grad_norm": 1.03125,
      "learning_rate": 0.000338145342481459,
      "loss": 4.7911,
      "mean_token_accuracy": 0.23858937323093415,
      "num_tokens": 82685674.0,
      "step": 37045
    },
    {
      "entropy": 5.320920085906982,
      "epoch": 4.161060257202224,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003381068062798848,
      "loss": 4.9135,
      "mean_token_accuracy": 0.2201535776257515,
      "num_tokens": 82696992.0,
      "step": 37050
    },
    {
      "entropy": 5.366306400299072,
      "epoch": 4.161621834110181,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003380682680689654,
      "loss": 4.864,
      "mean_token_accuracy": 0.2271619215607643,
      "num_tokens": 82708741.0,
      "step": 37055
    },
    {
      "entropy": 5.243252611160278,
      "epoch": 4.162183411018139,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033802972784992784,
      "loss": 4.8166,
      "mean_token_accuracy": 0.2321571946144104,
      "num_tokens": 82720687.0,
      "step": 37060
    },
    {
      "entropy": 5.169651222229004,
      "epoch": 4.1627449879260965,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003379911856239994,
      "loss": 4.8124,
      "mean_token_accuracy": 0.23005868941545488,
      "num_tokens": 82731863.0,
      "step": 37065
    },
    {
      "entropy": 5.195160675048828,
      "epoch": 4.163306564834054,
      "grad_norm": 1.0,
      "learning_rate": 0.0003379526413924071,
      "loss": 4.7629,
      "mean_token_accuracy": 0.23453584760427476,
      "num_tokens": 82742931.0,
      "step": 37070
    },
    {
      "entropy": 5.232096862792969,
      "epoch": 4.163868141742012,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003379140951563784,
      "loss": 4.7681,
      "mean_token_accuracy": 0.2294527292251587,
      "num_tokens": 82754244.0,
      "step": 37075
    },
    {
      "entropy": 5.171560382843017,
      "epoch": 4.164429718649969,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033787554691714033,
      "loss": 4.766,
      "mean_token_accuracy": 0.22923398911952972,
      "num_tokens": 82764929.0,
      "step": 37080
    },
    {
      "entropy": 5.190223121643067,
      "epoch": 4.164991295557926,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033783699667592066,
      "loss": 4.7381,
      "mean_token_accuracy": 0.23239219784736634,
      "num_tokens": 82775767.0,
      "step": 37085
    },
    {
      "entropy": 5.199125719070435,
      "epoch": 4.165552872465884,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033779844443394656,
      "loss": 4.7995,
      "mean_token_accuracy": 0.23100148737430573,
      "num_tokens": 82786902.0,
      "step": 37090
    },
    {
      "entropy": 5.337422227859497,
      "epoch": 4.166114449373842,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033775989019244556,
      "loss": 4.8503,
      "mean_token_accuracy": 0.22976560294628143,
      "num_tokens": 82797677.0,
      "step": 37095
    },
    {
      "entropy": 5.322229623794556,
      "epoch": 4.1666760262818,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033772133395264535,
      "loss": 4.8042,
      "mean_token_accuracy": 0.22773815542459488,
      "num_tokens": 82807246.0,
      "step": 37100
    },
    {
      "entropy": 5.238841962814331,
      "epoch": 4.167237603189757,
      "grad_norm": 0.8984375,
      "learning_rate": 0.00033768277571577364,
      "loss": 4.8054,
      "mean_token_accuracy": 0.23163640946149827,
      "num_tokens": 82818983.0,
      "step": 37105
    },
    {
      "entropy": 5.185670948028564,
      "epoch": 4.167799180097714,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033764421548305795,
      "loss": 4.7731,
      "mean_token_accuracy": 0.2347192332148552,
      "num_tokens": 82829340.0,
      "step": 37110
    },
    {
      "entropy": 5.227672958374024,
      "epoch": 4.168360757005672,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033760565325572614,
      "loss": 4.8471,
      "mean_token_accuracy": 0.22766609936952592,
      "num_tokens": 82840741.0,
      "step": 37115
    },
    {
      "entropy": 5.27437834739685,
      "epoch": 4.168922333913629,
      "grad_norm": 1.0,
      "learning_rate": 0.0003375670890350061,
      "loss": 4.8097,
      "mean_token_accuracy": 0.22757196873426438,
      "num_tokens": 82851284.0,
      "step": 37120
    },
    {
      "entropy": 5.064311933517456,
      "epoch": 4.169483910821587,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033752852282212557,
      "loss": 4.6206,
      "mean_token_accuracy": 0.23736967742443085,
      "num_tokens": 82861856.0,
      "step": 37125
    },
    {
      "entropy": 5.252471685409546,
      "epoch": 4.170045487729545,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003374899546183128,
      "loss": 4.8698,
      "mean_token_accuracy": 0.2217896819114685,
      "num_tokens": 82873354.0,
      "step": 37130
    },
    {
      "entropy": 5.265017890930176,
      "epoch": 4.170607064637502,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033745138442479544,
      "loss": 4.7938,
      "mean_token_accuracy": 0.23554246723651887,
      "num_tokens": 82883891.0,
      "step": 37135
    },
    {
      "entropy": 5.251681280136109,
      "epoch": 4.17116864154546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003374128122428019,
      "loss": 4.7116,
      "mean_token_accuracy": 0.2409118577837944,
      "num_tokens": 82893685.0,
      "step": 37140
    },
    {
      "entropy": 5.190596866607666,
      "epoch": 4.171730218453417,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003373742380735602,
      "loss": 4.793,
      "mean_token_accuracy": 0.24300142079591752,
      "num_tokens": 82904947.0,
      "step": 37145
    },
    {
      "entropy": 5.214969110488892,
      "epoch": 4.1722917953613745,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003373356619182985,
      "loss": 4.8828,
      "mean_token_accuracy": 0.22547731846570968,
      "num_tokens": 82915368.0,
      "step": 37150
    },
    {
      "entropy": 5.248837947845459,
      "epoch": 4.172853372269333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003372970837782453,
      "loss": 4.7805,
      "mean_token_accuracy": 0.23162636458873748,
      "num_tokens": 82926594.0,
      "step": 37155
    },
    {
      "entropy": 5.199381923675537,
      "epoch": 4.17341494917729,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003372585036546287,
      "loss": 4.7407,
      "mean_token_accuracy": 0.23713645339012146,
      "num_tokens": 82937699.0,
      "step": 37160
    },
    {
      "entropy": 5.268838024139404,
      "epoch": 4.173976526085247,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003372199215486772,
      "loss": 4.8688,
      "mean_token_accuracy": 0.2291138842701912,
      "num_tokens": 82948822.0,
      "step": 37165
    },
    {
      "entropy": 5.21164026260376,
      "epoch": 4.174538102993205,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003371813374616192,
      "loss": 4.7964,
      "mean_token_accuracy": 0.22755805402994156,
      "num_tokens": 82959597.0,
      "step": 37170
    },
    {
      "entropy": 5.238964462280274,
      "epoch": 4.175099679901162,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033714275139468336,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22897738665342332,
      "num_tokens": 82970740.0,
      "step": 37175
    },
    {
      "entropy": 5.302357482910156,
      "epoch": 4.17566125680912,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003371041633490982,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2232182264328003,
      "num_tokens": 82982923.0,
      "step": 37180
    },
    {
      "entropy": 5.180771207809448,
      "epoch": 4.176222833717078,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003370655733260924,
      "loss": 4.7393,
      "mean_token_accuracy": 0.2294350951910019,
      "num_tokens": 82993271.0,
      "step": 37185
    },
    {
      "entropy": 5.239220857620239,
      "epoch": 4.176784410625035,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00033702698132689467,
      "loss": 4.7953,
      "mean_token_accuracy": 0.23579512536525726,
      "num_tokens": 83004688.0,
      "step": 37190
    },
    {
      "entropy": 5.200460481643677,
      "epoch": 4.177345987532993,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003369883873527338,
      "loss": 4.8873,
      "mean_token_accuracy": 0.2244233086705208,
      "num_tokens": 83016310.0,
      "step": 37195
    },
    {
      "entropy": 5.304462623596192,
      "epoch": 4.17790756444095,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033694979140483865,
      "loss": 4.8765,
      "mean_token_accuracy": 0.22486332803964615,
      "num_tokens": 83027327.0,
      "step": 37200
    },
    {
      "entropy": 5.259927749633789,
      "epoch": 4.1784691413489075,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003369111934844381,
      "loss": 4.8253,
      "mean_token_accuracy": 0.22779244184494019,
      "num_tokens": 83038493.0,
      "step": 37205
    },
    {
      "entropy": 5.234377527236939,
      "epoch": 4.179030718256866,
      "grad_norm": 1.0,
      "learning_rate": 0.00033687259359276115,
      "loss": 4.7554,
      "mean_token_accuracy": 0.22995853126049043,
      "num_tokens": 83050082.0,
      "step": 37210
    },
    {
      "entropy": 5.195677757263184,
      "epoch": 4.179592295164823,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003368339917310368,
      "loss": 4.7317,
      "mean_token_accuracy": 0.23554655760526658,
      "num_tokens": 83061234.0,
      "step": 37215
    },
    {
      "entropy": 5.2066182613372805,
      "epoch": 4.18015387207278,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003367953879004942,
      "loss": 4.7734,
      "mean_token_accuracy": 0.2286195397377014,
      "num_tokens": 83071454.0,
      "step": 37220
    },
    {
      "entropy": 5.222982263565063,
      "epoch": 4.180715448980738,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003367567821023625,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22561008036136626,
      "num_tokens": 83083204.0,
      "step": 37225
    },
    {
      "entropy": 5.21359806060791,
      "epoch": 4.181277025888695,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003367181743378708,
      "loss": 4.7588,
      "mean_token_accuracy": 0.23595773577690124,
      "num_tokens": 83094654.0,
      "step": 37230
    },
    {
      "entropy": 5.2857788562774655,
      "epoch": 4.181838602796653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033667956460824866,
      "loss": 4.8645,
      "mean_token_accuracy": 0.2235479697585106,
      "num_tokens": 83105539.0,
      "step": 37235
    },
    {
      "entropy": 5.235623168945312,
      "epoch": 4.182400179704611,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003366409529147251,
      "loss": 4.7884,
      "mean_token_accuracy": 0.2342590183019638,
      "num_tokens": 83115741.0,
      "step": 37240
    },
    {
      "entropy": 5.2152739524841305,
      "epoch": 4.182961756612568,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033660233925852977,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2335145205259323,
      "num_tokens": 83126739.0,
      "step": 37245
    },
    {
      "entropy": 5.15121750831604,
      "epoch": 4.183523333520526,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033656372364089206,
      "loss": 4.6968,
      "mean_token_accuracy": 0.2393274948000908,
      "num_tokens": 83138821.0,
      "step": 37250
    },
    {
      "entropy": 5.2333595752716064,
      "epoch": 4.184084910428483,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033652510606304155,
      "loss": 4.8807,
      "mean_token_accuracy": 0.22700828760862352,
      "num_tokens": 83149757.0,
      "step": 37255
    },
    {
      "entropy": 5.25309271812439,
      "epoch": 4.18464648733644,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033648648652620766,
      "loss": 4.8428,
      "mean_token_accuracy": 0.2265539824962616,
      "num_tokens": 83161627.0,
      "step": 37260
    },
    {
      "entropy": 5.208529758453369,
      "epoch": 4.1852080642443985,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033644786503162036,
      "loss": 4.7496,
      "mean_token_accuracy": 0.23451498299837112,
      "num_tokens": 83172903.0,
      "step": 37265
    },
    {
      "entropy": 5.103759860992431,
      "epoch": 4.185769641152356,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003364092415805091,
      "loss": 4.693,
      "mean_token_accuracy": 0.23276172280311586,
      "num_tokens": 83184142.0,
      "step": 37270
    },
    {
      "entropy": 5.253635263442993,
      "epoch": 4.186331218060313,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033637061617410377,
      "loss": 4.841,
      "mean_token_accuracy": 0.22986388057470322,
      "num_tokens": 83195206.0,
      "step": 37275
    },
    {
      "entropy": 5.235664701461792,
      "epoch": 4.186892794968271,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003363319888136343,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2374179482460022,
      "num_tokens": 83207579.0,
      "step": 37280
    },
    {
      "entropy": 5.308981943130493,
      "epoch": 4.187454371876228,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003362933595003305,
      "loss": 4.8583,
      "mean_token_accuracy": 0.23219441175460814,
      "num_tokens": 83219964.0,
      "step": 37285
    },
    {
      "entropy": 5.212917995452881,
      "epoch": 4.188015948784186,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003362547282354223,
      "loss": 4.7589,
      "mean_token_accuracy": 0.2306332156062126,
      "num_tokens": 83231241.0,
      "step": 37290
    },
    {
      "entropy": 5.155891084671021,
      "epoch": 4.188577525692144,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003362160950201398,
      "loss": 4.7167,
      "mean_token_accuracy": 0.23954061418771744,
      "num_tokens": 83241784.0,
      "step": 37295
    },
    {
      "entropy": 5.185430240631104,
      "epoch": 4.189139102600101,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003361774598557132,
      "loss": 4.7411,
      "mean_token_accuracy": 0.23146923780441284,
      "num_tokens": 83252815.0,
      "step": 37300
    },
    {
      "entropy": 5.211003017425537,
      "epoch": 4.189700679508059,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033613882274337244,
      "loss": 4.8066,
      "mean_token_accuracy": 0.2381764605641365,
      "num_tokens": 83263938.0,
      "step": 37305
    },
    {
      "entropy": 5.2900137424469,
      "epoch": 4.190262256416016,
      "grad_norm": 1.125,
      "learning_rate": 0.0003361001836843479,
      "loss": 4.8672,
      "mean_token_accuracy": 0.22724520713090895,
      "num_tokens": 83274746.0,
      "step": 37310
    },
    {
      "entropy": 5.319981813430786,
      "epoch": 4.190823833323973,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003360615426798697,
      "loss": 4.9395,
      "mean_token_accuracy": 0.22289653569459916,
      "num_tokens": 83285707.0,
      "step": 37315
    },
    {
      "entropy": 5.2986267566680905,
      "epoch": 4.1913854102319315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003360228997311685,
      "loss": 4.8329,
      "mean_token_accuracy": 0.23218328207731248,
      "num_tokens": 83297175.0,
      "step": 37320
    },
    {
      "entropy": 5.2206212997436525,
      "epoch": 4.191946987139889,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003359842548394744,
      "loss": 4.7139,
      "mean_token_accuracy": 0.23874115496873854,
      "num_tokens": 83307731.0,
      "step": 37325
    },
    {
      "entropy": 5.2265218734741214,
      "epoch": 4.192508564047847,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003359456080060179,
      "loss": 4.803,
      "mean_token_accuracy": 0.23544129878282546,
      "num_tokens": 83318698.0,
      "step": 37330
    },
    {
      "entropy": 5.164061069488525,
      "epoch": 4.193070140955804,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003359069592320297,
      "loss": 4.7355,
      "mean_token_accuracy": 0.23492133021354675,
      "num_tokens": 83330085.0,
      "step": 37335
    },
    {
      "entropy": 5.278609848022461,
      "epoch": 4.193631717863761,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003358683085187403,
      "loss": 4.8462,
      "mean_token_accuracy": 0.22699519544839858,
      "num_tokens": 83340606.0,
      "step": 37340
    },
    {
      "entropy": 5.1919787406921385,
      "epoch": 4.194193294771719,
      "grad_norm": 1.0,
      "learning_rate": 0.00033582965586738037,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2349713608622551,
      "num_tokens": 83352177.0,
      "step": 37345
    },
    {
      "entropy": 5.279533433914184,
      "epoch": 4.1947548716796765,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003357910012791804,
      "loss": 4.9325,
      "mean_token_accuracy": 0.2230362981557846,
      "num_tokens": 83362474.0,
      "step": 37350
    },
    {
      "entropy": 5.28324327468872,
      "epoch": 4.195316448587634,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003357523447553716,
      "loss": 4.8502,
      "mean_token_accuracy": 0.2276044487953186,
      "num_tokens": 83373187.0,
      "step": 37355
    },
    {
      "entropy": 5.208130550384522,
      "epoch": 4.195878025495592,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033571368629718457,
      "loss": 4.7328,
      "mean_token_accuracy": 0.2362465441226959,
      "num_tokens": 83385110.0,
      "step": 37360
    },
    {
      "entropy": 5.235625886917115,
      "epoch": 4.196439602403549,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033567502590585007,
      "loss": 4.8269,
      "mean_token_accuracy": 0.2355415016412735,
      "num_tokens": 83396692.0,
      "step": 37365
    },
    {
      "entropy": 5.24248423576355,
      "epoch": 4.197001179311506,
      "grad_norm": 1.0,
      "learning_rate": 0.0003356363635825994,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2375287890434265,
      "num_tokens": 83408433.0,
      "step": 37370
    },
    {
      "entropy": 5.291813468933105,
      "epoch": 4.197562756219464,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003355976993286633,
      "loss": 4.8278,
      "mean_token_accuracy": 0.23197529166936875,
      "num_tokens": 83419169.0,
      "step": 37375
    },
    {
      "entropy": 5.201872444152832,
      "epoch": 4.198124333127422,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003355590331452729,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23923024982213975,
      "num_tokens": 83430247.0,
      "step": 37380
    },
    {
      "entropy": 5.178793907165527,
      "epoch": 4.19868591003538,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00033552036503365945,
      "loss": 4.7034,
      "mean_token_accuracy": 0.2379392370581627,
      "num_tokens": 83442294.0,
      "step": 37385
    },
    {
      "entropy": 5.150610446929932,
      "epoch": 4.199247486943337,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003354816949950541,
      "loss": 4.7387,
      "mean_token_accuracy": 0.23963966071605683,
      "num_tokens": 83453388.0,
      "step": 37390
    },
    {
      "entropy": 5.13798246383667,
      "epoch": 4.199809063851294,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033544302303068814,
      "loss": 4.7547,
      "mean_token_accuracy": 0.24122246950864792,
      "num_tokens": 83465151.0,
      "step": 37395
    },
    {
      "entropy": 5.259822750091553,
      "epoch": 4.200370640759252,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003354043491417929,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2301338329911232,
      "num_tokens": 83474731.0,
      "step": 37400
    },
    {
      "entropy": 5.193684196472168,
      "epoch": 4.2009322176672095,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033536567332959966,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23209916949272155,
      "num_tokens": 83486456.0,
      "step": 37405
    },
    {
      "entropy": 5.210886240005493,
      "epoch": 4.201493794575167,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033532699559534,
      "loss": 4.7738,
      "mean_token_accuracy": 0.23258630484342574,
      "num_tokens": 83496503.0,
      "step": 37410
    },
    {
      "entropy": 5.365537118911743,
      "epoch": 4.202055371483125,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033528831594024553,
      "loss": 4.8863,
      "mean_token_accuracy": 0.2246316909790039,
      "num_tokens": 83508198.0,
      "step": 37415
    },
    {
      "entropy": 5.24492301940918,
      "epoch": 4.202616948391082,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033524963436554753,
      "loss": 4.779,
      "mean_token_accuracy": 0.2305237263441086,
      "num_tokens": 83520255.0,
      "step": 37420
    },
    {
      "entropy": 5.207711601257325,
      "epoch": 4.20317852529904,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033521095087247794,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23055752813816072,
      "num_tokens": 83532083.0,
      "step": 37425
    },
    {
      "entropy": 5.300134992599487,
      "epoch": 4.203740102206997,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003351722654622683,
      "loss": 4.8784,
      "mean_token_accuracy": 0.23078581392765046,
      "num_tokens": 83542891.0,
      "step": 37430
    },
    {
      "entropy": 5.260670375823975,
      "epoch": 4.204301679114955,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003351335781361503,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2385306939482689,
      "num_tokens": 83554238.0,
      "step": 37435
    },
    {
      "entropy": 5.248376703262329,
      "epoch": 4.204863256022913,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033509488889535595,
      "loss": 4.8657,
      "mean_token_accuracy": 0.2203829064965248,
      "num_tokens": 83565681.0,
      "step": 37440
    },
    {
      "entropy": 5.119304418563843,
      "epoch": 4.20542483293087,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033505619774111696,
      "loss": 4.6469,
      "mean_token_accuracy": 0.2464430332183838,
      "num_tokens": 83576844.0,
      "step": 37445
    },
    {
      "entropy": 5.214753675460815,
      "epoch": 4.205986409838827,
      "grad_norm": 1.0,
      "learning_rate": 0.00033501750467466535,
      "loss": 4.7663,
      "mean_token_accuracy": 0.2412422314286232,
      "num_tokens": 83589501.0,
      "step": 37450
    },
    {
      "entropy": 5.275656604766846,
      "epoch": 4.206547986746785,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003349788096972331,
      "loss": 4.8504,
      "mean_token_accuracy": 0.23320903331041337,
      "num_tokens": 83600305.0,
      "step": 37455
    },
    {
      "entropy": 5.252416753768921,
      "epoch": 4.207109563654742,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003349401128100523,
      "loss": 4.8572,
      "mean_token_accuracy": 0.22384567707777023,
      "num_tokens": 83611908.0,
      "step": 37460
    },
    {
      "entropy": 5.209652662277222,
      "epoch": 4.2076711405627,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00033490141401435516,
      "loss": 4.7898,
      "mean_token_accuracy": 0.22923975735902785,
      "num_tokens": 83624156.0,
      "step": 37465
    },
    {
      "entropy": 5.186652565002442,
      "epoch": 4.208232717470658,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003348627133113736,
      "loss": 4.6858,
      "mean_token_accuracy": 0.24148286134004593,
      "num_tokens": 83634655.0,
      "step": 37470
    },
    {
      "entropy": 5.1406590938568115,
      "epoch": 4.208794294378615,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033482401070234006,
      "loss": 4.7238,
      "mean_token_accuracy": 0.2407629296183586,
      "num_tokens": 83645015.0,
      "step": 37475
    },
    {
      "entropy": 5.165185356140137,
      "epoch": 4.209355871286573,
      "grad_norm": 0.93359375,
      "learning_rate": 0.000334785306188487,
      "loss": 4.7897,
      "mean_token_accuracy": 0.23637537956237792,
      "num_tokens": 83656540.0,
      "step": 37480
    },
    {
      "entropy": 5.238648319244385,
      "epoch": 4.20991744819453,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003347465997710464,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23568926751613617,
      "num_tokens": 83666798.0,
      "step": 37485
    },
    {
      "entropy": 5.200516843795777,
      "epoch": 4.2104790251024875,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000334707891451251,
      "loss": 4.7511,
      "mean_token_accuracy": 0.23831358551979065,
      "num_tokens": 83677058.0,
      "step": 37490
    },
    {
      "entropy": 5.184558868408203,
      "epoch": 4.211040602010446,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003346691812303331,
      "loss": 4.7115,
      "mean_token_accuracy": 0.2340506508946419,
      "num_tokens": 83688126.0,
      "step": 37495
    },
    {
      "entropy": 5.262938976287842,
      "epoch": 4.211602178918403,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003346304691095254,
      "loss": 4.9207,
      "mean_token_accuracy": 0.22474397867918014,
      "num_tokens": 83700298.0,
      "step": 37500
    },
    {
      "entropy": 5.181654071807861,
      "epoch": 4.21216375582636,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003345917550900604,
      "loss": 4.6923,
      "mean_token_accuracy": 0.23494734168052672,
      "num_tokens": 83711037.0,
      "step": 37505
    },
    {
      "entropy": 5.236189651489258,
      "epoch": 4.212725332734318,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00033455303917317075,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23903975635766983,
      "num_tokens": 83722306.0,
      "step": 37510
    },
    {
      "entropy": 5.242357778549194,
      "epoch": 4.213286909642275,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00033451432136008937,
      "loss": 4.8091,
      "mean_token_accuracy": 0.23181018382310867,
      "num_tokens": 83733306.0,
      "step": 37515
    },
    {
      "entropy": 5.150763320922851,
      "epoch": 4.2138484865502335,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003344756016520488,
      "loss": 4.7496,
      "mean_token_accuracy": 0.2318197950720787,
      "num_tokens": 83745475.0,
      "step": 37520
    },
    {
      "entropy": 5.189968776702881,
      "epoch": 4.214410063458191,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033443688005028213,
      "loss": 4.7206,
      "mean_token_accuracy": 0.23475961685180663,
      "num_tokens": 83756992.0,
      "step": 37525
    },
    {
      "entropy": 5.187186336517334,
      "epoch": 4.214971640366148,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000334398156556022,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23215043246746064,
      "num_tokens": 83768881.0,
      "step": 37530
    },
    {
      "entropy": 5.347860860824585,
      "epoch": 4.215533217274106,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003343594311705016,
      "loss": 4.8436,
      "mean_token_accuracy": 0.22959304451942444,
      "num_tokens": 83781144.0,
      "step": 37535
    },
    {
      "entropy": 5.2881340980529785,
      "epoch": 4.216094794182063,
      "grad_norm": 0.96875,
      "learning_rate": 0.00033432070389495386,
      "loss": 4.9094,
      "mean_token_accuracy": 0.22623234242200851,
      "num_tokens": 83792449.0,
      "step": 37540
    },
    {
      "entropy": 5.196537590026855,
      "epoch": 4.2166563710900204,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003342819747306119,
      "loss": 4.6991,
      "mean_token_accuracy": 0.24051156789064407,
      "num_tokens": 83802889.0,
      "step": 37545
    },
    {
      "entropy": 5.255451345443726,
      "epoch": 4.2172179479979786,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033424324367870896,
      "loss": 4.8623,
      "mean_token_accuracy": 0.22478122413158416,
      "num_tokens": 83813527.0,
      "step": 37550
    },
    {
      "entropy": 5.103975439071656,
      "epoch": 4.217779524905936,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033420451074047807,
      "loss": 4.7375,
      "mean_token_accuracy": 0.241282095015049,
      "num_tokens": 83824963.0,
      "step": 37555
    },
    {
      "entropy": 5.268413496017456,
      "epoch": 4.218341101813893,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003341657759171526,
      "loss": 4.8684,
      "mean_token_accuracy": 0.22662003636360167,
      "num_tokens": 83836188.0,
      "step": 37560
    },
    {
      "entropy": 5.272253942489624,
      "epoch": 4.218902678721851,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003341270392099659,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23708248436450957,
      "num_tokens": 83848297.0,
      "step": 37565
    },
    {
      "entropy": 5.185402965545654,
      "epoch": 4.219464255629808,
      "grad_norm": 1.015625,
      "learning_rate": 0.00033408830062015135,
      "loss": 4.743,
      "mean_token_accuracy": 0.22988543808460235,
      "num_tokens": 83859074.0,
      "step": 37570
    },
    {
      "entropy": 5.148447704315186,
      "epoch": 4.220025832537766,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033404956014894237,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23124044984579087,
      "num_tokens": 83869921.0,
      "step": 37575
    },
    {
      "entropy": 5.218757390975952,
      "epoch": 4.220587409445724,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033401081779757254,
      "loss": 4.7951,
      "mean_token_accuracy": 0.23044316172599794,
      "num_tokens": 83881994.0,
      "step": 37580
    },
    {
      "entropy": 5.242766380310059,
      "epoch": 4.221148986353681,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003339720735672753,
      "loss": 4.8186,
      "mean_token_accuracy": 0.22568356841802598,
      "num_tokens": 83893530.0,
      "step": 37585
    },
    {
      "entropy": 5.207434368133545,
      "epoch": 4.221710563261639,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033393332745928453,
      "loss": 4.7522,
      "mean_token_accuracy": 0.23412871360778809,
      "num_tokens": 83904929.0,
      "step": 37590
    },
    {
      "entropy": 5.183398056030273,
      "epoch": 4.222272140169596,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033389457947483357,
      "loss": 4.7322,
      "mean_token_accuracy": 0.23536174595355988,
      "num_tokens": 83915888.0,
      "step": 37595
    },
    {
      "entropy": 5.206597089767456,
      "epoch": 4.222833717077553,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00033385582961515645,
      "loss": 4.8322,
      "mean_token_accuracy": 0.2325932964682579,
      "num_tokens": 83927481.0,
      "step": 37600
    },
    {
      "entropy": 5.269898319244385,
      "epoch": 4.2233952939855115,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033381707788148683,
      "loss": 4.8492,
      "mean_token_accuracy": 0.22885286211967468,
      "num_tokens": 83938644.0,
      "step": 37605
    },
    {
      "entropy": 5.177970218658447,
      "epoch": 4.223956870893469,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033377832427505864,
      "loss": 4.7282,
      "mean_token_accuracy": 0.24708638787269593,
      "num_tokens": 83949165.0,
      "step": 37610
    },
    {
      "entropy": 5.248934888839722,
      "epoch": 4.224518447801427,
      "grad_norm": 0.921875,
      "learning_rate": 0.00033373956879710593,
      "loss": 4.7927,
      "mean_token_accuracy": 0.22918577790260314,
      "num_tokens": 83961112.0,
      "step": 37615
    },
    {
      "entropy": 5.238050174713135,
      "epoch": 4.225080024709384,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033370081144886246,
      "loss": 4.8143,
      "mean_token_accuracy": 0.24003121256828308,
      "num_tokens": 83972451.0,
      "step": 37620
    },
    {
      "entropy": 5.141804885864258,
      "epoch": 4.225641601617341,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00033366205223156237,
      "loss": 4.6973,
      "mean_token_accuracy": 0.2381664216518402,
      "num_tokens": 83984082.0,
      "step": 37625
    },
    {
      "entropy": 5.308593654632569,
      "epoch": 4.226203178525299,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033362329114643987,
      "loss": 4.9288,
      "mean_token_accuracy": 0.21595220118761063,
      "num_tokens": 83995853.0,
      "step": 37630
    },
    {
      "entropy": 5.299464368820191,
      "epoch": 4.226764755433257,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033358452819472895,
      "loss": 4.8416,
      "mean_token_accuracy": 0.2277924582362175,
      "num_tokens": 84006440.0,
      "step": 37635
    },
    {
      "entropy": 5.35287184715271,
      "epoch": 4.227326332341214,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033354576337766396,
      "loss": 4.9405,
      "mean_token_accuracy": 0.22841888964176177,
      "num_tokens": 84018968.0,
      "step": 37640
    },
    {
      "entropy": 5.240641021728516,
      "epoch": 4.227887909249172,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003335069966964792,
      "loss": 4.7927,
      "mean_token_accuracy": 0.23986612409353256,
      "num_tokens": 84030607.0,
      "step": 37645
    },
    {
      "entropy": 5.199951028823852,
      "epoch": 4.228449486157129,
      "grad_norm": 0.984375,
      "learning_rate": 0.000333468228152409,
      "loss": 4.7752,
      "mean_token_accuracy": 0.2332095295190811,
      "num_tokens": 84042598.0,
      "step": 37650
    },
    {
      "entropy": 5.126388216018677,
      "epoch": 4.229011063065086,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033342945774668766,
      "loss": 4.713,
      "mean_token_accuracy": 0.23619846850633622,
      "num_tokens": 84052708.0,
      "step": 37655
    },
    {
      "entropy": 5.258275365829467,
      "epoch": 4.2295726399730444,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00033339068548054974,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22716949880123138,
      "num_tokens": 84063573.0,
      "step": 37660
    },
    {
      "entropy": 5.343243074417114,
      "epoch": 4.230134216881002,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033335191135522974,
      "loss": 4.817,
      "mean_token_accuracy": 0.22897749692201613,
      "num_tokens": 84074132.0,
      "step": 37665
    },
    {
      "entropy": 5.14973464012146,
      "epoch": 4.23069579378896,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033331313537196235,
      "loss": 4.7175,
      "mean_token_accuracy": 0.2562312766909599,
      "num_tokens": 84086424.0,
      "step": 37670
    },
    {
      "entropy": 5.254385280609131,
      "epoch": 4.231257370696917,
      "grad_norm": 1.09375,
      "learning_rate": 0.000333274357531982,
      "loss": 4.8362,
      "mean_token_accuracy": 0.22931170463562012,
      "num_tokens": 84098413.0,
      "step": 37675
    },
    {
      "entropy": 5.120986080169677,
      "epoch": 4.231818947604874,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003332355778365235,
      "loss": 4.7702,
      "mean_token_accuracy": 0.2263793647289276,
      "num_tokens": 84109156.0,
      "step": 37680
    },
    {
      "entropy": 5.204178285598755,
      "epoch": 4.232380524512832,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003331967962868217,
      "loss": 4.686,
      "mean_token_accuracy": 0.23418887406587602,
      "num_tokens": 84120677.0,
      "step": 37685
    },
    {
      "entropy": 5.238518714904785,
      "epoch": 4.2329421014207895,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003331580128841112,
      "loss": 4.79,
      "mean_token_accuracy": 0.23036497533321382,
      "num_tokens": 84132607.0,
      "step": 37690
    },
    {
      "entropy": 5.2430027484893795,
      "epoch": 4.233503678328747,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033311922762962716,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23630893379449844,
      "num_tokens": 84143958.0,
      "step": 37695
    },
    {
      "entropy": 5.296444988250732,
      "epoch": 4.234065255236705,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003330804405246043,
      "loss": 4.9047,
      "mean_token_accuracy": 0.22755301147699356,
      "num_tokens": 84155154.0,
      "step": 37700
    },
    {
      "entropy": 5.247061491012573,
      "epoch": 4.234626832144662,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003330416515702777,
      "loss": 4.8261,
      "mean_token_accuracy": 0.2284989759325981,
      "num_tokens": 84166100.0,
      "step": 37705
    },
    {
      "entropy": 5.294345569610596,
      "epoch": 4.23518840905262,
      "grad_norm": 1.046875,
      "learning_rate": 0.00033300286076788237,
      "loss": 4.849,
      "mean_token_accuracy": 0.2346808657050133,
      "num_tokens": 84177249.0,
      "step": 37710
    },
    {
      "entropy": 5.28318419456482,
      "epoch": 4.235749985960577,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003329640681186535,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23251403570175172,
      "num_tokens": 84187922.0,
      "step": 37715
    },
    {
      "entropy": 5.192152452468872,
      "epoch": 4.236311562868535,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033292527362382626,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23978643268346786,
      "num_tokens": 84198929.0,
      "step": 37720
    },
    {
      "entropy": 5.16879711151123,
      "epoch": 4.236873139776493,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033288647728463576,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22902590930461883,
      "num_tokens": 84209503.0,
      "step": 37725
    },
    {
      "entropy": 5.276904773712158,
      "epoch": 4.23743471668445,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033284767910231745,
      "loss": 4.873,
      "mean_token_accuracy": 0.22085750102996826,
      "num_tokens": 84220667.0,
      "step": 37730
    },
    {
      "entropy": 5.284882307052612,
      "epoch": 4.237996293592407,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003328088790781065,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2331231266260147,
      "num_tokens": 84232072.0,
      "step": 37735
    },
    {
      "entropy": 5.2896956443786625,
      "epoch": 4.238557870500365,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00033277007721323854,
      "loss": 4.8471,
      "mean_token_accuracy": 0.23242532163858415,
      "num_tokens": 84243765.0,
      "step": 37740
    },
    {
      "entropy": 5.311826610565186,
      "epoch": 4.2391194474083225,
      "grad_norm": 1.03125,
      "learning_rate": 0.00033273127350894884,
      "loss": 4.8705,
      "mean_token_accuracy": 0.2277536228299141,
      "num_tokens": 84254869.0,
      "step": 37745
    },
    {
      "entropy": 5.29176778793335,
      "epoch": 4.23968102431628,
      "grad_norm": 1.21875,
      "learning_rate": 0.00033269246796647296,
      "loss": 4.8771,
      "mean_token_accuracy": 0.22181699872016908,
      "num_tokens": 84264992.0,
      "step": 37750
    },
    {
      "entropy": 5.223863983154297,
      "epoch": 4.240242601224238,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003326536605870466,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23943956792354584,
      "num_tokens": 84275648.0,
      "step": 37755
    },
    {
      "entropy": 5.315495014190674,
      "epoch": 4.240804178132195,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033261485137190533,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23198092430830003,
      "num_tokens": 84287217.0,
      "step": 37760
    },
    {
      "entropy": 5.2217309951782225,
      "epoch": 4.241365755040153,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00033257604032228485,
      "loss": 4.8786,
      "mean_token_accuracy": 0.2222852662205696,
      "num_tokens": 84298764.0,
      "step": 37765
    },
    {
      "entropy": 5.289371347427368,
      "epoch": 4.24192733194811,
      "grad_norm": 0.953125,
      "learning_rate": 0.00033253722743942086,
      "loss": 4.8829,
      "mean_token_accuracy": 0.2288201257586479,
      "num_tokens": 84310016.0,
      "step": 37770
    },
    {
      "entropy": 5.241707277297974,
      "epoch": 4.2424889088560676,
      "grad_norm": 1.0,
      "learning_rate": 0.0003324984127245492,
      "loss": 4.8093,
      "mean_token_accuracy": 0.22798689901828767,
      "num_tokens": 84321836.0,
      "step": 37775
    },
    {
      "entropy": 5.193166065216064,
      "epoch": 4.243050485764026,
      "grad_norm": 1.0,
      "learning_rate": 0.0003324595961789058,
      "loss": 4.7065,
      "mean_token_accuracy": 0.23331710547208787,
      "num_tokens": 84332434.0,
      "step": 37780
    },
    {
      "entropy": 5.1280755519866945,
      "epoch": 4.243612062671983,
      "grad_norm": 0.984375,
      "learning_rate": 0.00033242077780372664,
      "loss": 4.7154,
      "mean_token_accuracy": 0.24099546521902085,
      "num_tokens": 84343997.0,
      "step": 37785
    },
    {
      "entropy": 5.222040224075317,
      "epoch": 4.24417363957994,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033238195760024766,
      "loss": 4.7638,
      "mean_token_accuracy": 0.24224919229745864,
      "num_tokens": 84354157.0,
      "step": 37790
    },
    {
      "entropy": 5.146730375289917,
      "epoch": 4.244735216487898,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033234313556970485,
      "loss": 4.6831,
      "mean_token_accuracy": 0.243278868496418,
      "num_tokens": 84365888.0,
      "step": 37795
    },
    {
      "entropy": 5.228413867950439,
      "epoch": 4.245296793395855,
      "grad_norm": 1.109375,
      "learning_rate": 0.00033230431171333435,
      "loss": 4.7801,
      "mean_token_accuracy": 0.23481448590755463,
      "num_tokens": 84376121.0,
      "step": 37800
    },
    {
      "entropy": 5.153992319107056,
      "epoch": 4.2458583703038135,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003322654860323723,
      "loss": 4.6572,
      "mean_token_accuracy": 0.2411666065454483,
      "num_tokens": 84387318.0,
      "step": 37805
    },
    {
      "entropy": 5.209800386428833,
      "epoch": 4.246419947211771,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033222665852805514,
      "loss": 4.7916,
      "mean_token_accuracy": 0.22401203066110612,
      "num_tokens": 84398952.0,
      "step": 37810
    },
    {
      "entropy": 5.222646808624267,
      "epoch": 4.246981524119728,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00033218782920161884,
      "loss": 4.8128,
      "mean_token_accuracy": 0.22499121725559235,
      "num_tokens": 84410010.0,
      "step": 37815
    },
    {
      "entropy": 5.254630422592163,
      "epoch": 4.247543101027686,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033214899805429984,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23295217603445054,
      "num_tokens": 84421879.0,
      "step": 37820
    },
    {
      "entropy": 5.296339321136474,
      "epoch": 4.248104677935643,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00033211016508733463,
      "loss": 4.8852,
      "mean_token_accuracy": 0.2245093360543251,
      "num_tokens": 84432811.0,
      "step": 37825
    },
    {
      "entropy": 5.246522092819214,
      "epoch": 4.2486662548436005,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033207133030195964,
      "loss": 4.8547,
      "mean_token_accuracy": 0.2339807227253914,
      "num_tokens": 84444330.0,
      "step": 37830
    },
    {
      "entropy": 5.2308876514434814,
      "epoch": 4.249227831751559,
      "grad_norm": 1.0,
      "learning_rate": 0.0003320324936994113,
      "loss": 4.774,
      "mean_token_accuracy": 0.23985983431339264,
      "num_tokens": 84455827.0,
      "step": 37835
    },
    {
      "entropy": 5.209240531921386,
      "epoch": 4.249789408659516,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00033199365528092625,
      "loss": 4.7158,
      "mean_token_accuracy": 0.23743005394935607,
      "num_tokens": 84465608.0,
      "step": 37840
    },
    {
      "entropy": 5.207386875152588,
      "epoch": 4.250350985567474,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033195481504774113,
      "loss": 4.7019,
      "mean_token_accuracy": 0.23445859998464585,
      "num_tokens": 84476142.0,
      "step": 37845
    },
    {
      "entropy": 5.2732030868530275,
      "epoch": 4.250912562475431,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033191597300109263,
      "loss": 4.8133,
      "mean_token_accuracy": 0.23052880316972732,
      "num_tokens": 84486496.0,
      "step": 37850
    },
    {
      "entropy": 5.160842037200927,
      "epoch": 4.251474139383388,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033187712914221744,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2404147833585739,
      "num_tokens": 84497529.0,
      "step": 37855
    },
    {
      "entropy": 5.188924169540405,
      "epoch": 4.2520357162913465,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003318382834723524,
      "loss": 4.7051,
      "mean_token_accuracy": 0.23600502461194992,
      "num_tokens": 84508016.0,
      "step": 37860
    },
    {
      "entropy": 5.340127086639404,
      "epoch": 4.252597293199304,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033179943599273445,
      "loss": 4.9092,
      "mean_token_accuracy": 0.22522350996732712,
      "num_tokens": 84519787.0,
      "step": 37865
    },
    {
      "entropy": 5.198639822006226,
      "epoch": 4.253158870107261,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003317605867046003,
      "loss": 4.7073,
      "mean_token_accuracy": 0.2267652541399002,
      "num_tokens": 84530908.0,
      "step": 37870
    },
    {
      "entropy": 5.329192447662353,
      "epoch": 4.253720447015219,
      "grad_norm": 1.0625,
      "learning_rate": 0.00033172173560918716,
      "loss": 4.9302,
      "mean_token_accuracy": 0.22215917110443115,
      "num_tokens": 84543502.0,
      "step": 37875
    },
    {
      "entropy": 5.181834363937378,
      "epoch": 4.254282023923176,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033168288270773197,
      "loss": 4.7721,
      "mean_token_accuracy": 0.23511089980602265,
      "num_tokens": 84554500.0,
      "step": 37880
    },
    {
      "entropy": 5.315736913681031,
      "epoch": 4.254843600831133,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033164402800147177,
      "loss": 4.9569,
      "mean_token_accuracy": 0.22303521633148193,
      "num_tokens": 84566704.0,
      "step": 37885
    },
    {
      "entropy": 5.236495161056519,
      "epoch": 4.2554051777390915,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003316051714916436,
      "loss": 4.7612,
      "mean_token_accuracy": 0.23442024141550064,
      "num_tokens": 84578264.0,
      "step": 37890
    },
    {
      "entropy": 5.15957989692688,
      "epoch": 4.255966754647049,
      "grad_norm": 1.078125,
      "learning_rate": 0.00033156631317948505,
      "loss": 4.5717,
      "mean_token_accuracy": 0.25482759177684783,
      "num_tokens": 84588513.0,
      "step": 37895
    },
    {
      "entropy": 5.195415592193603,
      "epoch": 4.256528331555007,
      "grad_norm": 1.0,
      "learning_rate": 0.000331527453066233,
      "loss": 4.7606,
      "mean_token_accuracy": 0.24289042353630066,
      "num_tokens": 84600384.0,
      "step": 37900
    },
    {
      "entropy": 5.30053071975708,
      "epoch": 4.257089908462964,
      "grad_norm": 1.09375,
      "learning_rate": 0.000331488591153125,
      "loss": 4.8173,
      "mean_token_accuracy": 0.22964724600315095,
      "num_tokens": 84610827.0,
      "step": 37905
    },
    {
      "entropy": 5.1816565990448,
      "epoch": 4.257651485370921,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003314497274413983,
      "loss": 4.7195,
      "mean_token_accuracy": 0.23988064974546433,
      "num_tokens": 84621130.0,
      "step": 37910
    },
    {
      "entropy": 5.22138090133667,
      "epoch": 4.258213062278879,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033141086193229043,
      "loss": 4.8348,
      "mean_token_accuracy": 0.23469134122133256,
      "num_tokens": 84633428.0,
      "step": 37915
    },
    {
      "entropy": 5.23441972732544,
      "epoch": 4.258774639186837,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003313719946270388,
      "loss": 4.8182,
      "mean_token_accuracy": 0.22783771753311158,
      "num_tokens": 84644519.0,
      "step": 37920
    },
    {
      "entropy": 5.296269226074219,
      "epoch": 4.259336216094794,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033133312552688094,
      "loss": 4.8825,
      "mean_token_accuracy": 0.22636161893606185,
      "num_tokens": 84654752.0,
      "step": 37925
    },
    {
      "entropy": 5.245888423919678,
      "epoch": 4.259897793002752,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00033129425463305465,
      "loss": 4.7464,
      "mean_token_accuracy": 0.2435311883687973,
      "num_tokens": 84666384.0,
      "step": 37930
    },
    {
      "entropy": 5.3172533988952635,
      "epoch": 4.260459369910709,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003312553819467974,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23032695651054383,
      "num_tokens": 84676630.0,
      "step": 37935
    },
    {
      "entropy": 5.146059131622314,
      "epoch": 4.261020946818666,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00033121650746934693,
      "loss": 4.7688,
      "mean_token_accuracy": 0.2367338180541992,
      "num_tokens": 84689407.0,
      "step": 37940
    },
    {
      "entropy": 5.2316100120544435,
      "epoch": 4.2615825237266245,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003311776312019412,
      "loss": 4.825,
      "mean_token_accuracy": 0.22631410509347916,
      "num_tokens": 84699237.0,
      "step": 37945
    },
    {
      "entropy": 5.212507677078247,
      "epoch": 4.262144100634582,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003311387531458177,
      "loss": 4.7794,
      "mean_token_accuracy": 0.23537420481443405,
      "num_tokens": 84710074.0,
      "step": 37950
    },
    {
      "entropy": 5.273351526260376,
      "epoch": 4.26270567754254,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003310998733022146,
      "loss": 4.94,
      "mean_token_accuracy": 0.2197895899415016,
      "num_tokens": 84721315.0,
      "step": 37955
    },
    {
      "entropy": 5.351147222518921,
      "epoch": 4.263267254450497,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003310609916723698,
      "loss": 4.8578,
      "mean_token_accuracy": 0.24019122123718262,
      "num_tokens": 84732654.0,
      "step": 37960
    },
    {
      "entropy": 5.268921422958374,
      "epoch": 4.263828831358454,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003310221082575213,
      "loss": 4.8966,
      "mean_token_accuracy": 0.2378625527024269,
      "num_tokens": 84745568.0,
      "step": 37965
    },
    {
      "entropy": 5.2738720893859865,
      "epoch": 4.264390408266412,
      "grad_norm": 1.09375,
      "learning_rate": 0.00033098322305890715,
      "loss": 4.8905,
      "mean_token_accuracy": 0.22741384357213973,
      "num_tokens": 84756443.0,
      "step": 37970
    },
    {
      "entropy": 5.3370178699493405,
      "epoch": 4.26495198517437,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00033094433607776545,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2380466416478157,
      "num_tokens": 84766789.0,
      "step": 37975
    },
    {
      "entropy": 5.311324739456177,
      "epoch": 4.265513562082327,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00033090544731533434,
      "loss": 4.9762,
      "mean_token_accuracy": 0.22177525311708451,
      "num_tokens": 84778186.0,
      "step": 37980
    },
    {
      "entropy": 5.2004687786102295,
      "epoch": 4.266075138990285,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003308665567728522,
      "loss": 4.7528,
      "mean_token_accuracy": 0.22973620742559434,
      "num_tokens": 84788431.0,
      "step": 37985
    },
    {
      "entropy": 5.260194730758667,
      "epoch": 4.266636715898242,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00033082766445155715,
      "loss": 4.9039,
      "mean_token_accuracy": 0.2304267942905426,
      "num_tokens": 84799881.0,
      "step": 37990
    },
    {
      "entropy": 5.284784555435181,
      "epoch": 4.2671982928062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033078877035268766,
      "loss": 4.858,
      "mean_token_accuracy": 0.2323773369193077,
      "num_tokens": 84810710.0,
      "step": 37995
    },
    {
      "entropy": 5.3125364780426025,
      "epoch": 4.267759869714157,
      "grad_norm": 1.0,
      "learning_rate": 0.00033074987447748205,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24136246740818024,
      "num_tokens": 84821932.0,
      "step": 38000
    },
    {
      "entropy": 5.1370597839355465,
      "epoch": 4.268321446622115,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00033071097682717883,
      "loss": 4.7573,
      "mean_token_accuracy": 0.23659318536520005,
      "num_tokens": 84832662.0,
      "step": 38005
    },
    {
      "entropy": 5.212203788757324,
      "epoch": 4.268883023530073,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003306720774030165,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23296640813350677,
      "num_tokens": 84843199.0,
      "step": 38010
    },
    {
      "entropy": 5.227132511138916,
      "epoch": 4.26944460043803,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003306331762062336,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23176999241113663,
      "num_tokens": 84853367.0,
      "step": 38015
    },
    {
      "entropy": 5.227924919128418,
      "epoch": 4.270006177345987,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003305942732380689,
      "loss": 4.7943,
      "mean_token_accuracy": 0.23930662125349045,
      "num_tokens": 84863527.0,
      "step": 38020
    },
    {
      "entropy": 5.115241813659668,
      "epoch": 4.270567754253945,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003305553684997609,
      "loss": 4.717,
      "mean_token_accuracy": 0.2389356642961502,
      "num_tokens": 84874929.0,
      "step": 38025
    },
    {
      "entropy": 5.211163854598999,
      "epoch": 4.2711293311619025,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00033051646199254843,
      "loss": 4.7496,
      "mean_token_accuracy": 0.2367352142930031,
      "num_tokens": 84885182.0,
      "step": 38030
    },
    {
      "entropy": 5.278507804870605,
      "epoch": 4.271690908069861,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003304775537176702,
      "loss": 4.8154,
      "mean_token_accuracy": 0.2264390170574188,
      "num_tokens": 84896115.0,
      "step": 38035
    },
    {
      "entropy": 5.268093299865723,
      "epoch": 4.272252484977818,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003304386436763652,
      "loss": 4.8512,
      "mean_token_accuracy": 0.2300826594233513,
      "num_tokens": 84908174.0,
      "step": 38040
    },
    {
      "entropy": 5.293190813064575,
      "epoch": 4.272814061885775,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003303997318698723,
      "loss": 4.8446,
      "mean_token_accuracy": 0.2254929944872856,
      "num_tokens": 84920893.0,
      "step": 38045
    },
    {
      "entropy": 5.267909669876099,
      "epoch": 4.273375638793733,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003303608182994304,
      "loss": 4.7731,
      "mean_token_accuracy": 0.2312474638223648,
      "num_tokens": 84931860.0,
      "step": 38050
    },
    {
      "entropy": 5.17956919670105,
      "epoch": 4.27393721570169,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003303219029662787,
      "loss": 4.7671,
      "mean_token_accuracy": 0.24355765879154206,
      "num_tokens": 84942533.0,
      "step": 38055
    },
    {
      "entropy": 5.123878145217896,
      "epoch": 4.274498792609648,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00033028298587165606,
      "loss": 4.6909,
      "mean_token_accuracy": 0.23397069722414016,
      "num_tokens": 84954018.0,
      "step": 38060
    },
    {
      "entropy": 5.287184524536133,
      "epoch": 4.275060369517606,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003302440670168017,
      "loss": 4.9223,
      "mean_token_accuracy": 0.22630200386047364,
      "num_tokens": 84966150.0,
      "step": 38065
    },
    {
      "entropy": 5.28118896484375,
      "epoch": 4.275621946425563,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003302051464029548,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23213020414114,
      "num_tokens": 84977668.0,
      "step": 38070
    },
    {
      "entropy": 5.199099206924439,
      "epoch": 4.27618352333352,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00033016622403135473,
      "loss": 4.7339,
      "mean_token_accuracy": 0.22994043678045273,
      "num_tokens": 84988735.0,
      "step": 38075
    },
    {
      "entropy": 5.204779100418091,
      "epoch": 4.276745100241478,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00033012729990324063,
      "loss": 4.827,
      "mean_token_accuracy": 0.22571006417274475,
      "num_tokens": 84999573.0,
      "step": 38080
    },
    {
      "entropy": 5.184571075439453,
      "epoch": 4.2773066771494355,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003300883740198519,
      "loss": 4.7555,
      "mean_token_accuracy": 0.22991671711206435,
      "num_tokens": 85010582.0,
      "step": 38085
    },
    {
      "entropy": 5.198814725875854,
      "epoch": 4.277868254057394,
      "grad_norm": 0.9375,
      "learning_rate": 0.00033004944638242805,
      "loss": 4.7205,
      "mean_token_accuracy": 0.23987737596035003,
      "num_tokens": 85021563.0,
      "step": 38090
    },
    {
      "entropy": 5.335615253448486,
      "epoch": 4.278429830965351,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003300105169922084,
      "loss": 5.0385,
      "mean_token_accuracy": 0.21889629513025283,
      "num_tokens": 85033231.0,
      "step": 38095
    },
    {
      "entropy": 5.197581577301025,
      "epoch": 4.278991407873308,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003299715858504326,
      "loss": 4.701,
      "mean_token_accuracy": 0.24005561470985412,
      "num_tokens": 85044110.0,
      "step": 38100
    },
    {
      "entropy": 5.230462312698364,
      "epoch": 4.279552984781266,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032993265295834013,
      "loss": 4.7313,
      "mean_token_accuracy": 0.2335362046957016,
      "num_tokens": 85055425.0,
      "step": 38105
    },
    {
      "entropy": 5.26796145439148,
      "epoch": 4.280114561689223,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003298937183171707,
      "loss": 4.8437,
      "mean_token_accuracy": 0.23178965747356414,
      "num_tokens": 85067358.0,
      "step": 38110
    },
    {
      "entropy": 5.270889949798584,
      "epoch": 4.2806761385971805,
      "grad_norm": 0.984375,
      "learning_rate": 0.000329854781928164,
      "loss": 4.8457,
      "mean_token_accuracy": 0.2289916232228279,
      "num_tokens": 85078967.0,
      "step": 38115
    },
    {
      "entropy": 5.197116994857788,
      "epoch": 4.281237715505139,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003298158437925598,
      "loss": 4.7089,
      "mean_token_accuracy": 0.2334579959511757,
      "num_tokens": 85088515.0,
      "step": 38120
    },
    {
      "entropy": 5.154993963241577,
      "epoch": 4.281799292413096,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003297769039115978,
      "loss": 4.7297,
      "mean_token_accuracy": 0.2346758082509041,
      "num_tokens": 85099535.0,
      "step": 38125
    },
    {
      "entropy": 5.361235237121582,
      "epoch": 4.282360869321053,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00032973796228651797,
      "loss": 4.994,
      "mean_token_accuracy": 0.21233387440443038,
      "num_tokens": 85110818.0,
      "step": 38130
    },
    {
      "entropy": 5.280814981460571,
      "epoch": 4.282922446229011,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032969901891856027,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22385382205247878,
      "num_tokens": 85120362.0,
      "step": 38135
    },
    {
      "entropy": 5.202555847167969,
      "epoch": 4.283484023136968,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032966007380896443,
      "loss": 4.7868,
      "mean_token_accuracy": 0.23564910441637038,
      "num_tokens": 85131358.0,
      "step": 38140
    },
    {
      "entropy": 5.254212951660156,
      "epoch": 4.2840456000449265,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003296211269589708,
      "loss": 4.8007,
      "mean_token_accuracy": 0.23439698070287704,
      "num_tokens": 85143395.0,
      "step": 38145
    },
    {
      "entropy": 5.225847339630127,
      "epoch": 4.284607176952884,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003295821783698192,
      "loss": 4.7373,
      "mean_token_accuracy": 0.23879900723695754,
      "num_tokens": 85155654.0,
      "step": 38150
    },
    {
      "entropy": 5.221353530883789,
      "epoch": 4.285168753860841,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032954322804274995,
      "loss": 4.9018,
      "mean_token_accuracy": 0.22230051904916764,
      "num_tokens": 85168285.0,
      "step": 38155
    },
    {
      "entropy": 5.251021385192871,
      "epoch": 4.285730330768799,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032950427597900315,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22487429529428482,
      "num_tokens": 85180058.0,
      "step": 38160
    },
    {
      "entropy": 5.2507209300994875,
      "epoch": 4.286291907676756,
      "grad_norm": 1.03125,
      "learning_rate": 0.000329465322179819,
      "loss": 4.8496,
      "mean_token_accuracy": 0.23014408051967622,
      "num_tokens": 85190990.0,
      "step": 38165
    },
    {
      "entropy": 5.352065229415894,
      "epoch": 4.2868534845847135,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032942636664643794,
      "loss": 4.8506,
      "mean_token_accuracy": 0.23461171388626098,
      "num_tokens": 85201136.0,
      "step": 38170
    },
    {
      "entropy": 5.278194618225098,
      "epoch": 4.287415061492672,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032938740938010024,
      "loss": 4.8483,
      "mean_token_accuracy": 0.22957116812467576,
      "num_tokens": 85212424.0,
      "step": 38175
    },
    {
      "entropy": 5.335921812057495,
      "epoch": 4.287976638400629,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003293484503820463,
      "loss": 4.8737,
      "mean_token_accuracy": 0.22680389434099196,
      "num_tokens": 85223767.0,
      "step": 38180
    },
    {
      "entropy": 5.32161226272583,
      "epoch": 4.288538215308587,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032930948965351664,
      "loss": 4.8579,
      "mean_token_accuracy": 0.2359195828437805,
      "num_tokens": 85234374.0,
      "step": 38185
    },
    {
      "entropy": 5.2107641220092775,
      "epoch": 4.289099792216544,
      "grad_norm": 1.21875,
      "learning_rate": 0.0003292705271957518,
      "loss": 4.757,
      "mean_token_accuracy": 0.2380441352725029,
      "num_tokens": 85245322.0,
      "step": 38190
    },
    {
      "entropy": 5.359225749969482,
      "epoch": 4.289661369124501,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032923156300999223,
      "loss": 4.9012,
      "mean_token_accuracy": 0.22152847647666932,
      "num_tokens": 85257114.0,
      "step": 38195
    },
    {
      "entropy": 5.270093488693237,
      "epoch": 4.2902229460324595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032919259709747874,
      "loss": 4.7916,
      "mean_token_accuracy": 0.23357363045215607,
      "num_tokens": 85267564.0,
      "step": 38200
    },
    {
      "entropy": 5.18270092010498,
      "epoch": 4.290784522940417,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032915362945945195,
      "loss": 4.7438,
      "mean_token_accuracy": 0.23665302395820617,
      "num_tokens": 85277763.0,
      "step": 38205
    },
    {
      "entropy": 5.282632112503052,
      "epoch": 4.291346099848374,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032911466009715254,
      "loss": 4.8976,
      "mean_token_accuracy": 0.22818298488855362,
      "num_tokens": 85288504.0,
      "step": 38210
    },
    {
      "entropy": 5.311814022064209,
      "epoch": 4.291907676756332,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00032907568901182136,
      "loss": 4.776,
      "mean_token_accuracy": 0.23949673771858215,
      "num_tokens": 85299928.0,
      "step": 38215
    },
    {
      "entropy": 5.241078567504883,
      "epoch": 4.292469253664289,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003290367162046993,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23654469847679138,
      "num_tokens": 85311052.0,
      "step": 38220
    },
    {
      "entropy": 5.266176509857178,
      "epoch": 4.293030830572247,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032899774167702723,
      "loss": 4.8659,
      "mean_token_accuracy": 0.23039778769016267,
      "num_tokens": 85322023.0,
      "step": 38225
    },
    {
      "entropy": 5.196544790267945,
      "epoch": 4.2935924074802045,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0003289587654300461,
      "loss": 4.8183,
      "mean_token_accuracy": 0.23250722289085388,
      "num_tokens": 85334160.0,
      "step": 38230
    },
    {
      "entropy": 5.324274492263794,
      "epoch": 4.294153984388162,
      "grad_norm": 1.078125,
      "learning_rate": 0.000328919787464997,
      "loss": 4.9213,
      "mean_token_accuracy": 0.2190966010093689,
      "num_tokens": 85345378.0,
      "step": 38235
    },
    {
      "entropy": 5.278559970855713,
      "epoch": 4.29471556129612,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000328880807783121,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2357930824160576,
      "num_tokens": 85357610.0,
      "step": 38240
    },
    {
      "entropy": 5.289321947097778,
      "epoch": 4.295277138204077,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032884182638565914,
      "loss": 4.7939,
      "mean_token_accuracy": 0.22592489570379257,
      "num_tokens": 85368749.0,
      "step": 38245
    },
    {
      "entropy": 5.23977952003479,
      "epoch": 4.295838715112034,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00032880284327385255,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23315060138702393,
      "num_tokens": 85379488.0,
      "step": 38250
    },
    {
      "entropy": 5.204528427124023,
      "epoch": 4.296400292019992,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003287638584489426,
      "loss": 4.7756,
      "mean_token_accuracy": 0.23286470919847488,
      "num_tokens": 85390581.0,
      "step": 38255
    },
    {
      "entropy": 5.210098123550415,
      "epoch": 4.29696186892795,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003287248719121706,
      "loss": 4.7844,
      "mean_token_accuracy": 0.22570316791534423,
      "num_tokens": 85402430.0,
      "step": 38260
    },
    {
      "entropy": 5.206792831420898,
      "epoch": 4.297523445835907,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032868588366477785,
      "loss": 4.679,
      "mean_token_accuracy": 0.23866692632436753,
      "num_tokens": 85414932.0,
      "step": 38265
    },
    {
      "entropy": 5.200941658020019,
      "epoch": 4.298085022743865,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032864689370800564,
      "loss": 4.7391,
      "mean_token_accuracy": 0.24464642256498337,
      "num_tokens": 85426484.0,
      "step": 38270
    },
    {
      "entropy": 5.177139186859131,
      "epoch": 4.298646599651822,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003286079020430957,
      "loss": 4.7769,
      "mean_token_accuracy": 0.2352145329117775,
      "num_tokens": 85436864.0,
      "step": 38275
    },
    {
      "entropy": 5.184155035018921,
      "epoch": 4.29920817655978,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032856890867128926,
      "loss": 4.7295,
      "mean_token_accuracy": 0.23656991273164749,
      "num_tokens": 85446303.0,
      "step": 38280
    },
    {
      "entropy": 5.249675607681274,
      "epoch": 4.2997697534677375,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003285299135938279,
      "loss": 4.8013,
      "mean_token_accuracy": 0.23368048369884492,
      "num_tokens": 85457218.0,
      "step": 38285
    },
    {
      "entropy": 5.251171493530274,
      "epoch": 4.300331330375695,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032849091681195347,
      "loss": 4.7732,
      "mean_token_accuracy": 0.23417459428310394,
      "num_tokens": 85469085.0,
      "step": 38290
    },
    {
      "entropy": 5.192622184753418,
      "epoch": 4.300892907283653,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032845191832690747,
      "loss": 4.75,
      "mean_token_accuracy": 0.235232113301754,
      "num_tokens": 85478966.0,
      "step": 38295
    },
    {
      "entropy": 5.178266906738282,
      "epoch": 4.30145448419161,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003284129181399317,
      "loss": 4.7451,
      "mean_token_accuracy": 0.24335668832063675,
      "num_tokens": 85489742.0,
      "step": 38300
    },
    {
      "entropy": 5.167383670806885,
      "epoch": 4.302016061099567,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003283739162522678,
      "loss": 4.6773,
      "mean_token_accuracy": 0.25153719037771227,
      "num_tokens": 85500766.0,
      "step": 38305
    },
    {
      "entropy": 5.224397611618042,
      "epoch": 4.302577638007525,
      "grad_norm": 1.0,
      "learning_rate": 0.0003283349126651578,
      "loss": 4.7376,
      "mean_token_accuracy": 0.2307155668735504,
      "num_tokens": 85512206.0,
      "step": 38310
    },
    {
      "entropy": 5.191664791107177,
      "epoch": 4.303139214915483,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00032829590737984345,
      "loss": 4.7965,
      "mean_token_accuracy": 0.23324814885854722,
      "num_tokens": 85523881.0,
      "step": 38315
    },
    {
      "entropy": 5.152264356613159,
      "epoch": 4.30370079182344,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032825690039756675,
      "loss": 4.7553,
      "mean_token_accuracy": 0.23327648937702178,
      "num_tokens": 85535130.0,
      "step": 38320
    },
    {
      "entropy": 5.285225772857666,
      "epoch": 4.304262368731398,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032821789171956966,
      "loss": 4.805,
      "mean_token_accuracy": 0.23180000782012938,
      "num_tokens": 85547261.0,
      "step": 38325
    },
    {
      "entropy": 5.290974473953247,
      "epoch": 4.304823945639355,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003281788813470943,
      "loss": 4.821,
      "mean_token_accuracy": 0.23002952933311463,
      "num_tokens": 85558300.0,
      "step": 38330
    },
    {
      "entropy": 5.1598992347717285,
      "epoch": 4.305385522547313,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003281398692813827,
      "loss": 4.7662,
      "mean_token_accuracy": 0.23458000719547273,
      "num_tokens": 85569370.0,
      "step": 38335
    },
    {
      "entropy": 5.215747356414795,
      "epoch": 4.30594709945527,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000328100855523677,
      "loss": 4.7511,
      "mean_token_accuracy": 0.24071647226810455,
      "num_tokens": 85579527.0,
      "step": 38340
    },
    {
      "entropy": 5.252534866333008,
      "epoch": 4.306508676363228,
      "grad_norm": 1.0,
      "learning_rate": 0.00032806184007521955,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23126779347658158,
      "num_tokens": 85590791.0,
      "step": 38345
    },
    {
      "entropy": 5.193552017211914,
      "epoch": 4.307070253271186,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003280228229372525,
      "loss": 4.7116,
      "mean_token_accuracy": 0.23588921427726744,
      "num_tokens": 85601210.0,
      "step": 38350
    },
    {
      "entropy": 5.221908521652222,
      "epoch": 4.307631830179143,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003279838041110182,
      "loss": 4.8116,
      "mean_token_accuracy": 0.22968484461307526,
      "num_tokens": 85611811.0,
      "step": 38355
    },
    {
      "entropy": 5.25264687538147,
      "epoch": 4.3081934070871,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032794478359775904,
      "loss": 4.9136,
      "mean_token_accuracy": 0.22155057191848754,
      "num_tokens": 85622046.0,
      "step": 38360
    },
    {
      "entropy": 5.267361068725586,
      "epoch": 4.308754983995058,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032790576139871743,
      "loss": 4.8036,
      "mean_token_accuracy": 0.23550867438316345,
      "num_tokens": 85633094.0,
      "step": 38365
    },
    {
      "entropy": 5.190475845336914,
      "epoch": 4.3093165609030155,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003278667375151359,
      "loss": 4.7209,
      "mean_token_accuracy": 0.23513114750385283,
      "num_tokens": 85644224.0,
      "step": 38370
    },
    {
      "entropy": 5.2792407989501955,
      "epoch": 4.309878137810974,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032782771194825685,
      "loss": 4.884,
      "mean_token_accuracy": 0.22655295580625534,
      "num_tokens": 85654950.0,
      "step": 38375
    },
    {
      "entropy": 5.254873418807984,
      "epoch": 4.310439714718931,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032778868469932304,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23004738241434097,
      "num_tokens": 85665250.0,
      "step": 38380
    },
    {
      "entropy": 5.311117649078369,
      "epoch": 4.311001291626888,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032774965576957705,
      "loss": 4.8873,
      "mean_token_accuracy": 0.22609642148017883,
      "num_tokens": 85676856.0,
      "step": 38385
    },
    {
      "entropy": 5.385608816146851,
      "epoch": 4.311562868534846,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003277106251602615,
      "loss": 4.9751,
      "mean_token_accuracy": 0.22285052686929702,
      "num_tokens": 85689733.0,
      "step": 38390
    },
    {
      "entropy": 5.22466926574707,
      "epoch": 4.312124445442803,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032767159287261914,
      "loss": 4.7256,
      "mean_token_accuracy": 0.2345612183213234,
      "num_tokens": 85699678.0,
      "step": 38395
    },
    {
      "entropy": 5.26018877029419,
      "epoch": 4.312686022350761,
      "grad_norm": 1.109375,
      "learning_rate": 0.000327632558907893,
      "loss": 4.9313,
      "mean_token_accuracy": 0.22704072743654252,
      "num_tokens": 85712130.0,
      "step": 38400
    },
    {
      "entropy": 5.223860502243042,
      "epoch": 4.313247599258719,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032759352326732573,
      "loss": 4.6847,
      "mean_token_accuracy": 0.2414837121963501,
      "num_tokens": 85723025.0,
      "step": 38405
    },
    {
      "entropy": 5.216822767257691,
      "epoch": 4.313809176166676,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032755448595216027,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23368801027536393,
      "num_tokens": 85733551.0,
      "step": 38410
    },
    {
      "entropy": 5.262781286239624,
      "epoch": 4.314370753074634,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032751544696363945,
      "loss": 4.8369,
      "mean_token_accuracy": 0.22762360721826552,
      "num_tokens": 85745633.0,
      "step": 38415
    },
    {
      "entropy": 5.211758756637574,
      "epoch": 4.314932329982591,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032747640630300664,
      "loss": 4.775,
      "mean_token_accuracy": 0.23622321635484694,
      "num_tokens": 85757522.0,
      "step": 38420
    },
    {
      "entropy": 5.298891687393189,
      "epoch": 4.3154939068905485,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032743736397150455,
      "loss": 4.783,
      "mean_token_accuracy": 0.2432649701833725,
      "num_tokens": 85768666.0,
      "step": 38425
    },
    {
      "entropy": 5.161177110671997,
      "epoch": 4.316055483798507,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003273983199703765,
      "loss": 4.732,
      "mean_token_accuracy": 0.23505178987979888,
      "num_tokens": 85780527.0,
      "step": 38430
    },
    {
      "entropy": 5.188716316223145,
      "epoch": 4.316617060706464,
      "grad_norm": 1.0,
      "learning_rate": 0.0003273592743008656,
      "loss": 4.7843,
      "mean_token_accuracy": 0.2336143970489502,
      "num_tokens": 85792166.0,
      "step": 38435
    },
    {
      "entropy": 5.336127901077271,
      "epoch": 4.317178637614421,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032732022696421514,
      "loss": 4.8968,
      "mean_token_accuracy": 0.22311235666275026,
      "num_tokens": 85803706.0,
      "step": 38440
    },
    {
      "entropy": 5.261284255981446,
      "epoch": 4.317740214522379,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003272811779616684,
      "loss": 4.8231,
      "mean_token_accuracy": 0.23504829853773118,
      "num_tokens": 85815515.0,
      "step": 38445
    },
    {
      "entropy": 5.247927856445313,
      "epoch": 4.318301791430336,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032724212729446864,
      "loss": 4.7385,
      "mean_token_accuracy": 0.23554009050130845,
      "num_tokens": 85826359.0,
      "step": 38450
    },
    {
      "entropy": 5.200520658493042,
      "epoch": 4.3188633683382935,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003272030749638593,
      "loss": 4.8213,
      "mean_token_accuracy": 0.2349433109164238,
      "num_tokens": 85837977.0,
      "step": 38455
    },
    {
      "entropy": 5.28941912651062,
      "epoch": 4.319424945246252,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003271640209710837,
      "loss": 4.873,
      "mean_token_accuracy": 0.2257858470082283,
      "num_tokens": 85850356.0,
      "step": 38460
    },
    {
      "entropy": 5.271939134597778,
      "epoch": 4.319986522154209,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003271249653173855,
      "loss": 4.8468,
      "mean_token_accuracy": 0.22777175903320312,
      "num_tokens": 85863101.0,
      "step": 38465
    },
    {
      "entropy": 5.286149740219116,
      "epoch": 4.320548099062167,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032708590800400826,
      "loss": 4.7898,
      "mean_token_accuracy": 0.23326119780540466,
      "num_tokens": 85873374.0,
      "step": 38470
    },
    {
      "entropy": 5.301827669143677,
      "epoch": 4.321109675970124,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003270468490321955,
      "loss": 4.8942,
      "mean_token_accuracy": 0.22895084023475648,
      "num_tokens": 85886167.0,
      "step": 38475
    },
    {
      "entropy": 5.232157230377197,
      "epoch": 4.321671252878081,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032700778840319086,
      "loss": 4.7951,
      "mean_token_accuracy": 0.22944723069667816,
      "num_tokens": 85897040.0,
      "step": 38480
    },
    {
      "entropy": 5.210456037521363,
      "epoch": 4.3222328297860395,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000326968726118238,
      "loss": 4.7678,
      "mean_token_accuracy": 0.22928351908922195,
      "num_tokens": 85908386.0,
      "step": 38485
    },
    {
      "entropy": 5.202938222885132,
      "epoch": 4.322794406693997,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032692966217858074,
      "loss": 4.7815,
      "mean_token_accuracy": 0.24659544229507446,
      "num_tokens": 85920313.0,
      "step": 38490
    },
    {
      "entropy": 5.224696493148803,
      "epoch": 4.323355983601954,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003268905965854629,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2318374916911125,
      "num_tokens": 85931273.0,
      "step": 38495
    },
    {
      "entropy": 5.2830384254455565,
      "epoch": 4.323917560509912,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032685152934012844,
      "loss": 4.8722,
      "mean_token_accuracy": 0.2238446518778801,
      "num_tokens": 85942952.0,
      "step": 38500
    },
    {
      "entropy": 5.203948879241944,
      "epoch": 4.324479137417869,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003268124604438211,
      "loss": 4.7483,
      "mean_token_accuracy": 0.23163330405950547,
      "num_tokens": 85953331.0,
      "step": 38505
    },
    {
      "entropy": 5.257327556610107,
      "epoch": 4.3250407143258265,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032677338989778494,
      "loss": 4.8094,
      "mean_token_accuracy": 0.2374363660812378,
      "num_tokens": 85965142.0,
      "step": 38510
    },
    {
      "entropy": 5.2643616676330565,
      "epoch": 4.325602291233785,
      "grad_norm": 1.078125,
      "learning_rate": 0.000326734317703264,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23782408833503724,
      "num_tokens": 85976258.0,
      "step": 38515
    },
    {
      "entropy": 5.213579511642456,
      "epoch": 4.326163868141742,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003266952438615023,
      "loss": 4.8238,
      "mean_token_accuracy": 0.2341081127524376,
      "num_tokens": 85987976.0,
      "step": 38520
    },
    {
      "entropy": 5.148120260238647,
      "epoch": 4.3267254450497,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032665616837374397,
      "loss": 4.6962,
      "mean_token_accuracy": 0.23198213428258896,
      "num_tokens": 85999283.0,
      "step": 38525
    },
    {
      "entropy": 5.1640400886535645,
      "epoch": 4.327287021957657,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003266170912412332,
      "loss": 4.7495,
      "mean_token_accuracy": 0.238247749209404,
      "num_tokens": 86010484.0,
      "step": 38530
    },
    {
      "entropy": 5.264367008209229,
      "epoch": 4.327848598865614,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003265780124652143,
      "loss": 4.864,
      "mean_token_accuracy": 0.22818603515625,
      "num_tokens": 86021582.0,
      "step": 38535
    },
    {
      "entropy": 5.1925966262817385,
      "epoch": 4.3284101757735725,
      "grad_norm": 0.90234375,
      "learning_rate": 0.00032653893204693136,
      "loss": 4.7794,
      "mean_token_accuracy": 0.22743798643350602,
      "num_tokens": 86033847.0,
      "step": 38540
    },
    {
      "entropy": 5.2356085777282715,
      "epoch": 4.32897175268153,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003264998499876289,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23327402174472808,
      "num_tokens": 86044870.0,
      "step": 38545
    },
    {
      "entropy": 5.174020099639892,
      "epoch": 4.329533329589487,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003264607662885512,
      "loss": 4.6693,
      "mean_token_accuracy": 0.23875287622213365,
      "num_tokens": 86055807.0,
      "step": 38550
    },
    {
      "entropy": 5.248003911972046,
      "epoch": 4.330094906497445,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00032642168095094274,
      "loss": 4.8908,
      "mean_token_accuracy": 0.22686259299516678,
      "num_tokens": 86066869.0,
      "step": 38555
    },
    {
      "entropy": 5.291968870162964,
      "epoch": 4.330656483405402,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003263825939760481,
      "loss": 4.8829,
      "mean_token_accuracy": 0.22762629389762878,
      "num_tokens": 86078870.0,
      "step": 38560
    },
    {
      "entropy": 5.212283992767334,
      "epoch": 4.33121806031336,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032634350536511167,
      "loss": 4.8442,
      "mean_token_accuracy": 0.22568914890289307,
      "num_tokens": 86090940.0,
      "step": 38565
    },
    {
      "entropy": 5.1882415294647215,
      "epoch": 4.3317796372213175,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00032630441511937813,
      "loss": 4.7152,
      "mean_token_accuracy": 0.23573580235242844,
      "num_tokens": 86101171.0,
      "step": 38570
    },
    {
      "entropy": 5.1559347152709964,
      "epoch": 4.332341214129275,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032626532324009204,
      "loss": 4.7567,
      "mean_token_accuracy": 0.22601438462734222,
      "num_tokens": 86111614.0,
      "step": 38575
    },
    {
      "entropy": 5.207160568237304,
      "epoch": 4.332902791037233,
      "grad_norm": 1.0,
      "learning_rate": 0.0003262262297284983,
      "loss": 4.7905,
      "mean_token_accuracy": 0.2411384865641594,
      "num_tokens": 86124131.0,
      "step": 38580
    },
    {
      "entropy": 5.29119348526001,
      "epoch": 4.33346436794519,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032618713458584147,
      "loss": 4.8063,
      "mean_token_accuracy": 0.2319396749138832,
      "num_tokens": 86135584.0,
      "step": 38585
    },
    {
      "entropy": 5.214950275421143,
      "epoch": 4.334025944853147,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003261480378133665,
      "loss": 4.6868,
      "mean_token_accuracy": 0.24149044007062911,
      "num_tokens": 86146163.0,
      "step": 38590
    },
    {
      "entropy": 5.207509708404541,
      "epoch": 4.334587521761105,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003261089394123181,
      "loss": 4.8617,
      "mean_token_accuracy": 0.2291869655251503,
      "num_tokens": 86157199.0,
      "step": 38595
    },
    {
      "entropy": 5.222863578796387,
      "epoch": 4.335149098669063,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003260698393839413,
      "loss": 4.7605,
      "mean_token_accuracy": 0.23419930338859557,
      "num_tokens": 86167523.0,
      "step": 38600
    },
    {
      "entropy": 5.242627859115601,
      "epoch": 4.335710675577021,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032603073772948096,
      "loss": 4.757,
      "mean_token_accuracy": 0.23626542836427689,
      "num_tokens": 86179341.0,
      "step": 38605
    },
    {
      "entropy": 5.223317861557007,
      "epoch": 4.336272252484978,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00032599163445018204,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23048143535852433,
      "num_tokens": 86190154.0,
      "step": 38610
    },
    {
      "entropy": 5.184272909164429,
      "epoch": 4.336833829392935,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003259525295472899,
      "loss": 4.7914,
      "mean_token_accuracy": 0.23691242486238479,
      "num_tokens": 86201598.0,
      "step": 38615
    },
    {
      "entropy": 5.242425727844238,
      "epoch": 4.337395406300893,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032591342302204935,
      "loss": 4.7598,
      "mean_token_accuracy": 0.2339010179042816,
      "num_tokens": 86212033.0,
      "step": 38620
    },
    {
      "entropy": 5.224689960479736,
      "epoch": 4.3379569832088505,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003258743148757057,
      "loss": 4.8278,
      "mean_token_accuracy": 0.2244638353586197,
      "num_tokens": 86222931.0,
      "step": 38625
    },
    {
      "entropy": 5.192465496063233,
      "epoch": 4.338518560116808,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032583520510950405,
      "loss": 4.7413,
      "mean_token_accuracy": 0.23213358521461486,
      "num_tokens": 86234022.0,
      "step": 38630
    },
    {
      "entropy": 5.157550859451294,
      "epoch": 4.339080137024766,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032579609372468984,
      "loss": 4.7205,
      "mean_token_accuracy": 0.23559615314006804,
      "num_tokens": 86244267.0,
      "step": 38635
    },
    {
      "entropy": 5.204600477218628,
      "epoch": 4.339641713932723,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032575698072250827,
      "loss": 4.7494,
      "mean_token_accuracy": 0.24036502540111543,
      "num_tokens": 86254200.0,
      "step": 38640
    },
    {
      "entropy": 5.250797462463379,
      "epoch": 4.34020329084068,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003257178661042047,
      "loss": 4.8397,
      "mean_token_accuracy": 0.2296455204486847,
      "num_tokens": 86266523.0,
      "step": 38645
    },
    {
      "entropy": 5.313766765594482,
      "epoch": 4.340764867748638,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032567874987102463,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2267376109957695,
      "num_tokens": 86276949.0,
      "step": 38650
    },
    {
      "entropy": 5.225643491744995,
      "epoch": 4.341326444656596,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003256396320242136,
      "loss": 4.8029,
      "mean_token_accuracy": 0.2341684654355049,
      "num_tokens": 86287412.0,
      "step": 38655
    },
    {
      "entropy": 5.243691062927246,
      "epoch": 4.341888021564554,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000325600512565017,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23862721920013427,
      "num_tokens": 86298540.0,
      "step": 38660
    },
    {
      "entropy": 5.142025709152222,
      "epoch": 4.342449598472511,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032556139149468046,
      "loss": 4.7362,
      "mean_token_accuracy": 0.24279806315898894,
      "num_tokens": 86309351.0,
      "step": 38665
    },
    {
      "entropy": 5.303529596328735,
      "epoch": 4.343011175380468,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032552226881444956,
      "loss": 4.8568,
      "mean_token_accuracy": 0.22948520332574845,
      "num_tokens": 86320873.0,
      "step": 38670
    },
    {
      "entropy": 5.292119455337525,
      "epoch": 4.343572752288426,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00032548314452557006,
      "loss": 4.8624,
      "mean_token_accuracy": 0.2263875886797905,
      "num_tokens": 86332889.0,
      "step": 38675
    },
    {
      "entropy": 5.171069574356079,
      "epoch": 4.344134329196383,
      "grad_norm": 1.109375,
      "learning_rate": 0.00032544401862928766,
      "loss": 4.6715,
      "mean_token_accuracy": 0.23870826363563538,
      "num_tokens": 86344155.0,
      "step": 38680
    },
    {
      "entropy": 5.197946500778198,
      "epoch": 4.344695906104341,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032540489112684804,
      "loss": 4.7704,
      "mean_token_accuracy": 0.239088337123394,
      "num_tokens": 86355718.0,
      "step": 38685
    },
    {
      "entropy": 5.269540643692016,
      "epoch": 4.345257483012299,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032536576201949724,
      "loss": 4.8952,
      "mean_token_accuracy": 0.22086803913116454,
      "num_tokens": 86367174.0,
      "step": 38690
    },
    {
      "entropy": 5.264174127578736,
      "epoch": 4.345819059920256,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003253266313084811,
      "loss": 4.8514,
      "mean_token_accuracy": 0.2269853189587593,
      "num_tokens": 86379086.0,
      "step": 38695
    },
    {
      "entropy": 5.321811103820801,
      "epoch": 4.346380636828213,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003252874989950454,
      "loss": 4.8679,
      "mean_token_accuracy": 0.23107229471206664,
      "num_tokens": 86390113.0,
      "step": 38700
    },
    {
      "entropy": 5.316815376281738,
      "epoch": 4.346942213736171,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032524836508043625,
      "loss": 4.9054,
      "mean_token_accuracy": 0.2246581181883812,
      "num_tokens": 86400824.0,
      "step": 38705
    },
    {
      "entropy": 5.217395639419555,
      "epoch": 4.3475037906441285,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032520922956589973,
      "loss": 4.8056,
      "mean_token_accuracy": 0.23019414097070695,
      "num_tokens": 86412972.0,
      "step": 38710
    },
    {
      "entropy": 5.271756219863891,
      "epoch": 4.348065367552087,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032517009245268175,
      "loss": 4.7579,
      "mean_token_accuracy": 0.2335440397262573,
      "num_tokens": 86425332.0,
      "step": 38715
    },
    {
      "entropy": 5.192366123199463,
      "epoch": 4.348626944460044,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032513095374202863,
      "loss": 4.7188,
      "mean_token_accuracy": 0.2381754234433174,
      "num_tokens": 86435615.0,
      "step": 38720
    },
    {
      "entropy": 5.25582857131958,
      "epoch": 4.349188521368001,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032509181343518646,
      "loss": 4.8998,
      "mean_token_accuracy": 0.23127583414316177,
      "num_tokens": 86446875.0,
      "step": 38725
    },
    {
      "entropy": 5.216358089447022,
      "epoch": 4.349750098275959,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00032505267153340163,
      "loss": 4.7542,
      "mean_token_accuracy": 0.23994012027978898,
      "num_tokens": 86458284.0,
      "step": 38730
    },
    {
      "entropy": 5.318609762191772,
      "epoch": 4.350311675183916,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003250135280379202,
      "loss": 4.9059,
      "mean_token_accuracy": 0.21640936136245728,
      "num_tokens": 86470977.0,
      "step": 38735
    },
    {
      "entropy": 5.285028553009033,
      "epoch": 4.350873252091874,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003249743829499888,
      "loss": 4.8772,
      "mean_token_accuracy": 0.22680825740098953,
      "num_tokens": 86482578.0,
      "step": 38740
    },
    {
      "entropy": 5.206098461151123,
      "epoch": 4.351434828999832,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003249352362708535,
      "loss": 4.7084,
      "mean_token_accuracy": 0.23777247369289398,
      "num_tokens": 86493123.0,
      "step": 38745
    },
    {
      "entropy": 5.329891157150269,
      "epoch": 4.351996405907789,
      "grad_norm": 1.0625,
      "learning_rate": 0.000324896088001761,
      "loss": 4.9605,
      "mean_token_accuracy": 0.22683517187833785,
      "num_tokens": 86503865.0,
      "step": 38750
    },
    {
      "entropy": 5.208509302139282,
      "epoch": 4.352557982815747,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003248569381439577,
      "loss": 4.8009,
      "mean_token_accuracy": 0.23501321226358413,
      "num_tokens": 86515006.0,
      "step": 38755
    },
    {
      "entropy": 5.211210155487061,
      "epoch": 4.353119559723704,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003248177866986901,
      "loss": 4.7825,
      "mean_token_accuracy": 0.23012216836214067,
      "num_tokens": 86526256.0,
      "step": 38760
    },
    {
      "entropy": 5.246545648574829,
      "epoch": 4.3536811366316615,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003247786336672049,
      "loss": 4.7212,
      "mean_token_accuracy": 0.24281422048807144,
      "num_tokens": 86537050.0,
      "step": 38765
    },
    {
      "entropy": 5.181524276733398,
      "epoch": 4.35424271353962,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032473947905074866,
      "loss": 4.731,
      "mean_token_accuracy": 0.24190652072429658,
      "num_tokens": 86547678.0,
      "step": 38770
    },
    {
      "entropy": 5.154521512985229,
      "epoch": 4.354804290447577,
      "grad_norm": 1.0,
      "learning_rate": 0.00032470032285056814,
      "loss": 4.7315,
      "mean_token_accuracy": 0.23849258124828338,
      "num_tokens": 86558747.0,
      "step": 38775
    },
    {
      "entropy": 5.187523269653321,
      "epoch": 4.355365867355534,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032466116506791004,
      "loss": 4.817,
      "mean_token_accuracy": 0.2313641294836998,
      "num_tokens": 86569500.0,
      "step": 38780
    },
    {
      "entropy": 5.179836463928223,
      "epoch": 4.355927444263492,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032462200570402124,
      "loss": 4.8222,
      "mean_token_accuracy": 0.22834180891513825,
      "num_tokens": 86581312.0,
      "step": 38785
    },
    {
      "entropy": 5.231000232696533,
      "epoch": 4.356489021171449,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003245828447601484,
      "loss": 4.8412,
      "mean_token_accuracy": 0.23264707773923873,
      "num_tokens": 86592526.0,
      "step": 38790
    },
    {
      "entropy": 5.265898561477661,
      "epoch": 4.357050598079407,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003245436822375387,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22843435108661653,
      "num_tokens": 86604284.0,
      "step": 38795
    },
    {
      "entropy": 5.210625553131104,
      "epoch": 4.357612174987365,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003245045181374389,
      "loss": 4.752,
      "mean_token_accuracy": 0.2378621593117714,
      "num_tokens": 86615567.0,
      "step": 38800
    },
    {
      "entropy": 5.119342470169068,
      "epoch": 4.358173751895322,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000324465352461096,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2434703841805458,
      "num_tokens": 86626188.0,
      "step": 38805
    },
    {
      "entropy": 5.267591428756714,
      "epoch": 4.35873532880328,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003244261852097572,
      "loss": 4.8662,
      "mean_token_accuracy": 0.22800345122814178,
      "num_tokens": 86638468.0,
      "step": 38810
    },
    {
      "entropy": 5.331792783737183,
      "epoch": 4.359296905711237,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00032438701638466943,
      "loss": 4.8223,
      "mean_token_accuracy": 0.2373081475496292,
      "num_tokens": 86649119.0,
      "step": 38815
    },
    {
      "entropy": 5.2625397682189945,
      "epoch": 4.359858482619194,
      "grad_norm": 1.0,
      "learning_rate": 0.00032434784598707987,
      "loss": 4.802,
      "mean_token_accuracy": 0.23408755511045456,
      "num_tokens": 86660924.0,
      "step": 38820
    },
    {
      "entropy": 5.233278512954712,
      "epoch": 4.3604200595271525,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003243086740182358,
      "loss": 4.8182,
      "mean_token_accuracy": 0.2359543189406395,
      "num_tokens": 86672655.0,
      "step": 38825
    },
    {
      "entropy": 5.202617073059082,
      "epoch": 4.36098163643511,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032426950047938436,
      "loss": 4.7345,
      "mean_token_accuracy": 0.2397816002368927,
      "num_tokens": 86684146.0,
      "step": 38830
    },
    {
      "entropy": 5.313151121139526,
      "epoch": 4.361543213343067,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000324230325371773,
      "loss": 4.9547,
      "mean_token_accuracy": 0.22334452271461486,
      "num_tokens": 86695435.0,
      "step": 38835
    },
    {
      "entropy": 5.4362293720245365,
      "epoch": 4.362104790251025,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003241911486966489,
      "loss": 4.9807,
      "mean_token_accuracy": 0.2223629340529442,
      "num_tokens": 86706387.0,
      "step": 38840
    },
    {
      "entropy": 5.254540300369262,
      "epoch": 4.362666367158982,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003241519704552596,
      "loss": 4.7249,
      "mean_token_accuracy": 0.23576538860797883,
      "num_tokens": 86717661.0,
      "step": 38845
    },
    {
      "entropy": 5.196700239181519,
      "epoch": 4.36322794406694,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032411279064885253,
      "loss": 4.7793,
      "mean_token_accuracy": 0.2299754425883293,
      "num_tokens": 86729045.0,
      "step": 38850
    },
    {
      "entropy": 5.224139738082886,
      "epoch": 4.363789520974898,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032407360927867506,
      "loss": 4.9369,
      "mean_token_accuracy": 0.22506697177886964,
      "num_tokens": 86740763.0,
      "step": 38855
    },
    {
      "entropy": 5.26365852355957,
      "epoch": 4.364351097882855,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003240344263459749,
      "loss": 4.8534,
      "mean_token_accuracy": 0.23309439420700073,
      "num_tokens": 86752369.0,
      "step": 38860
    },
    {
      "entropy": 5.2478855609893795,
      "epoch": 4.364912674790813,
      "grad_norm": 1.0,
      "learning_rate": 0.00032399524185199957,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22896899580955504,
      "num_tokens": 86765947.0,
      "step": 38865
    },
    {
      "entropy": 5.183007669448853,
      "epoch": 4.36547425169877,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032395605579799673,
      "loss": 4.6671,
      "mean_token_accuracy": 0.24128337800502778,
      "num_tokens": 86777202.0,
      "step": 38870
    },
    {
      "entropy": 5.228087043762207,
      "epoch": 4.366035828606727,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032391686818521414,
      "loss": 4.7794,
      "mean_token_accuracy": 0.22888812869787217,
      "num_tokens": 86787838.0,
      "step": 38875
    },
    {
      "entropy": 5.305694103240967,
      "epoch": 4.3665974055146854,
      "grad_norm": 0.921875,
      "learning_rate": 0.00032387767901489946,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2281346797943115,
      "num_tokens": 86799875.0,
      "step": 38880
    },
    {
      "entropy": 5.265244054794311,
      "epoch": 4.367158982422643,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032383848828830054,
      "loss": 4.8244,
      "mean_token_accuracy": 0.2308244749903679,
      "num_tokens": 86811018.0,
      "step": 38885
    },
    {
      "entropy": 5.195840072631836,
      "epoch": 4.3677205593306,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00032379929600666525,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2365916520357132,
      "num_tokens": 86821556.0,
      "step": 38890
    },
    {
      "entropy": 5.090424299240112,
      "epoch": 4.368282136238558,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032376010217124133,
      "loss": 4.6491,
      "mean_token_accuracy": 0.24863025546073914,
      "num_tokens": 86832041.0,
      "step": 38895
    },
    {
      "entropy": 5.199869012832641,
      "epoch": 4.368843713146515,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032372090678327694,
      "loss": 4.7608,
      "mean_token_accuracy": 0.23242097645998,
      "num_tokens": 86844723.0,
      "step": 38900
    },
    {
      "entropy": 5.284578657150268,
      "epoch": 4.369405290054473,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032368170984401997,
      "loss": 4.8152,
      "mean_token_accuracy": 0.23326994627714157,
      "num_tokens": 86856211.0,
      "step": 38905
    },
    {
      "entropy": 5.256839084625244,
      "epoch": 4.3699668669624305,
      "grad_norm": 1.09375,
      "learning_rate": 0.00032364251135471846,
      "loss": 4.8313,
      "mean_token_accuracy": 0.2269367128610611,
      "num_tokens": 86867701.0,
      "step": 38910
    },
    {
      "entropy": 5.319783782958984,
      "epoch": 4.370528443870388,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003236033113166205,
      "loss": 4.9275,
      "mean_token_accuracy": 0.2260647103190422,
      "num_tokens": 86879838.0,
      "step": 38915
    },
    {
      "entropy": 5.298369741439819,
      "epoch": 4.371090020778346,
      "grad_norm": 0.96875,
      "learning_rate": 0.00032356410973097423,
      "loss": 4.885,
      "mean_token_accuracy": 0.23200421631336213,
      "num_tokens": 86891101.0,
      "step": 38920
    },
    {
      "entropy": 5.270803785324096,
      "epoch": 4.371651597686303,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00032352490659902795,
      "loss": 4.7834,
      "mean_token_accuracy": 0.23995203226804734,
      "num_tokens": 86902796.0,
      "step": 38925
    },
    {
      "entropy": 5.2443084716796875,
      "epoch": 4.37221317459426,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003234857019220298,
      "loss": 4.8497,
      "mean_token_accuracy": 0.22981565445661545,
      "num_tokens": 86914381.0,
      "step": 38930
    },
    {
      "entropy": 5.300256729125977,
      "epoch": 4.372774751502218,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000323446495701228,
      "loss": 4.9599,
      "mean_token_accuracy": 0.22768183499574662,
      "num_tokens": 86926082.0,
      "step": 38935
    },
    {
      "entropy": 5.240850019454956,
      "epoch": 4.373336328410176,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003234072879378711,
      "loss": 4.7514,
      "mean_token_accuracy": 0.2401571527123451,
      "num_tokens": 86936644.0,
      "step": 38940
    },
    {
      "entropy": 5.263034820556641,
      "epoch": 4.373897905318134,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003233680786332074,
      "loss": 4.8478,
      "mean_token_accuracy": 0.22783418744802475,
      "num_tokens": 86947586.0,
      "step": 38945
    },
    {
      "entropy": 5.297974109649658,
      "epoch": 4.374459482226091,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003233288677884853,
      "loss": 4.8127,
      "mean_token_accuracy": 0.2323236033320427,
      "num_tokens": 86958265.0,
      "step": 38950
    },
    {
      "entropy": 5.256190586090088,
      "epoch": 4.375021059134048,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003232896554049533,
      "loss": 4.7733,
      "mean_token_accuracy": 0.2337459772825241,
      "num_tokens": 86969119.0,
      "step": 38955
    },
    {
      "entropy": 5.204686212539673,
      "epoch": 4.375582636042006,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032325044148385986,
      "loss": 4.8195,
      "mean_token_accuracy": 0.22757731080055238,
      "num_tokens": 86980232.0,
      "step": 38960
    },
    {
      "entropy": 5.214489459991455,
      "epoch": 4.3761442129499635,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032321122602645374,
      "loss": 4.7171,
      "mean_token_accuracy": 0.2327033206820488,
      "num_tokens": 86993590.0,
      "step": 38965
    },
    {
      "entropy": 5.262153959274292,
      "epoch": 4.376705789857921,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003231720090339834,
      "loss": 4.83,
      "mean_token_accuracy": 0.23017940074205398,
      "num_tokens": 87004905.0,
      "step": 38970
    },
    {
      "entropy": 5.207371520996094,
      "epoch": 4.377267366765879,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032313279050769767,
      "loss": 4.7985,
      "mean_token_accuracy": 0.23729910403490068,
      "num_tokens": 87016920.0,
      "step": 38975
    },
    {
      "entropy": 5.314700269699097,
      "epoch": 4.377828943673836,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003230935704488452,
      "loss": 4.938,
      "mean_token_accuracy": 0.22305944859981536,
      "num_tokens": 87027740.0,
      "step": 38980
    },
    {
      "entropy": 5.326733589172363,
      "epoch": 4.378390520581794,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032305434885867483,
      "loss": 4.8093,
      "mean_token_accuracy": 0.2269554167985916,
      "num_tokens": 87039742.0,
      "step": 38985
    },
    {
      "entropy": 5.191789293289185,
      "epoch": 4.378952097489751,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00032301512573843535,
      "loss": 4.7423,
      "mean_token_accuracy": 0.2299266129732132,
      "num_tokens": 87049750.0,
      "step": 38990
    },
    {
      "entropy": 5.250551795959472,
      "epoch": 4.3795136743977086,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003229759010893756,
      "loss": 4.8348,
      "mean_token_accuracy": 0.2394724443554878,
      "num_tokens": 87062890.0,
      "step": 38995
    },
    {
      "entropy": 5.249023008346557,
      "epoch": 4.380075251305667,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003229366749127446,
      "loss": 4.8302,
      "mean_token_accuracy": 0.23092907816171646,
      "num_tokens": 87074450.0,
      "step": 39000
    },
    {
      "epoch": 4.380075251305667,
      "eval_entropy": 5.085296160786736,
      "eval_loss": 5.000723838806152,
      "eval_mean_token_accuracy": 0.2292825282580895,
      "eval_num_tokens": 87074450.0,
      "eval_runtime": 24.1665,
      "eval_samples_per_second": 1283.307,
      "eval_steps_per_second": 160.429,
      "step": 39000
    },
    {
      "entropy": 5.359755516052246,
      "epoch": 4.380636828213624,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003228974472097913,
      "loss": 4.9253,
      "mean_token_accuracy": 0.22619855850934983,
      "num_tokens": 87085931.0,
      "step": 39005
    },
    {
      "entropy": 5.264909458160401,
      "epoch": 4.381198405121581,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032285821798176456,
      "loss": 4.8835,
      "mean_token_accuracy": 0.22880505323410033,
      "num_tokens": 87096655.0,
      "step": 39010
    },
    {
      "entropy": 5.258397912979126,
      "epoch": 4.381759982029539,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032281898722991376,
      "loss": 4.8732,
      "mean_token_accuracy": 0.22373248934745787,
      "num_tokens": 87108359.0,
      "step": 39015
    },
    {
      "entropy": 5.301404523849487,
      "epoch": 4.382321558937496,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00032277975495548775,
      "loss": 4.9114,
      "mean_token_accuracy": 0.2279702603816986,
      "num_tokens": 87119028.0,
      "step": 39020
    },
    {
      "entropy": 5.191596603393554,
      "epoch": 4.382883135845454,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003227405211597359,
      "loss": 4.7344,
      "mean_token_accuracy": 0.2372366890311241,
      "num_tokens": 87130574.0,
      "step": 39025
    },
    {
      "entropy": 5.273971033096314,
      "epoch": 4.383444712753412,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032270128584390735,
      "loss": 4.7759,
      "mean_token_accuracy": 0.23112665712833405,
      "num_tokens": 87141052.0,
      "step": 39030
    },
    {
      "entropy": 5.280712890625,
      "epoch": 4.384006289661369,
      "grad_norm": 1.0,
      "learning_rate": 0.0003226620490092513,
      "loss": 4.859,
      "mean_token_accuracy": 0.22761166244745254,
      "num_tokens": 87151323.0,
      "step": 39035
    },
    {
      "entropy": 5.18376841545105,
      "epoch": 4.384567866569327,
      "grad_norm": 0.984375,
      "learning_rate": 0.00032262281065701716,
      "loss": 4.7256,
      "mean_token_accuracy": 0.22574040591716765,
      "num_tokens": 87162614.0,
      "step": 39040
    },
    {
      "entropy": 5.176966810226441,
      "epoch": 4.385129443477284,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003225835707884542,
      "loss": 4.7293,
      "mean_token_accuracy": 0.23279133141040803,
      "num_tokens": 87172712.0,
      "step": 39045
    },
    {
      "entropy": 5.26340684890747,
      "epoch": 4.3856910203852415,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003225443294048119,
      "loss": 4.8049,
      "mean_token_accuracy": 0.23518168926239014,
      "num_tokens": 87183612.0,
      "step": 39050
    },
    {
      "entropy": 5.255292320251465,
      "epoch": 4.3862525972932,
      "grad_norm": 1.046875,
      "learning_rate": 0.00032250508650733976,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23256277590990065,
      "num_tokens": 87194707.0,
      "step": 39055
    },
    {
      "entropy": 5.164405632019043,
      "epoch": 4.386814174201157,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003224658420972871,
      "loss": 4.7193,
      "mean_token_accuracy": 0.24051294028759002,
      "num_tokens": 87206451.0,
      "step": 39060
    },
    {
      "entropy": 5.270030736923218,
      "epoch": 4.387375751109114,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003224265961759038,
      "loss": 4.9334,
      "mean_token_accuracy": 0.22319797724485396,
      "num_tokens": 87217542.0,
      "step": 39065
    },
    {
      "entropy": 5.2145562171936035,
      "epoch": 4.387937328017072,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003223873487444392,
      "loss": 4.7394,
      "mean_token_accuracy": 0.2403210163116455,
      "num_tokens": 87228399.0,
      "step": 39070
    },
    {
      "entropy": 5.252208566665649,
      "epoch": 4.388498904925029,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003223480998041431,
      "loss": 4.8349,
      "mean_token_accuracy": 0.23008241057395934,
      "num_tokens": 87239875.0,
      "step": 39075
    },
    {
      "entropy": 5.291070938110352,
      "epoch": 4.389060481832987,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032230884935626505,
      "loss": 4.8588,
      "mean_token_accuracy": 0.22812968492507935,
      "num_tokens": 87252919.0,
      "step": 39080
    },
    {
      "entropy": 5.187070083618164,
      "epoch": 4.389622058740945,
      "grad_norm": 0.97265625,
      "learning_rate": 0.000322269597402055,
      "loss": 4.7677,
      "mean_token_accuracy": 0.23661463409662248,
      "num_tokens": 87264458.0,
      "step": 39085
    },
    {
      "entropy": 5.269531774520874,
      "epoch": 4.390183635648902,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003222303439427625,
      "loss": 4.8179,
      "mean_token_accuracy": 0.23958767354488372,
      "num_tokens": 87274935.0,
      "step": 39090
    },
    {
      "entropy": 5.178520393371582,
      "epoch": 4.39074521255686,
      "grad_norm": 1.015625,
      "learning_rate": 0.00032219108897963767,
      "loss": 4.7598,
      "mean_token_accuracy": 0.23790866732597352,
      "num_tokens": 87285870.0,
      "step": 39095
    },
    {
      "entropy": 5.208259105682373,
      "epoch": 4.391306789464817,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003221518325139302,
      "loss": 4.7413,
      "mean_token_accuracy": 0.23201300352811813,
      "num_tokens": 87297374.0,
      "step": 39100
    },
    {
      "entropy": 5.2874753952026365,
      "epoch": 4.3918683663727744,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003221125745468901,
      "loss": 4.8056,
      "mean_token_accuracy": 0.233444182574749,
      "num_tokens": 87307597.0,
      "step": 39105
    },
    {
      "entropy": 5.146993970870971,
      "epoch": 4.3924299432807326,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003220733150797673,
      "loss": 4.7074,
      "mean_token_accuracy": 0.2366822600364685,
      "num_tokens": 87318459.0,
      "step": 39110
    },
    {
      "entropy": 5.333412837982178,
      "epoch": 4.39299152018869,
      "grad_norm": 0.87109375,
      "learning_rate": 0.0003220340541138119,
      "loss": 4.9637,
      "mean_token_accuracy": 0.22280818670988084,
      "num_tokens": 87330613.0,
      "step": 39115
    },
    {
      "entropy": 5.3236133575439455,
      "epoch": 4.393553097096647,
      "grad_norm": 0.953125,
      "learning_rate": 0.00032199479165027407,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22842433899641038,
      "num_tokens": 87341988.0,
      "step": 39120
    },
    {
      "entropy": 5.229107809066773,
      "epoch": 4.394114674004605,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032195552769040376,
      "loss": 4.748,
      "mean_token_accuracy": 0.2304905906319618,
      "num_tokens": 87353532.0,
      "step": 39125
    },
    {
      "entropy": 5.2743854999542235,
      "epoch": 4.394676250912562,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003219162622354513,
      "loss": 4.8422,
      "mean_token_accuracy": 0.2379297733306885,
      "num_tokens": 87364754.0,
      "step": 39130
    },
    {
      "entropy": 5.209136867523194,
      "epoch": 4.39523782782052,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00032187699528666674,
      "loss": 4.7883,
      "mean_token_accuracy": 0.22802451848983765,
      "num_tokens": 87375983.0,
      "step": 39135
    },
    {
      "entropy": 5.215414428710938,
      "epoch": 4.395799404728478,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003218377268453006,
      "loss": 4.7862,
      "mean_token_accuracy": 0.23883452266454697,
      "num_tokens": 87388169.0,
      "step": 39140
    },
    {
      "entropy": 5.3036659240722654,
      "epoch": 4.396360981636435,
      "grad_norm": 1.046875,
      "learning_rate": 0.000321798456912603,
      "loss": 4.85,
      "mean_token_accuracy": 0.2329813838005066,
      "num_tokens": 87399341.0,
      "step": 39145
    },
    {
      "entropy": 5.2165967464447025,
      "epoch": 4.396922558544393,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032175918548982437,
      "loss": 4.7814,
      "mean_token_accuracy": 0.23087027221918105,
      "num_tokens": 87410941.0,
      "step": 39150
    },
    {
      "entropy": 5.202066421508789,
      "epoch": 4.39748413545235,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00032171991257821513,
      "loss": 4.7555,
      "mean_token_accuracy": 0.2351413205265999,
      "num_tokens": 87421833.0,
      "step": 39155
    },
    {
      "entropy": 5.15942349433899,
      "epoch": 4.398045712360307,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003216806381790258,
      "loss": 4.6796,
      "mean_token_accuracy": 0.238067689538002,
      "num_tokens": 87431648.0,
      "step": 39160
    },
    {
      "entropy": 5.288822078704834,
      "epoch": 4.3986072892682655,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003216413622935069,
      "loss": 4.8901,
      "mean_token_accuracy": 0.22721879482269286,
      "num_tokens": 87444207.0,
      "step": 39165
    },
    {
      "entropy": 5.208715105056763,
      "epoch": 4.399168866176223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00032160208492290886,
      "loss": 4.7211,
      "mean_token_accuracy": 0.2434866189956665,
      "num_tokens": 87455919.0,
      "step": 39170
    },
    {
      "entropy": 5.190032196044922,
      "epoch": 4.399730443084181,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032156280606848237,
      "loss": 4.7582,
      "mean_token_accuracy": 0.2358864054083824,
      "num_tokens": 87466979.0,
      "step": 39175
    },
    {
      "entropy": 5.24551100730896,
      "epoch": 4.400292019992138,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003215235257314781,
      "loss": 4.8324,
      "mean_token_accuracy": 0.22693494260311126,
      "num_tokens": 87478193.0,
      "step": 39180
    },
    {
      "entropy": 5.227803659439087,
      "epoch": 4.400853596900095,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032148424391314665,
      "loss": 4.8064,
      "mean_token_accuracy": 0.23253122717142105,
      "num_tokens": 87488961.0,
      "step": 39185
    },
    {
      "entropy": 5.3099206447601315,
      "epoch": 4.401415173808053,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032144496061473895,
      "loss": 4.8611,
      "mean_token_accuracy": 0.23396962136030197,
      "num_tokens": 87500508.0,
      "step": 39190
    },
    {
      "entropy": 5.232723331451416,
      "epoch": 4.401976750716011,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003214056758375056,
      "loss": 4.7526,
      "mean_token_accuracy": 0.23840825706720353,
      "num_tokens": 87510755.0,
      "step": 39195
    },
    {
      "entropy": 5.221077537536621,
      "epoch": 4.402538327623968,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003213663895826976,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23213931173086166,
      "num_tokens": 87522425.0,
      "step": 39200
    },
    {
      "entropy": 5.263457250595093,
      "epoch": 4.403099904531926,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003213271018515657,
      "loss": 4.8477,
      "mean_token_accuracy": 0.2349825382232666,
      "num_tokens": 87534209.0,
      "step": 39205
    },
    {
      "entropy": 5.176665163040161,
      "epoch": 4.403661481439883,
      "grad_norm": 1.0,
      "learning_rate": 0.000321287812645361,
      "loss": 4.7297,
      "mean_token_accuracy": 0.23674239963293076,
      "num_tokens": 87544647.0,
      "step": 39210
    },
    {
      "entropy": 5.19109001159668,
      "epoch": 4.40422305834784,
      "grad_norm": 1.03125,
      "learning_rate": 0.00032124852196533434,
      "loss": 4.6889,
      "mean_token_accuracy": 0.24187662452459335,
      "num_tokens": 87555924.0,
      "step": 39215
    },
    {
      "entropy": 5.169349241256714,
      "epoch": 4.404784635255798,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003212092298127368,
      "loss": 4.7984,
      "mean_token_accuracy": 0.23546454459428787,
      "num_tokens": 87566998.0,
      "step": 39220
    },
    {
      "entropy": 5.186750316619873,
      "epoch": 4.405346212163756,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00032116993618881945,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23066940903663635,
      "num_tokens": 87579516.0,
      "step": 39225
    },
    {
      "entropy": 5.33924708366394,
      "epoch": 4.405907789071714,
      "grad_norm": 0.9375,
      "learning_rate": 0.00032113064109483345,
      "loss": 4.847,
      "mean_token_accuracy": 0.23090673089027405,
      "num_tokens": 87591512.0,
      "step": 39230
    },
    {
      "entropy": 5.241888332366943,
      "epoch": 4.406469365979671,
      "grad_norm": 1.0,
      "learning_rate": 0.00032109134453202993,
      "loss": 4.7798,
      "mean_token_accuracy": 0.2299850732088089,
      "num_tokens": 87602783.0,
      "step": 39235
    },
    {
      "entropy": 5.1624939918518065,
      "epoch": 4.407030942887628,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00032105204650166016,
      "loss": 4.7652,
      "mean_token_accuracy": 0.23276780247688295,
      "num_tokens": 87613585.0,
      "step": 39240
    },
    {
      "entropy": 5.302844142913818,
      "epoch": 4.407592519795586,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032101274700497537,
      "loss": 4.8605,
      "mean_token_accuracy": 0.23215901255607604,
      "num_tokens": 87624003.0,
      "step": 39245
    },
    {
      "entropy": 5.251820802688599,
      "epoch": 4.4081540967035435,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003209734460432269,
      "loss": 4.7629,
      "mean_token_accuracy": 0.2357669323682785,
      "num_tokens": 87634429.0,
      "step": 39250
    },
    {
      "entropy": 5.216601419448852,
      "epoch": 4.408715673611501,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000320934143617666,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23417943120002746,
      "num_tokens": 87645887.0,
      "step": 39255
    },
    {
      "entropy": 5.242512607574463,
      "epoch": 4.409277250519459,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003208948397295441,
      "loss": 4.7871,
      "mean_token_accuracy": 0.22966949343681337,
      "num_tokens": 87657439.0,
      "step": 39260
    },
    {
      "entropy": 5.19781084060669,
      "epoch": 4.409838827427416,
      "grad_norm": 1.0625,
      "learning_rate": 0.00032085553438011285,
      "loss": 4.7538,
      "mean_token_accuracy": 0.2337652027606964,
      "num_tokens": 87668461.0,
      "step": 39265
    },
    {
      "entropy": 5.343347835540771,
      "epoch": 4.410400404335373,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003208162275706235,
      "loss": 4.8706,
      "mean_token_accuracy": 0.22298690676689148,
      "num_tokens": 87680595.0,
      "step": 39270
    },
    {
      "entropy": 5.180088233947754,
      "epoch": 4.410961981243331,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003207769193023277,
      "loss": 4.7561,
      "mean_token_accuracy": 0.23327647000551224,
      "num_tokens": 87691563.0,
      "step": 39275
    },
    {
      "entropy": 5.244667291641235,
      "epoch": 4.411523558151289,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00032073760957647705,
      "loss": 4.8374,
      "mean_token_accuracy": 0.23004306405782698,
      "num_tokens": 87702052.0,
      "step": 39280
    },
    {
      "entropy": 5.213583898544312,
      "epoch": 4.412085135059247,
      "grad_norm": 1.0,
      "learning_rate": 0.0003206982983943232,
      "loss": 4.8045,
      "mean_token_accuracy": 0.2301102489233017,
      "num_tokens": 87713123.0,
      "step": 39285
    },
    {
      "entropy": 5.285949945449829,
      "epoch": 4.412646711967204,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003206589857571178,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22599416077136994,
      "num_tokens": 87723294.0,
      "step": 39290
    },
    {
      "entropy": 5.277732896804809,
      "epoch": 4.413208288875161,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00032061967166611255,
      "loss": 4.7851,
      "mean_token_accuracy": 0.23947374224662782,
      "num_tokens": 87733751.0,
      "step": 39295
    },
    {
      "entropy": 5.2062726497650145,
      "epoch": 4.413769865783119,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00032058035612255934,
      "loss": 4.7549,
      "mean_token_accuracy": 0.24097288697957991,
      "num_tokens": 87744584.0,
      "step": 39300
    },
    {
      "entropy": 5.262624502182007,
      "epoch": 4.4143314426910765,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003205410391277098,
      "loss": 4.7829,
      "mean_token_accuracy": 0.23878986835479737,
      "num_tokens": 87755420.0,
      "step": 39305
    },
    {
      "entropy": 5.276637220382691,
      "epoch": 4.414893019599034,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00032050172068281594,
      "loss": 4.8344,
      "mean_token_accuracy": 0.22823706567287444,
      "num_tokens": 87767213.0,
      "step": 39310
    },
    {
      "entropy": 5.2191246509552,
      "epoch": 4.415454596506992,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003204624007891297,
      "loss": 4.8113,
      "mean_token_accuracy": 0.23210841417312622,
      "num_tokens": 87778058.0,
      "step": 39315
    },
    {
      "entropy": 5.250999546051025,
      "epoch": 4.416016173414949,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000320423079447903,
      "loss": 4.8179,
      "mean_token_accuracy": 0.22651427686214448,
      "num_tokens": 87788279.0,
      "step": 39320
    },
    {
      "entropy": 5.217709732055664,
      "epoch": 4.416577750322907,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00032038375666038776,
      "loss": 4.7547,
      "mean_token_accuracy": 0.23148721158504487,
      "num_tokens": 87799643.0,
      "step": 39325
    },
    {
      "entropy": 5.18208589553833,
      "epoch": 4.417139327230864,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003203444324278361,
      "loss": 4.7324,
      "mean_token_accuracy": 0.23903499394655228,
      "num_tokens": 87810493.0,
      "step": 39330
    },
    {
      "entropy": 5.244478607177735,
      "epoch": 4.4177009041388215,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003203051067515002,
      "loss": 4.8149,
      "mean_token_accuracy": 0.22977941781282424,
      "num_tokens": 87821876.0,
      "step": 39335
    },
    {
      "entropy": 5.247426319122314,
      "epoch": 4.41826248104678,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003202657796326321,
      "loss": 4.7433,
      "mean_token_accuracy": 0.24631451517343522,
      "num_tokens": 87832058.0,
      "step": 39340
    },
    {
      "entropy": 5.240084314346314,
      "epoch": 4.418824057954737,
      "grad_norm": 1.046875,
      "learning_rate": 0.000320226451072484,
      "loss": 4.8049,
      "mean_token_accuracy": 0.22819913178682327,
      "num_tokens": 87842549.0,
      "step": 39345
    },
    {
      "entropy": 5.156752490997315,
      "epoch": 4.419385634862694,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00032018712107230815,
      "loss": 4.7718,
      "mean_token_accuracy": 0.2387339249253273,
      "num_tokens": 87853507.0,
      "step": 39350
    },
    {
      "entropy": 5.2964009761810305,
      "epoch": 4.419947211770652,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003201477896333568,
      "loss": 4.8137,
      "mean_token_accuracy": 0.23194118291139604,
      "num_tokens": 87864546.0,
      "step": 39355
    },
    {
      "entropy": 5.276336669921875,
      "epoch": 4.420508788678609,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003201084567568824,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24171165972948075,
      "num_tokens": 87876553.0,
      "step": 39360
    },
    {
      "entropy": 5.207035064697266,
      "epoch": 4.4210703655865675,
      "grad_norm": 1.078125,
      "learning_rate": 0.00032006912244413727,
      "loss": 4.7425,
      "mean_token_accuracy": 0.23600226938724517,
      "num_tokens": 87886257.0,
      "step": 39365
    },
    {
      "entropy": 5.179496002197266,
      "epoch": 4.421631942494525,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003200297866963737,
      "loss": 4.6783,
      "mean_token_accuracy": 0.2402724042534828,
      "num_tokens": 87897704.0,
      "step": 39370
    },
    {
      "entropy": 5.177543878555298,
      "epoch": 4.422193519402482,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003199904495148443,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23843200802803038,
      "num_tokens": 87908892.0,
      "step": 39375
    },
    {
      "entropy": 5.229150724411011,
      "epoch": 4.42275509631044,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003199511109008016,
      "loss": 4.8553,
      "mean_token_accuracy": 0.23486700057983398,
      "num_tokens": 87921502.0,
      "step": 39380
    },
    {
      "entropy": 5.27090892791748,
      "epoch": 4.423316673218397,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031991177085549807,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23449390530586242,
      "num_tokens": 87933250.0,
      "step": 39385
    },
    {
      "entropy": 5.205315208435058,
      "epoch": 4.4238782501263545,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031987242938018626,
      "loss": 4.7314,
      "mean_token_accuracy": 0.24229945987462997,
      "num_tokens": 87945715.0,
      "step": 39390
    },
    {
      "entropy": 5.1785211086273195,
      "epoch": 4.424439827034313,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031983308647611896,
      "loss": 4.7282,
      "mean_token_accuracy": 0.2430281311273575,
      "num_tokens": 87956973.0,
      "step": 39395
    },
    {
      "entropy": 5.272212982177734,
      "epoch": 4.42500140394227,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003197937421445488,
      "loss": 4.8596,
      "mean_token_accuracy": 0.2345861777663231,
      "num_tokens": 87968341.0,
      "step": 39400
    },
    {
      "entropy": 5.190668821334839,
      "epoch": 4.425562980850227,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003197543963867285,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23016917258501052,
      "num_tokens": 87978980.0,
      "step": 39405
    },
    {
      "entropy": 5.28774209022522,
      "epoch": 4.426124557758185,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003197150492039108,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23804595321416855,
      "num_tokens": 87990612.0,
      "step": 39410
    },
    {
      "entropy": 5.25946798324585,
      "epoch": 4.426686134666142,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003196757005973487,
      "loss": 4.7512,
      "mean_token_accuracy": 0.23246829956769943,
      "num_tokens": 88001052.0,
      "step": 39415
    },
    {
      "entropy": 5.220422077178955,
      "epoch": 4.4272477115741005,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031963635056829487,
      "loss": 4.8161,
      "mean_token_accuracy": 0.23092371225357056,
      "num_tokens": 88013102.0,
      "step": 39420
    },
    {
      "entropy": 5.248900127410889,
      "epoch": 4.427809288482058,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031959699911800237,
      "loss": 4.8674,
      "mean_token_accuracy": 0.23280680924654007,
      "num_tokens": 88023952.0,
      "step": 39425
    },
    {
      "entropy": 5.224085664749145,
      "epoch": 4.428370865390015,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003195576462477241,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23143236637115477,
      "num_tokens": 88035148.0,
      "step": 39430
    },
    {
      "entropy": 5.270566606521607,
      "epoch": 4.428932442297973,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031951829195871315,
      "loss": 4.8703,
      "mean_token_accuracy": 0.22879950553178788,
      "num_tokens": 88046294.0,
      "step": 39435
    },
    {
      "entropy": 5.215512752532959,
      "epoch": 4.42949401920593,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031947893625222244,
      "loss": 4.7804,
      "mean_token_accuracy": 0.22914931625127793,
      "num_tokens": 88058225.0,
      "step": 39440
    },
    {
      "entropy": 5.2372150897979735,
      "epoch": 4.430055596113887,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031943957912950523,
      "loss": 4.8178,
      "mean_token_accuracy": 0.2280558839440346,
      "num_tokens": 88069994.0,
      "step": 39445
    },
    {
      "entropy": 5.278449487686157,
      "epoch": 4.4306171730218455,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003194002205918145,
      "loss": 4.798,
      "mean_token_accuracy": 0.23580312877893447,
      "num_tokens": 88081912.0,
      "step": 39450
    },
    {
      "entropy": 5.275518560409546,
      "epoch": 4.431178749929803,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003193608606404036,
      "loss": 4.7822,
      "mean_token_accuracy": 0.23921035081148148,
      "num_tokens": 88092457.0,
      "step": 39455
    },
    {
      "entropy": 5.334955358505249,
      "epoch": 4.43174032683776,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003193214992765257,
      "loss": 4.9488,
      "mean_token_accuracy": 0.21850700080394744,
      "num_tokens": 88104407.0,
      "step": 39460
    },
    {
      "entropy": 5.187772130966186,
      "epoch": 4.432301903745718,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031928213650143407,
      "loss": 4.7422,
      "mean_token_accuracy": 0.23854355812072753,
      "num_tokens": 88114653.0,
      "step": 39465
    },
    {
      "entropy": 5.31070613861084,
      "epoch": 4.432863480653675,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000319242772316382,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2399245023727417,
      "num_tokens": 88125477.0,
      "step": 39470
    },
    {
      "entropy": 5.215651416778565,
      "epoch": 4.433425057561633,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003192034067226228,
      "loss": 4.7045,
      "mean_token_accuracy": 0.2450124055147171,
      "num_tokens": 88136419.0,
      "step": 39475
    },
    {
      "entropy": 5.177561855316162,
      "epoch": 4.433986634469591,
      "grad_norm": 1.125,
      "learning_rate": 0.00031916403972141017,
      "loss": 4.7057,
      "mean_token_accuracy": 0.24516120851039885,
      "num_tokens": 88146578.0,
      "step": 39480
    },
    {
      "entropy": 5.0536524772644045,
      "epoch": 4.434548211377548,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003191246713139973,
      "loss": 4.646,
      "mean_token_accuracy": 0.2399264767765999,
      "num_tokens": 88157351.0,
      "step": 39485
    },
    {
      "entropy": 5.2414521217346195,
      "epoch": 4.435109788285506,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031908530150163777,
      "loss": 4.8176,
      "mean_token_accuracy": 0.23455170840024947,
      "num_tokens": 88168460.0,
      "step": 39490
    },
    {
      "entropy": 5.303431987762451,
      "epoch": 4.435671365193463,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00031904593028558517,
      "loss": 4.8197,
      "mean_token_accuracy": 0.23202313333749772,
      "num_tokens": 88178994.0,
      "step": 39495
    },
    {
      "entropy": 5.234448385238648,
      "epoch": 4.43623294210142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031900655766709305,
      "loss": 4.7328,
      "mean_token_accuracy": 0.23401332646608353,
      "num_tokens": 88188916.0,
      "step": 39500
    },
    {
      "entropy": 5.152022981643677,
      "epoch": 4.4367945190093785,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003189671836474151,
      "loss": 4.7915,
      "mean_token_accuracy": 0.2336057424545288,
      "num_tokens": 88200833.0,
      "step": 39505
    },
    {
      "entropy": 5.237713241577149,
      "epoch": 4.437356095917336,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003189278082278049,
      "loss": 4.8891,
      "mean_token_accuracy": 0.2254093810915947,
      "num_tokens": 88212022.0,
      "step": 39510
    },
    {
      "entropy": 5.256261777877808,
      "epoch": 4.437917672825294,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031888843140951627,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24303473234176637,
      "num_tokens": 88222327.0,
      "step": 39515
    },
    {
      "entropy": 5.117857027053833,
      "epoch": 4.438479249733251,
      "grad_norm": 1.0,
      "learning_rate": 0.0003188490531938031,
      "loss": 4.6708,
      "mean_token_accuracy": 0.2457813560962677,
      "num_tokens": 88233141.0,
      "step": 39520
    },
    {
      "entropy": 5.286913204193115,
      "epoch": 4.439040826641208,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031880967358191884,
      "loss": 4.8793,
      "mean_token_accuracy": 0.22483659535646439,
      "num_tokens": 88245663.0,
      "step": 39525
    },
    {
      "entropy": 5.262447023391724,
      "epoch": 4.439602403549166,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031877029257511774,
      "loss": 4.7571,
      "mean_token_accuracy": 0.23564204275608064,
      "num_tokens": 88256291.0,
      "step": 39530
    },
    {
      "entropy": 5.30471658706665,
      "epoch": 4.440163980457124,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031873091017465356,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22592180967330933,
      "num_tokens": 88268382.0,
      "step": 39535
    },
    {
      "entropy": 5.14695463180542,
      "epoch": 4.440725557365081,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003186915263817802,
      "loss": 4.747,
      "mean_token_accuracy": 0.24218785762786865,
      "num_tokens": 88279638.0,
      "step": 39540
    },
    {
      "entropy": 5.248746919631958,
      "epoch": 4.441287134273039,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031865214119775173,
      "loss": 4.8295,
      "mean_token_accuracy": 0.23606770634651184,
      "num_tokens": 88290599.0,
      "step": 39545
    },
    {
      "entropy": 5.268609142303466,
      "epoch": 4.441848711180996,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000318612754623822,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23418606370687484,
      "num_tokens": 88301572.0,
      "step": 39550
    },
    {
      "entropy": 5.310630893707275,
      "epoch": 4.442410288088954,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031857336666124537,
      "loss": 4.8717,
      "mean_token_accuracy": 0.2312408909201622,
      "num_tokens": 88312978.0,
      "step": 39555
    },
    {
      "entropy": 5.238432693481445,
      "epoch": 4.442971864996911,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003185339773112758,
      "loss": 4.8503,
      "mean_token_accuracy": 0.2278346449136734,
      "num_tokens": 88324023.0,
      "step": 39560
    },
    {
      "entropy": 5.213559293746949,
      "epoch": 4.443533441904869,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0003184945865751675,
      "loss": 4.7385,
      "mean_token_accuracy": 0.22931316941976548,
      "num_tokens": 88335366.0,
      "step": 39565
    },
    {
      "entropy": 5.246920347213745,
      "epoch": 4.444095018812827,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031845519445417475,
      "loss": 4.7972,
      "mean_token_accuracy": 0.23312366902828216,
      "num_tokens": 88347786.0,
      "step": 39570
    },
    {
      "entropy": 5.3088301658630375,
      "epoch": 4.444656595720784,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003184158009495518,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23040181994438172,
      "num_tokens": 88358720.0,
      "step": 39575
    },
    {
      "entropy": 5.2476671695709225,
      "epoch": 4.445218172628741,
      "grad_norm": 1.0,
      "learning_rate": 0.0003183764060625528,
      "loss": 4.8508,
      "mean_token_accuracy": 0.22916352450847627,
      "num_tokens": 88370152.0,
      "step": 39580
    },
    {
      "entropy": 5.198625326156616,
      "epoch": 4.445779749536699,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031833700979443227,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23159035593271254,
      "num_tokens": 88381385.0,
      "step": 39585
    },
    {
      "entropy": 5.229145336151123,
      "epoch": 4.4463413264446565,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031829761214644453,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23175157606601715,
      "num_tokens": 88391450.0,
      "step": 39590
    },
    {
      "entropy": 5.148619699478149,
      "epoch": 4.446902903352614,
      "grad_norm": 1.078125,
      "learning_rate": 0.000318258213119844,
      "loss": 4.7029,
      "mean_token_accuracy": 0.24810463190078735,
      "num_tokens": 88402310.0,
      "step": 39595
    },
    {
      "entropy": 5.203392124176025,
      "epoch": 4.447464480260572,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003182188127158852,
      "loss": 4.8065,
      "mean_token_accuracy": 0.22136763483285904,
      "num_tokens": 88412286.0,
      "step": 39600
    },
    {
      "entropy": 5.304801750183105,
      "epoch": 4.448026057168529,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003181794109358227,
      "loss": 4.8488,
      "mean_token_accuracy": 0.23054244220256806,
      "num_tokens": 88423335.0,
      "step": 39605
    },
    {
      "entropy": 5.183753252029419,
      "epoch": 4.448587634076487,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003181400077809109,
      "loss": 4.7391,
      "mean_token_accuracy": 0.23917068392038346,
      "num_tokens": 88434328.0,
      "step": 39610
    },
    {
      "entropy": 5.234421253204346,
      "epoch": 4.449149210984444,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031810060325240466,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23451482653617858,
      "num_tokens": 88445611.0,
      "step": 39615
    },
    {
      "entropy": 5.253994178771973,
      "epoch": 4.449710787892402,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003180611973515583,
      "loss": 4.8011,
      "mean_token_accuracy": 0.23446896821260452,
      "num_tokens": 88456207.0,
      "step": 39620
    },
    {
      "entropy": 5.3210186004638675,
      "epoch": 4.45027236480036,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031802179007962687,
      "loss": 4.9462,
      "mean_token_accuracy": 0.22485428899526597,
      "num_tokens": 88467322.0,
      "step": 39625
    },
    {
      "entropy": 5.242198467254639,
      "epoch": 4.450833941708317,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003179823814378649,
      "loss": 4.859,
      "mean_token_accuracy": 0.23114392310380935,
      "num_tokens": 88479703.0,
      "step": 39630
    },
    {
      "entropy": 5.158197832107544,
      "epoch": 4.451395518616274,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00031794297142752727,
      "loss": 4.7416,
      "mean_token_accuracy": 0.2383045956492424,
      "num_tokens": 88490284.0,
      "step": 39635
    },
    {
      "entropy": 5.225045919418335,
      "epoch": 4.451957095524232,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003179035600498687,
      "loss": 4.8302,
      "mean_token_accuracy": 0.22716230452060698,
      "num_tokens": 88502232.0,
      "step": 39640
    },
    {
      "entropy": 5.204051685333252,
      "epoch": 4.4525186724321895,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031786414730614423,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23440620452165603,
      "num_tokens": 88513788.0,
      "step": 39645
    },
    {
      "entropy": 5.209995746612549,
      "epoch": 4.453080249340147,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003178247331976086,
      "loss": 4.8109,
      "mean_token_accuracy": 0.22939417362213135,
      "num_tokens": 88524942.0,
      "step": 39650
    },
    {
      "entropy": 5.143832969665527,
      "epoch": 4.453641826248105,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00031778531772551696,
      "loss": 4.6633,
      "mean_token_accuracy": 0.24060776382684707,
      "num_tokens": 88535692.0,
      "step": 39655
    },
    {
      "entropy": 5.203570604324341,
      "epoch": 4.454203403156062,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003177459008911241,
      "loss": 4.759,
      "mean_token_accuracy": 0.23687535971403123,
      "num_tokens": 88547007.0,
      "step": 39660
    },
    {
      "entropy": 5.295115327835083,
      "epoch": 4.45476498006402,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031770648269568524,
      "loss": 4.8637,
      "mean_token_accuracy": 0.23633944392204284,
      "num_tokens": 88558197.0,
      "step": 39665
    },
    {
      "entropy": 5.303886270523071,
      "epoch": 4.455326556971977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003176670631404554,
      "loss": 4.8679,
      "mean_token_accuracy": 0.22867240607738495,
      "num_tokens": 88569488.0,
      "step": 39670
    },
    {
      "entropy": 5.252887916564942,
      "epoch": 4.4558881338799345,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031762764222668976,
      "loss": 4.7998,
      "mean_token_accuracy": 0.22912787795066833,
      "num_tokens": 88582054.0,
      "step": 39675
    },
    {
      "entropy": 5.212625885009766,
      "epoch": 4.456449710787893,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031758821995564343,
      "loss": 4.7124,
      "mean_token_accuracy": 0.2370191365480423,
      "num_tokens": 88592528.0,
      "step": 39680
    },
    {
      "entropy": 5.2534956455230715,
      "epoch": 4.45701128769585,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0003175487963285716,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23522479236125945,
      "num_tokens": 88604136.0,
      "step": 39685
    },
    {
      "entropy": 5.294599437713623,
      "epoch": 4.457572864603807,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003175093713467297,
      "loss": 4.8537,
      "mean_token_accuracy": 0.22732323706150054,
      "num_tokens": 88615629.0,
      "step": 39690
    },
    {
      "entropy": 5.145235061645508,
      "epoch": 4.458134441511765,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031746994501137283,
      "loss": 4.7049,
      "mean_token_accuracy": 0.24298588782548905,
      "num_tokens": 88626100.0,
      "step": 39695
    },
    {
      "entropy": 5.187931537628174,
      "epoch": 4.458696018419722,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003174305173237565,
      "loss": 4.852,
      "mean_token_accuracy": 0.22829025685787202,
      "num_tokens": 88638948.0,
      "step": 39700
    },
    {
      "entropy": 5.267453002929687,
      "epoch": 4.4592575953276805,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000317391088285136,
      "loss": 4.8356,
      "mean_token_accuracy": 0.2340323284268379,
      "num_tokens": 88650488.0,
      "step": 39705
    },
    {
      "entropy": 5.3485315322875975,
      "epoch": 4.459819172235638,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003173516578967668,
      "loss": 4.8868,
      "mean_token_accuracy": 0.22383270412683487,
      "num_tokens": 88661584.0,
      "step": 39710
    },
    {
      "entropy": 5.173656415939331,
      "epoch": 4.460380749143595,
      "grad_norm": 0.921875,
      "learning_rate": 0.00031731222615990443,
      "loss": 4.7576,
      "mean_token_accuracy": 0.23252669125795364,
      "num_tokens": 88673051.0,
      "step": 39715
    },
    {
      "entropy": 5.259626388549805,
      "epoch": 4.460942326051553,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031727279307580436,
      "loss": 4.8823,
      "mean_token_accuracy": 0.22883908599615096,
      "num_tokens": 88684122.0,
      "step": 39720
    },
    {
      "entropy": 5.340874195098877,
      "epoch": 4.46150390295951,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031723335864572216,
      "loss": 4.9064,
      "mean_token_accuracy": 0.22950998246669768,
      "num_tokens": 88694620.0,
      "step": 39725
    },
    {
      "entropy": 5.3574176788330075,
      "epoch": 4.4620654798674675,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031719392287091343,
      "loss": 4.8967,
      "mean_token_accuracy": 0.228570057451725,
      "num_tokens": 88706137.0,
      "step": 39730
    },
    {
      "entropy": 5.200046730041504,
      "epoch": 4.462627056775426,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003171544857526338,
      "loss": 4.7322,
      "mean_token_accuracy": 0.2303127869963646,
      "num_tokens": 88717326.0,
      "step": 39735
    },
    {
      "entropy": 5.250757122039795,
      "epoch": 4.463188633683383,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003171150472921391,
      "loss": 4.7657,
      "mean_token_accuracy": 0.23384772837162018,
      "num_tokens": 88727125.0,
      "step": 39740
    },
    {
      "entropy": 5.220198631286621,
      "epoch": 4.463750210591341,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031707560749068487,
      "loss": 4.7486,
      "mean_token_accuracy": 0.2329125478863716,
      "num_tokens": 88737098.0,
      "step": 39745
    },
    {
      "entropy": 5.224533367156982,
      "epoch": 4.464311787499298,
      "grad_norm": 1.046875,
      "learning_rate": 0.000317036166349527,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23571185022592545,
      "num_tokens": 88748454.0,
      "step": 39750
    },
    {
      "entropy": 5.207876348495484,
      "epoch": 4.464873364407255,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003169967238699213,
      "loss": 4.7857,
      "mean_token_accuracy": 0.23347359150648117,
      "num_tokens": 88760165.0,
      "step": 39755
    },
    {
      "entropy": 5.218605470657349,
      "epoch": 4.4654349413152135,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00031695728005312363,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23089263439178467,
      "num_tokens": 88772124.0,
      "step": 39760
    },
    {
      "entropy": 5.168553209304809,
      "epoch": 4.465996518223171,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003169178349003898,
      "loss": 4.7237,
      "mean_token_accuracy": 0.2374151185154915,
      "num_tokens": 88782429.0,
      "step": 39765
    },
    {
      "entropy": 5.201882171630859,
      "epoch": 4.466558095131128,
      "grad_norm": 1.0,
      "learning_rate": 0.000316878388412976,
      "loss": 4.8669,
      "mean_token_accuracy": 0.23152701407670975,
      "num_tokens": 88794060.0,
      "step": 39770
    },
    {
      "entropy": 5.314175176620483,
      "epoch": 4.467119672039086,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000316838940592138,
      "loss": 4.8403,
      "mean_token_accuracy": 0.23509023189544678,
      "num_tokens": 88806174.0,
      "step": 39775
    },
    {
      "entropy": 5.184953403472901,
      "epoch": 4.467681248947043,
      "grad_norm": 1.0,
      "learning_rate": 0.00031679949143913183,
      "loss": 4.6907,
      "mean_token_accuracy": 0.23541554808616638,
      "num_tokens": 88817587.0,
      "step": 39780
    },
    {
      "entropy": 5.240903520584107,
      "epoch": 4.468242825855,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003167600409552137,
      "loss": 4.7582,
      "mean_token_accuracy": 0.2415437176823616,
      "num_tokens": 88829435.0,
      "step": 39785
    },
    {
      "entropy": 5.184481143951416,
      "epoch": 4.4688044027629585,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003167205891416397,
      "loss": 4.8578,
      "mean_token_accuracy": 0.2269921198487282,
      "num_tokens": 88840244.0,
      "step": 39790
    },
    {
      "entropy": 5.345177221298218,
      "epoch": 4.469365979670916,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000316681135999666,
      "loss": 4.9048,
      "mean_token_accuracy": 0.23118101358413695,
      "num_tokens": 88852085.0,
      "step": 39795
    },
    {
      "entropy": 5.3150818824768065,
      "epoch": 4.469927556578874,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003166416815305486,
      "loss": 4.9011,
      "mean_token_accuracy": 0.2280762180685997,
      "num_tokens": 88864605.0,
      "step": 39800
    },
    {
      "entropy": 5.265274906158448,
      "epoch": 4.470489133486831,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031660222573554406,
      "loss": 4.783,
      "mean_token_accuracy": 0.23274572640657426,
      "num_tokens": 88876209.0,
      "step": 39805
    },
    {
      "entropy": 5.2121467113494875,
      "epoch": 4.471050710394788,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003165627686159085,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23195394724607468,
      "num_tokens": 88886346.0,
      "step": 39810
    },
    {
      "entropy": 5.187676334381104,
      "epoch": 4.471612287302746,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003165233101728982,
      "loss": 4.7962,
      "mean_token_accuracy": 0.23443193286657332,
      "num_tokens": 88898190.0,
      "step": 39815
    },
    {
      "entropy": 5.247471714019776,
      "epoch": 4.472173864210704,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031648385040776965,
      "loss": 4.8842,
      "mean_token_accuracy": 0.22001990228891372,
      "num_tokens": 88909957.0,
      "step": 39820
    },
    {
      "entropy": 5.415828514099121,
      "epoch": 4.472735441118661,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003164443893217793,
      "loss": 4.9605,
      "mean_token_accuracy": 0.2243284821510315,
      "num_tokens": 88922993.0,
      "step": 39825
    },
    {
      "entropy": 5.160286378860474,
      "epoch": 4.473297018026619,
      "grad_norm": 1.0,
      "learning_rate": 0.0003164049269161834,
      "loss": 4.6969,
      "mean_token_accuracy": 0.24054068326950073,
      "num_tokens": 88934506.0,
      "step": 39830
    },
    {
      "entropy": 5.2017467498779295,
      "epoch": 4.473858594934576,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031636546319223845,
      "loss": 4.7964,
      "mean_token_accuracy": 0.2363939791917801,
      "num_tokens": 88946963.0,
      "step": 39835
    },
    {
      "entropy": 5.267075872421264,
      "epoch": 4.474420171842533,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003163259981512013,
      "loss": 4.8626,
      "mean_token_accuracy": 0.23339330404996872,
      "num_tokens": 88956995.0,
      "step": 39840
    },
    {
      "entropy": 5.332167387008667,
      "epoch": 4.4749817487504915,
      "grad_norm": 1.0,
      "learning_rate": 0.0003162865317943283,
      "loss": 4.8695,
      "mean_token_accuracy": 0.2314709886908531,
      "num_tokens": 88967507.0,
      "step": 39845
    },
    {
      "entropy": 5.318812322616577,
      "epoch": 4.475543325658449,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031624706412287614,
      "loss": 4.8645,
      "mean_token_accuracy": 0.23368800133466722,
      "num_tokens": 88979943.0,
      "step": 39850
    },
    {
      "entropy": 5.294915819168091,
      "epoch": 4.476104902566407,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003162075951381014,
      "loss": 4.8945,
      "mean_token_accuracy": 0.2246240720152855,
      "num_tokens": 88990840.0,
      "step": 39855
    },
    {
      "entropy": 5.311063385009765,
      "epoch": 4.476666479474364,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031616812484126084,
      "loss": 4.878,
      "mean_token_accuracy": 0.22136164605617523,
      "num_tokens": 89002245.0,
      "step": 39860
    },
    {
      "entropy": 5.341882562637329,
      "epoch": 4.477228056382321,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003161286532336113,
      "loss": 4.8706,
      "mean_token_accuracy": 0.23300074487924577,
      "num_tokens": 89013520.0,
      "step": 39865
    },
    {
      "entropy": 5.277456903457642,
      "epoch": 4.477789633290279,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00031608918031640946,
      "loss": 4.7909,
      "mean_token_accuracy": 0.23514578938484193,
      "num_tokens": 89025377.0,
      "step": 39870
    },
    {
      "entropy": 5.20879020690918,
      "epoch": 4.478351210198237,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003160497060909121,
      "loss": 4.7814,
      "mean_token_accuracy": 0.2292601779103279,
      "num_tokens": 89035299.0,
      "step": 39875
    },
    {
      "entropy": 5.267144584655762,
      "epoch": 4.478912787106194,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031601023055837624,
      "loss": 4.8466,
      "mean_token_accuracy": 0.23111254572868348,
      "num_tokens": 89046255.0,
      "step": 39880
    },
    {
      "entropy": 5.32869234085083,
      "epoch": 4.479474364014152,
      "grad_norm": 1.21875,
      "learning_rate": 0.00031597075372005875,
      "loss": 4.905,
      "mean_token_accuracy": 0.22143975496292115,
      "num_tokens": 89057873.0,
      "step": 39885
    },
    {
      "entropy": 5.218490076065064,
      "epoch": 4.480035940922109,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00031593127557721657,
      "loss": 4.6952,
      "mean_token_accuracy": 0.2370813235640526,
      "num_tokens": 89068830.0,
      "step": 39890
    },
    {
      "entropy": 5.194696760177612,
      "epoch": 4.480597517830067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003158917961311067,
      "loss": 4.8048,
      "mean_token_accuracy": 0.23848582059144974,
      "num_tokens": 89080328.0,
      "step": 39895
    },
    {
      "entropy": 5.277004098892212,
      "epoch": 4.481159094738024,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003158523153829862,
      "loss": 4.8181,
      "mean_token_accuracy": 0.22576351314783097,
      "num_tokens": 89090118.0,
      "step": 39900
    },
    {
      "entropy": 5.308381986618042,
      "epoch": 4.481720671645982,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003158128333341121,
      "loss": 4.8167,
      "mean_token_accuracy": 0.2316090703010559,
      "num_tokens": 89103338.0,
      "step": 39905
    },
    {
      "entropy": 5.281309461593628,
      "epoch": 4.48228224855394,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031577334998574155,
      "loss": 4.7708,
      "mean_token_accuracy": 0.23708402663469313,
      "num_tokens": 89113721.0,
      "step": 39910
    },
    {
      "entropy": 5.334532928466797,
      "epoch": 4.482843825461897,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003157338653391317,
      "loss": 4.8722,
      "mean_token_accuracy": 0.23247384577989577,
      "num_tokens": 89125425.0,
      "step": 39915
    },
    {
      "entropy": 5.252283000946045,
      "epoch": 4.483405402369854,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003156943793955398,
      "loss": 4.8459,
      "mean_token_accuracy": 0.23239072561264038,
      "num_tokens": 89135407.0,
      "step": 39920
    },
    {
      "entropy": 5.294947290420533,
      "epoch": 4.483966979277812,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031565489215622304,
      "loss": 4.861,
      "mean_token_accuracy": 0.23232651203870774,
      "num_tokens": 89147563.0,
      "step": 39925
    },
    {
      "entropy": 5.309891223907471,
      "epoch": 4.4845285561857695,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003156154036224388,
      "loss": 4.8408,
      "mean_token_accuracy": 0.23135845959186555,
      "num_tokens": 89158336.0,
      "step": 39930
    },
    {
      "entropy": 5.172334861755371,
      "epoch": 4.485090133093728,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003155759137954444,
      "loss": 4.7104,
      "mean_token_accuracy": 0.23990746289491655,
      "num_tokens": 89168894.0,
      "step": 39935
    },
    {
      "entropy": 5.275461864471436,
      "epoch": 4.485651710001685,
      "grad_norm": 1.140625,
      "learning_rate": 0.00031553642267649715,
      "loss": 4.8435,
      "mean_token_accuracy": 0.23609223663806916,
      "num_tokens": 89180701.0,
      "step": 39940
    },
    {
      "entropy": 5.252439260482788,
      "epoch": 4.486213286909642,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003154969302668544,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23524011075496673,
      "num_tokens": 89191192.0,
      "step": 39945
    },
    {
      "entropy": 5.240339708328247,
      "epoch": 4.4867748638176,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003154574365677737,
      "loss": 4.7425,
      "mean_token_accuracy": 0.23853860497474672,
      "num_tokens": 89202377.0,
      "step": 39950
    },
    {
      "entropy": 5.177823543548584,
      "epoch": 4.487336440725557,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003154179415805126,
      "loss": 4.7159,
      "mean_token_accuracy": 0.2429713323712349,
      "num_tokens": 89212869.0,
      "step": 39955
    },
    {
      "entropy": 5.213218927383423,
      "epoch": 4.487898017633515,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0003153784453063285,
      "loss": 4.8143,
      "mean_token_accuracy": 0.2288116231560707,
      "num_tokens": 89224558.0,
      "step": 39960
    },
    {
      "entropy": 5.197320747375488,
      "epoch": 4.488459594541473,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003153389477464791,
      "loss": 4.7362,
      "mean_token_accuracy": 0.23942475467920304,
      "num_tokens": 89235645.0,
      "step": 39965
    },
    {
      "entropy": 5.186260223388672,
      "epoch": 4.48902117144943,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003152994489022219,
      "loss": 4.7394,
      "mean_token_accuracy": 0.23680182546377182,
      "num_tokens": 89247063.0,
      "step": 39970
    },
    {
      "entropy": 5.231364440917969,
      "epoch": 4.489582748357387,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031525994877481465,
      "loss": 4.8577,
      "mean_token_accuracy": 0.22532792538404464,
      "num_tokens": 89258297.0,
      "step": 39975
    },
    {
      "entropy": 5.237774181365967,
      "epoch": 4.490144325265345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000315220447365515,
      "loss": 4.7736,
      "mean_token_accuracy": 0.23494705706834793,
      "num_tokens": 89269245.0,
      "step": 39980
    },
    {
      "entropy": 5.295721578598022,
      "epoch": 4.4907059021733025,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031518094467558074,
      "loss": 4.8633,
      "mean_token_accuracy": 0.22701956778764726,
      "num_tokens": 89281959.0,
      "step": 39985
    },
    {
      "entropy": 5.241893529891968,
      "epoch": 4.491267479081261,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031514144070626956,
      "loss": 4.7601,
      "mean_token_accuracy": 0.2398478552699089,
      "num_tokens": 89292656.0,
      "step": 39990
    },
    {
      "entropy": 5.239465427398682,
      "epoch": 4.491829055989218,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031510193545883944,
      "loss": 4.8556,
      "mean_token_accuracy": 0.23101551830768585,
      "num_tokens": 89304198.0,
      "step": 39995
    },
    {
      "entropy": 5.227236032485962,
      "epoch": 4.492390632897175,
      "grad_norm": 1.109375,
      "learning_rate": 0.000315062428934548,
      "loss": 4.8142,
      "mean_token_accuracy": 0.23766385316848754,
      "num_tokens": 89315443.0,
      "step": 40000
    },
    {
      "entropy": 5.224753189086914,
      "epoch": 4.492952209805133,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031502292113465334,
      "loss": 4.8395,
      "mean_token_accuracy": 0.23082965463399888,
      "num_tokens": 89326597.0,
      "step": 40005
    },
    {
      "entropy": 5.29864239692688,
      "epoch": 4.49351378671309,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031498341206041337,
      "loss": 4.8307,
      "mean_token_accuracy": 0.22887840270996093,
      "num_tokens": 89338385.0,
      "step": 40010
    },
    {
      "entropy": 5.203031206130982,
      "epoch": 4.4940753636210475,
      "grad_norm": 0.984375,
      "learning_rate": 0.000314943901713086,
      "loss": 4.7605,
      "mean_token_accuracy": 0.240002278983593,
      "num_tokens": 89350950.0,
      "step": 40015
    },
    {
      "entropy": 5.212242317199707,
      "epoch": 4.494636940529006,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003149043900939294,
      "loss": 4.8298,
      "mean_token_accuracy": 0.23008304685354233,
      "num_tokens": 89361154.0,
      "step": 40020
    },
    {
      "entropy": 5.274960947036743,
      "epoch": 4.495198517436963,
      "grad_norm": 1.0,
      "learning_rate": 0.00031486487720420146,
      "loss": 4.8459,
      "mean_token_accuracy": 0.22809950709342958,
      "num_tokens": 89373188.0,
      "step": 40025
    },
    {
      "entropy": 5.263661098480225,
      "epoch": 4.49576009434492,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031482536304516047,
      "loss": 4.7782,
      "mean_token_accuracy": 0.23636733293533324,
      "num_tokens": 89383651.0,
      "step": 40030
    },
    {
      "entropy": 5.157902479171753,
      "epoch": 4.496321671252878,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003147858476180644,
      "loss": 4.7123,
      "mean_token_accuracy": 0.23561801612377167,
      "num_tokens": 89395385.0,
      "step": 40035
    },
    {
      "entropy": 5.188540458679199,
      "epoch": 4.496883248160835,
      "grad_norm": 1.0,
      "learning_rate": 0.00031474633092417155,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23705560117959976,
      "num_tokens": 89406071.0,
      "step": 40040
    },
    {
      "entropy": 5.329393815994263,
      "epoch": 4.4974448250687935,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031470681296474006,
      "loss": 4.9169,
      "mean_token_accuracy": 0.22667704969644548,
      "num_tokens": 89417044.0,
      "step": 40045
    },
    {
      "entropy": 5.249815845489502,
      "epoch": 4.498006401976751,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003146672937410283,
      "loss": 4.7604,
      "mean_token_accuracy": 0.22982120364904404,
      "num_tokens": 89428253.0,
      "step": 40050
    },
    {
      "entropy": 5.267969846725464,
      "epoch": 4.498567978884708,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031462777325429447,
      "loss": 4.8188,
      "mean_token_accuracy": 0.22446689903736114,
      "num_tokens": 89438760.0,
      "step": 40055
    },
    {
      "entropy": 5.18108024597168,
      "epoch": 4.499129555792666,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000314588251505797,
      "loss": 4.7557,
      "mean_token_accuracy": 0.2484777480363846,
      "num_tokens": 89450335.0,
      "step": 40060
    },
    {
      "entropy": 5.311463499069214,
      "epoch": 4.499691132700623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031454872849679425,
      "loss": 4.8294,
      "mean_token_accuracy": 0.22904188930988312,
      "num_tokens": 89461077.0,
      "step": 40065
    },
    {
      "entropy": 5.29195122718811,
      "epoch": 4.500252709608581,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003145092042285446,
      "loss": 4.8393,
      "mean_token_accuracy": 0.22437268048524855,
      "num_tokens": 89472281.0,
      "step": 40070
    },
    {
      "entropy": 5.24699182510376,
      "epoch": 4.500814286516539,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003144696787023066,
      "loss": 4.7665,
      "mean_token_accuracy": 0.2361178755760193,
      "num_tokens": 89483030.0,
      "step": 40075
    },
    {
      "entropy": 5.149210262298584,
      "epoch": 4.501375863424496,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031443015191933864,
      "loss": 4.6876,
      "mean_token_accuracy": 0.2337690159678459,
      "num_tokens": 89493355.0,
      "step": 40080
    },
    {
      "entropy": 5.228713417053223,
      "epoch": 4.501937440332454,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031439062388089953,
      "loss": 4.8688,
      "mean_token_accuracy": 0.23244127333164216,
      "num_tokens": 89504253.0,
      "step": 40085
    },
    {
      "entropy": 5.179410982131958,
      "epoch": 4.502499017240411,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003143510945882475,
      "loss": 4.8247,
      "mean_token_accuracy": 0.2339414104819298,
      "num_tokens": 89514720.0,
      "step": 40090
    },
    {
      "entropy": 5.278791904449463,
      "epoch": 4.503060594148368,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00031431156404264143,
      "loss": 4.7962,
      "mean_token_accuracy": 0.23543696850538254,
      "num_tokens": 89526583.0,
      "step": 40095
    },
    {
      "entropy": 5.252575349807739,
      "epoch": 4.5036221710563265,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031427203224533996,
      "loss": 4.834,
      "mean_token_accuracy": 0.23138999193906784,
      "num_tokens": 89537791.0,
      "step": 40100
    },
    {
      "entropy": 5.318744945526123,
      "epoch": 4.504183747964284,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031423249919760167,
      "loss": 4.8697,
      "mean_token_accuracy": 0.22909336239099504,
      "num_tokens": 89549480.0,
      "step": 40105
    },
    {
      "entropy": 5.2543071746826175,
      "epoch": 4.504745324872241,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003141929649006854,
      "loss": 4.7768,
      "mean_token_accuracy": 0.23345431089401245,
      "num_tokens": 89559531.0,
      "step": 40110
    },
    {
      "entropy": 5.223902320861816,
      "epoch": 4.505306901780199,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031415342935585,
      "loss": 4.8187,
      "mean_token_accuracy": 0.22380415797233583,
      "num_tokens": 89570238.0,
      "step": 40115
    },
    {
      "entropy": 5.251343584060669,
      "epoch": 4.505868478688156,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003141138925643542,
      "loss": 4.8057,
      "mean_token_accuracy": 0.2362707018852234,
      "num_tokens": 89581300.0,
      "step": 40120
    },
    {
      "entropy": 5.309865808486938,
      "epoch": 4.506430055596114,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003140743545274569,
      "loss": 4.8588,
      "mean_token_accuracy": 0.23149511516094207,
      "num_tokens": 89592461.0,
      "step": 40125
    },
    {
      "entropy": 5.3240045547485355,
      "epoch": 4.5069916325040715,
      "grad_norm": 1.0,
      "learning_rate": 0.00031403481524641696,
      "loss": 4.941,
      "mean_token_accuracy": 0.2195713222026825,
      "num_tokens": 89603380.0,
      "step": 40130
    },
    {
      "entropy": 5.267281913757325,
      "epoch": 4.507553209412029,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031399527472249336,
      "loss": 4.8187,
      "mean_token_accuracy": 0.2318210706114769,
      "num_tokens": 89614268.0,
      "step": 40135
    },
    {
      "entropy": 5.298549365997315,
      "epoch": 4.508114786319987,
      "grad_norm": 1.09375,
      "learning_rate": 0.00031395573295694507,
      "loss": 4.8068,
      "mean_token_accuracy": 0.2278694987297058,
      "num_tokens": 89625480.0,
      "step": 40140
    },
    {
      "entropy": 5.285141468048096,
      "epoch": 4.508676363227944,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031391618995103116,
      "loss": 4.8488,
      "mean_token_accuracy": 0.2294372871518135,
      "num_tokens": 89637437.0,
      "step": 40145
    },
    {
      "entropy": 5.217415857315063,
      "epoch": 4.509237940135901,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003138766457060107,
      "loss": 4.7454,
      "mean_token_accuracy": 0.2354426473379135,
      "num_tokens": 89649196.0,
      "step": 40150
    },
    {
      "entropy": 5.1454620361328125,
      "epoch": 4.509799517043859,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003138371002231427,
      "loss": 4.7462,
      "mean_token_accuracy": 0.23386417031288148,
      "num_tokens": 89659536.0,
      "step": 40155
    },
    {
      "entropy": 5.268584585189819,
      "epoch": 4.510361093951817,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003137975535036864,
      "loss": 4.851,
      "mean_token_accuracy": 0.23435092717409134,
      "num_tokens": 89671021.0,
      "step": 40160
    },
    {
      "entropy": 5.221821403503418,
      "epoch": 4.510922670859774,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00031375800554890094,
      "loss": 4.714,
      "mean_token_accuracy": 0.23885829597711564,
      "num_tokens": 89681689.0,
      "step": 40165
    },
    {
      "entropy": 5.243422651290894,
      "epoch": 4.511484247767732,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031371845636004556,
      "loss": 4.7922,
      "mean_token_accuracy": 0.23504580855369567,
      "num_tokens": 89692353.0,
      "step": 40170
    },
    {
      "entropy": 5.205980062484741,
      "epoch": 4.512045824675689,
      "grad_norm": 1.125,
      "learning_rate": 0.0003136789059383795,
      "loss": 4.8173,
      "mean_token_accuracy": 0.23355720192193985,
      "num_tokens": 89702852.0,
      "step": 40175
    },
    {
      "entropy": 5.292448091506958,
      "epoch": 4.512607401583647,
      "grad_norm": 0.8984375,
      "learning_rate": 0.0003136393542851622,
      "loss": 4.8551,
      "mean_token_accuracy": 0.22892528623342515,
      "num_tokens": 89716125.0,
      "step": 40180
    },
    {
      "entropy": 5.20902681350708,
      "epoch": 4.5131689784916045,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031359980140165274,
      "loss": 4.7585,
      "mean_token_accuracy": 0.2330080524086952,
      "num_tokens": 89726611.0,
      "step": 40185
    },
    {
      "entropy": 5.251026773452759,
      "epoch": 4.513730555399562,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003135602472891106,
      "loss": 4.8866,
      "mean_token_accuracy": 0.23249602168798447,
      "num_tokens": 89738563.0,
      "step": 40190
    },
    {
      "entropy": 5.239123249053955,
      "epoch": 4.51429213230752,
      "grad_norm": 1.078125,
      "learning_rate": 0.00031352069194879527,
      "loss": 4.7419,
      "mean_token_accuracy": 0.23622345179319382,
      "num_tokens": 89748842.0,
      "step": 40195
    },
    {
      "entropy": 5.202417802810669,
      "epoch": 4.514853709215477,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003134811353819662,
      "loss": 4.7987,
      "mean_token_accuracy": 0.23927332311868668,
      "num_tokens": 89760747.0,
      "step": 40200
    },
    {
      "entropy": 5.179620838165283,
      "epoch": 4.515415286123434,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003134415775898828,
      "loss": 4.6877,
      "mean_token_accuracy": 0.23976208716630937,
      "num_tokens": 89773059.0,
      "step": 40205
    },
    {
      "entropy": 5.2446338653564455,
      "epoch": 4.515976863031392,
      "grad_norm": 1.125,
      "learning_rate": 0.0003134020185738047,
      "loss": 4.7495,
      "mean_token_accuracy": 0.241462442278862,
      "num_tokens": 89783347.0,
      "step": 40210
    },
    {
      "entropy": 5.241715669631958,
      "epoch": 4.51653843993935,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00031336245833499146,
      "loss": 4.8006,
      "mean_token_accuracy": 0.23645733147859574,
      "num_tokens": 89793094.0,
      "step": 40215
    },
    {
      "entropy": 5.24695520401001,
      "epoch": 4.517100016847307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003133228968747026,
      "loss": 4.7989,
      "mean_token_accuracy": 0.23175788521766663,
      "num_tokens": 89804591.0,
      "step": 40220
    },
    {
      "entropy": 5.316673278808594,
      "epoch": 4.517661593755265,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003132833341941979,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23559387475252153,
      "num_tokens": 89815749.0,
      "step": 40225
    },
    {
      "entropy": 5.296691370010376,
      "epoch": 4.518223170663222,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000313243770294737,
      "loss": 4.813,
      "mean_token_accuracy": 0.23083861768245698,
      "num_tokens": 89826940.0,
      "step": 40230
    },
    {
      "entropy": 5.304778432846069,
      "epoch": 4.51878474757118,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031320420517757955,
      "loss": 4.8732,
      "mean_token_accuracy": 0.23148408383131028,
      "num_tokens": 89838955.0,
      "step": 40235
    },
    {
      "entropy": 5.210058259963989,
      "epoch": 4.519346324479137,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003131646388439854,
      "loss": 4.8361,
      "mean_token_accuracy": 0.2265905663371086,
      "num_tokens": 89850177.0,
      "step": 40240
    },
    {
      "entropy": 5.216290807723999,
      "epoch": 4.519907901387095,
      "grad_norm": 1.0,
      "learning_rate": 0.0003131250712952144,
      "loss": 4.8058,
      "mean_token_accuracy": 0.22908548563718795,
      "num_tokens": 89861185.0,
      "step": 40245
    },
    {
      "entropy": 5.334852838516236,
      "epoch": 4.520469478295053,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031308550253252634,
      "loss": 4.8386,
      "mean_token_accuracy": 0.2345749944448471,
      "num_tokens": 89873000.0,
      "step": 40250
    },
    {
      "entropy": 5.258608341217041,
      "epoch": 4.52103105520301,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00031304593255718113,
      "loss": 4.7745,
      "mean_token_accuracy": 0.23121029138565063,
      "num_tokens": 89884223.0,
      "step": 40255
    },
    {
      "entropy": 5.302819347381591,
      "epoch": 4.521592632110968,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031300636137043857,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23084105551242828,
      "num_tokens": 89895655.0,
      "step": 40260
    },
    {
      "entropy": 5.211422777175903,
      "epoch": 4.522154209018925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031296678897355877,
      "loss": 4.7337,
      "mean_token_accuracy": 0.23338425159454346,
      "num_tokens": 89906475.0,
      "step": 40265
    },
    {
      "entropy": 5.207676982879638,
      "epoch": 4.5227157859268825,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003129272153678017,
      "loss": 4.8215,
      "mean_token_accuracy": 0.2360489100217819,
      "num_tokens": 89917523.0,
      "step": 40270
    },
    {
      "entropy": 5.114949655532837,
      "epoch": 4.523277362834841,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003128876405544274,
      "loss": 4.6571,
      "mean_token_accuracy": 0.23841649740934373,
      "num_tokens": 89928106.0,
      "step": 40275
    },
    {
      "entropy": 5.159280490875244,
      "epoch": 4.523838939742798,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031284806453469596,
      "loss": 4.7141,
      "mean_token_accuracy": 0.241891048848629,
      "num_tokens": 89938638.0,
      "step": 40280
    },
    {
      "entropy": 5.228719758987427,
      "epoch": 4.524400516650755,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031280848730986735,
      "loss": 4.7755,
      "mean_token_accuracy": 0.24063454568386078,
      "num_tokens": 89949760.0,
      "step": 40285
    },
    {
      "entropy": 5.32233247756958,
      "epoch": 4.524962093558713,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003127689088812019,
      "loss": 4.8993,
      "mean_token_accuracy": 0.22990622669458388,
      "num_tokens": 89960578.0,
      "step": 40290
    },
    {
      "entropy": 5.170852613449097,
      "epoch": 4.52552367046667,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003127293292499598,
      "loss": 4.7511,
      "mean_token_accuracy": 0.2395279049873352,
      "num_tokens": 89971370.0,
      "step": 40295
    },
    {
      "entropy": 5.205481433868409,
      "epoch": 4.526085247374628,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003126897484174012,
      "loss": 4.7069,
      "mean_token_accuracy": 0.23690689355134964,
      "num_tokens": 89982068.0,
      "step": 40300
    },
    {
      "entropy": 5.218219232559204,
      "epoch": 4.526646824282586,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003126501663847863,
      "loss": 4.7189,
      "mean_token_accuracy": 0.23761225193738938,
      "num_tokens": 89993108.0,
      "step": 40305
    },
    {
      "entropy": 5.238379669189453,
      "epoch": 4.527208401190543,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003126105831533756,
      "loss": 4.8077,
      "mean_token_accuracy": 0.2253062903881073,
      "num_tokens": 90004604.0,
      "step": 40310
    },
    {
      "entropy": 5.287255477905274,
      "epoch": 4.527769978098501,
      "grad_norm": 1.171875,
      "learning_rate": 0.0003125709987244293,
      "loss": 4.8798,
      "mean_token_accuracy": 0.22984106093645096,
      "num_tokens": 90015241.0,
      "step": 40315
    },
    {
      "entropy": 5.124442720413208,
      "epoch": 4.528331555006458,
      "grad_norm": 0.96875,
      "learning_rate": 0.00031253141309920784,
      "loss": 4.6799,
      "mean_token_accuracy": 0.24212672114372252,
      "num_tokens": 90026511.0,
      "step": 40320
    },
    {
      "entropy": 5.1374969482421875,
      "epoch": 4.5288931319144154,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031249182627897165,
      "loss": 4.6641,
      "mean_token_accuracy": 0.24276194125413894,
      "num_tokens": 90038994.0,
      "step": 40325
    },
    {
      "entropy": 5.24053807258606,
      "epoch": 4.5294547088223736,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00031245223826498116,
      "loss": 4.8199,
      "mean_token_accuracy": 0.22620220333337784,
      "num_tokens": 90049183.0,
      "step": 40330
    },
    {
      "entropy": 5.220356750488281,
      "epoch": 4.530016285730331,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031241264905849687,
      "loss": 4.764,
      "mean_token_accuracy": 0.24041813015937805,
      "num_tokens": 90060642.0,
      "step": 40335
    },
    {
      "entropy": 5.258073472976685,
      "epoch": 4.530577862638288,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031237305866077926,
      "loss": 4.8264,
      "mean_token_accuracy": 0.2299100548028946,
      "num_tokens": 90069709.0,
      "step": 40340
    },
    {
      "entropy": 5.232828283309937,
      "epoch": 4.531139439546246,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031233346707308906,
      "loss": 4.7691,
      "mean_token_accuracy": 0.2348434403538704,
      "num_tokens": 90080219.0,
      "step": 40345
    },
    {
      "entropy": 5.288220930099487,
      "epoch": 4.531701016454203,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031229387429668675,
      "loss": 4.898,
      "mean_token_accuracy": 0.22466425597667694,
      "num_tokens": 90091769.0,
      "step": 40350
    },
    {
      "entropy": 5.294201612472534,
      "epoch": 4.5322625933621605,
      "grad_norm": 0.9375,
      "learning_rate": 0.00031225428033283305,
      "loss": 4.7861,
      "mean_token_accuracy": 0.23592448532581328,
      "num_tokens": 90102273.0,
      "step": 40355
    },
    {
      "entropy": 5.217380428314209,
      "epoch": 4.532824170270119,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003122146851827886,
      "loss": 4.7753,
      "mean_token_accuracy": 0.2292950063943863,
      "num_tokens": 90112311.0,
      "step": 40360
    },
    {
      "entropy": 5.128723621368408,
      "epoch": 4.533385747178076,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031217508884781413,
      "loss": 4.6995,
      "mean_token_accuracy": 0.2371765196323395,
      "num_tokens": 90123361.0,
      "step": 40365
    },
    {
      "entropy": 5.160928058624267,
      "epoch": 4.533947324086034,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003121354913291704,
      "loss": 4.6759,
      "mean_token_accuracy": 0.2380952790379524,
      "num_tokens": 90134811.0,
      "step": 40370
    },
    {
      "entropy": 5.290014028549194,
      "epoch": 4.534508900993991,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003120958926281182,
      "loss": 4.8261,
      "mean_token_accuracy": 0.23297327607870102,
      "num_tokens": 90145218.0,
      "step": 40375
    },
    {
      "entropy": 5.2760358333587645,
      "epoch": 4.535070477901948,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00031205629274591844,
      "loss": 4.8727,
      "mean_token_accuracy": 0.23686936497688293,
      "num_tokens": 90157217.0,
      "step": 40380
    },
    {
      "entropy": 5.2353699684143065,
      "epoch": 4.5356320548099065,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00031201669168383193,
      "loss": 4.7939,
      "mean_token_accuracy": 0.2336752489209175,
      "num_tokens": 90167752.0,
      "step": 40385
    },
    {
      "entropy": 5.194595575332642,
      "epoch": 4.536193631717864,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00031197708944311964,
      "loss": 4.7699,
      "mean_token_accuracy": 0.24269434064626694,
      "num_tokens": 90178309.0,
      "step": 40390
    },
    {
      "entropy": 5.170888662338257,
      "epoch": 4.536755208625821,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003119374860250425,
      "loss": 4.6999,
      "mean_token_accuracy": 0.24698664397001266,
      "num_tokens": 90189158.0,
      "step": 40395
    },
    {
      "entropy": 5.252411413192749,
      "epoch": 4.537316785533779,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003118978814308614,
      "loss": 4.7736,
      "mean_token_accuracy": 0.2370266631245613,
      "num_tokens": 90199611.0,
      "step": 40400
    },
    {
      "entropy": 5.288829755783081,
      "epoch": 4.537878362441736,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003118582756618375,
      "loss": 4.8313,
      "mean_token_accuracy": 0.22830185294151306,
      "num_tokens": 90210852.0,
      "step": 40405
    },
    {
      "entropy": 5.2542712688446045,
      "epoch": 4.5384399393496935,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003118186687192318,
      "loss": 4.804,
      "mean_token_accuracy": 0.22906601727008818,
      "num_tokens": 90221193.0,
      "step": 40410
    },
    {
      "entropy": 5.211963701248169,
      "epoch": 4.539001516257652,
      "grad_norm": 1.125,
      "learning_rate": 0.00031177906060430536,
      "loss": 4.8194,
      "mean_token_accuracy": 0.22848130613565446,
      "num_tokens": 90232811.0,
      "step": 40415
    },
    {
      "entropy": 5.259741687774659,
      "epoch": 4.539563093165609,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031173945131831936,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23030529916286469,
      "num_tokens": 90243566.0,
      "step": 40420
    },
    {
      "entropy": 5.2946861743927,
      "epoch": 4.540124670073567,
      "grad_norm": 1.125,
      "learning_rate": 0.000311699840862535,
      "loss": 4.8237,
      "mean_token_accuracy": 0.24042216092348098,
      "num_tokens": 90255292.0,
      "step": 40425
    },
    {
      "entropy": 5.243488931655884,
      "epoch": 4.540686246981524,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003116602292382135,
      "loss": 4.8288,
      "mean_token_accuracy": 0.2326616570353508,
      "num_tokens": 90266691.0,
      "step": 40430
    },
    {
      "entropy": 5.207397747039795,
      "epoch": 4.541247823889481,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00031162061644661606,
      "loss": 4.8079,
      "mean_token_accuracy": 0.24029551595449447,
      "num_tokens": 90278240.0,
      "step": 40435
    },
    {
      "entropy": 5.288090419769287,
      "epoch": 4.5418094007974394,
      "grad_norm": 1.015625,
      "learning_rate": 0.000311581002489004,
      "loss": 4.8405,
      "mean_token_accuracy": 0.23362436890602112,
      "num_tokens": 90288203.0,
      "step": 40440
    },
    {
      "entropy": 5.25248384475708,
      "epoch": 4.542370977705397,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003115413873666387,
      "loss": 4.8339,
      "mean_token_accuracy": 0.23512151092290878,
      "num_tokens": 90299326.0,
      "step": 40445
    },
    {
      "entropy": 5.221164608001709,
      "epoch": 4.542932554613355,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003115017710807813,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24006936550140381,
      "num_tokens": 90309504.0,
      "step": 40450
    },
    {
      "entropy": 5.1825908660888675,
      "epoch": 4.543494131521312,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003114621536326934,
      "loss": 4.7781,
      "mean_token_accuracy": 0.23732460290193558,
      "num_tokens": 90320919.0,
      "step": 40455
    },
    {
      "entropy": 5.175068140029907,
      "epoch": 4.544055708429269,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003114225350236365,
      "loss": 4.6817,
      "mean_token_accuracy": 0.24480066001415252,
      "num_tokens": 90331864.0,
      "step": 40460
    },
    {
      "entropy": 5.202399873733521,
      "epoch": 4.544617285337227,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003113829152548719,
      "loss": 4.7376,
      "mean_token_accuracy": 0.2347819685935974,
      "num_tokens": 90342563.0,
      "step": 40465
    },
    {
      "entropy": 5.193118524551392,
      "epoch": 4.5451788622451845,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003113432943276611,
      "loss": 4.7031,
      "mean_token_accuracy": 0.23407386541366576,
      "num_tokens": 90353464.0,
      "step": 40470
    },
    {
      "entropy": 5.202253293991089,
      "epoch": 4.545740439153142,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031130367224326585,
      "loss": 4.7712,
      "mean_token_accuracy": 0.23510017842054368,
      "num_tokens": 90363255.0,
      "step": 40475
    },
    {
      "entropy": 5.300638484954834,
      "epoch": 4.5463020160611,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003112640490029475,
      "loss": 4.8632,
      "mean_token_accuracy": 0.23049821853637695,
      "num_tokens": 90374478.0,
      "step": 40480
    },
    {
      "entropy": 5.279642534255982,
      "epoch": 4.546863592969057,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031122442460796774,
      "loss": 4.8232,
      "mean_token_accuracy": 0.2377368316054344,
      "num_tokens": 90386916.0,
      "step": 40485
    },
    {
      "entropy": 5.176132869720459,
      "epoch": 4.547425169877014,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003111847990595883,
      "loss": 4.6856,
      "mean_token_accuracy": 0.23516541868448257,
      "num_tokens": 90397407.0,
      "step": 40490
    },
    {
      "entropy": 5.225453090667725,
      "epoch": 4.547986746784972,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003111451723590708,
      "loss": 4.8029,
      "mean_token_accuracy": 0.22882143557071685,
      "num_tokens": 90408762.0,
      "step": 40495
    },
    {
      "entropy": 5.229246616363525,
      "epoch": 4.54854832369293,
      "grad_norm": 1.140625,
      "learning_rate": 0.000311105544507677,
      "loss": 4.85,
      "mean_token_accuracy": 0.22724255323410034,
      "num_tokens": 90418942.0,
      "step": 40500
    },
    {
      "entropy": 5.357480001449585,
      "epoch": 4.549109900600888,
      "grad_norm": 1.03125,
      "learning_rate": 0.00031106591550666865,
      "loss": 4.9514,
      "mean_token_accuracy": 0.22323980927467346,
      "num_tokens": 90430052.0,
      "step": 40505
    },
    {
      "entropy": 5.342213344573975,
      "epoch": 4.549671477508845,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003110262853573076,
      "loss": 4.9032,
      "mean_token_accuracy": 0.2297995164990425,
      "num_tokens": 90442631.0,
      "step": 40510
    },
    {
      "entropy": 5.257503318786621,
      "epoch": 4.550233054416802,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003109866540608556,
      "loss": 4.8131,
      "mean_token_accuracy": 0.23831469267606736,
      "num_tokens": 90454429.0,
      "step": 40515
    },
    {
      "entropy": 5.241280746459961,
      "epoch": 4.55079463132476,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031094702161857455,
      "loss": 4.8797,
      "mean_token_accuracy": 0.2252253159880638,
      "num_tokens": 90466367.0,
      "step": 40520
    },
    {
      "entropy": 5.271191930770874,
      "epoch": 4.5513562082327175,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031090738803172646,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23921174108982085,
      "num_tokens": 90478480.0,
      "step": 40525
    },
    {
      "entropy": 5.218852519989014,
      "epoch": 4.551917785140675,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031086775330157324,
      "loss": 4.7873,
      "mean_token_accuracy": 0.2220443680882454,
      "num_tokens": 90490501.0,
      "step": 40530
    },
    {
      "entropy": 5.198581695556641,
      "epoch": 4.552479362048633,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003108281174293768,
      "loss": 4.7567,
      "mean_token_accuracy": 0.23488612920045854,
      "num_tokens": 90500647.0,
      "step": 40535
    },
    {
      "entropy": 5.153819179534912,
      "epoch": 4.55304093895659,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00031078848041639926,
      "loss": 4.6991,
      "mean_token_accuracy": 0.24152608960866928,
      "num_tokens": 90510963.0,
      "step": 40540
    },
    {
      "entropy": 5.2715339183807375,
      "epoch": 4.553602515864547,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00031074884226390246,
      "loss": 4.8426,
      "mean_token_accuracy": 0.2318502336740494,
      "num_tokens": 90523475.0,
      "step": 40545
    },
    {
      "entropy": 5.257555294036865,
      "epoch": 4.554164092772505,
      "grad_norm": 0.984375,
      "learning_rate": 0.0003107092029731488,
      "loss": 4.7828,
      "mean_token_accuracy": 0.23365146815776824,
      "num_tokens": 90535348.0,
      "step": 40550
    },
    {
      "entropy": 5.178422641754151,
      "epoch": 4.5547256696804626,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00031066956254540025,
      "loss": 4.7297,
      "mean_token_accuracy": 0.234313003718853,
      "num_tokens": 90545841.0,
      "step": 40555
    },
    {
      "entropy": 5.225968885421753,
      "epoch": 4.555287246588421,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00031062992098191897,
      "loss": 4.8156,
      "mean_token_accuracy": 0.229535511136055,
      "num_tokens": 90557221.0,
      "step": 40560
    },
    {
      "entropy": 5.1517575740814205,
      "epoch": 4.555848823496378,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003105902782839672,
      "loss": 4.6592,
      "mean_token_accuracy": 0.24434657543897628,
      "num_tokens": 90568080.0,
      "step": 40565
    },
    {
      "entropy": 5.265501689910889,
      "epoch": 4.556410400404335,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003105506344528071,
      "loss": 4.7898,
      "mean_token_accuracy": 0.23521328568458558,
      "num_tokens": 90579075.0,
      "step": 40570
    },
    {
      "entropy": 5.256441068649292,
      "epoch": 4.556971977312293,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000310510989489701,
      "loss": 4.8715,
      "mean_token_accuracy": 0.23497994095087052,
      "num_tokens": 90589937.0,
      "step": 40575
    },
    {
      "entropy": 5.183305978775024,
      "epoch": 4.55753355422025,
      "grad_norm": 1.015625,
      "learning_rate": 0.00031047134339591136,
      "loss": 4.7547,
      "mean_token_accuracy": 0.23810143768787384,
      "num_tokens": 90600262.0,
      "step": 40580
    },
    {
      "entropy": 5.296771812438965,
      "epoch": 4.558095131128208,
      "grad_norm": 0.984375,
      "learning_rate": 0.00031043169617270033,
      "loss": 4.874,
      "mean_token_accuracy": 0.23106696158647538,
      "num_tokens": 90612161.0,
      "step": 40585
    },
    {
      "entropy": 5.311317873001099,
      "epoch": 4.558656708036166,
      "grad_norm": 0.953125,
      "learning_rate": 0.00031039204782133033,
      "loss": 4.836,
      "mean_token_accuracy": 0.22691119462251663,
      "num_tokens": 90624304.0,
      "step": 40590
    },
    {
      "entropy": 5.214407968521118,
      "epoch": 4.559218284944123,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003103523983430638,
      "loss": 4.7574,
      "mean_token_accuracy": 0.22954113632440568,
      "num_tokens": 90636196.0,
      "step": 40595
    },
    {
      "entropy": 5.185132265090942,
      "epoch": 4.55977986185208,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00031031274773916326,
      "loss": 4.7706,
      "mean_token_accuracy": 0.23733463138341904,
      "num_tokens": 90649024.0,
      "step": 40600
    },
    {
      "entropy": 5.219745206832886,
      "epoch": 4.560341438760038,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00031027309601089104,
      "loss": 4.7711,
      "mean_token_accuracy": 0.22564496845006943,
      "num_tokens": 90661110.0,
      "step": 40605
    },
    {
      "entropy": 5.22375807762146,
      "epoch": 4.5609030156679955,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003102334431595098,
      "loss": 4.7714,
      "mean_token_accuracy": 0.23460846543312072,
      "num_tokens": 90671691.0,
      "step": 40610
    },
    {
      "entropy": 5.25214991569519,
      "epoch": 4.561464592575954,
      "grad_norm": 1.046875,
      "learning_rate": 0.00031019378918628203,
      "loss": 4.7624,
      "mean_token_accuracy": 0.23820443600416183,
      "num_tokens": 90683147.0,
      "step": 40615
    },
    {
      "entropy": 5.255657482147217,
      "epoch": 4.562026169483911,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00031015413409247046,
      "loss": 4.8217,
      "mean_token_accuracy": 0.2329704686999321,
      "num_tokens": 90694614.0,
      "step": 40620
    },
    {
      "entropy": 5.219903230667114,
      "epoch": 4.562587746391868,
      "grad_norm": 1.0625,
      "learning_rate": 0.00031011447787933756,
      "loss": 4.876,
      "mean_token_accuracy": 0.22858633399009703,
      "num_tokens": 90705326.0,
      "step": 40625
    },
    {
      "entropy": 5.270701169967651,
      "epoch": 4.563149323299826,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003100748205481461,
      "loss": 4.8095,
      "mean_token_accuracy": 0.22896412163972854,
      "num_tokens": 90716140.0,
      "step": 40630
    },
    {
      "entropy": 5.114110374450684,
      "epoch": 4.563710900207783,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003100351621001587,
      "loss": 4.603,
      "mean_token_accuracy": 0.25316258817911147,
      "num_tokens": 90727181.0,
      "step": 40635
    },
    {
      "entropy": 5.17583999633789,
      "epoch": 4.5642724771157415,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003099955025366382,
      "loss": 4.8031,
      "mean_token_accuracy": 0.2368406116962433,
      "num_tokens": 90737931.0,
      "step": 40640
    },
    {
      "entropy": 5.204661560058594,
      "epoch": 4.564834054023699,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030995584185884733,
      "loss": 4.8189,
      "mean_token_accuracy": 0.2311497688293457,
      "num_tokens": 90749011.0,
      "step": 40645
    },
    {
      "entropy": 5.22731294631958,
      "epoch": 4.565395630931656,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003099161800680488,
      "loss": 4.7596,
      "mean_token_accuracy": 0.22808935940265657,
      "num_tokens": 90760413.0,
      "step": 40650
    },
    {
      "entropy": 5.195943307876587,
      "epoch": 4.565957207839614,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003098765171655056,
      "loss": 4.8286,
      "mean_token_accuracy": 0.23043037056922913,
      "num_tokens": 90770973.0,
      "step": 40655
    },
    {
      "entropy": 5.262861919403076,
      "epoch": 4.566518784747571,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003098368531524806,
      "loss": 4.7977,
      "mean_token_accuracy": 0.2370335191488266,
      "num_tokens": 90782204.0,
      "step": 40660
    },
    {
      "entropy": 5.259708595275879,
      "epoch": 4.567080361655528,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003097971880302367,
      "loss": 4.7711,
      "mean_token_accuracy": 0.24045614898204803,
      "num_tokens": 90791906.0,
      "step": 40665
    },
    {
      "entropy": 5.186523199081421,
      "epoch": 4.5676419385634865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030975752180003674,
      "loss": 4.6987,
      "mean_token_accuracy": 0.2417054533958435,
      "num_tokens": 90802153.0,
      "step": 40670
    },
    {
      "entropy": 5.166971683502197,
      "epoch": 4.568203515471444,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000309717854463144,
      "loss": 4.7573,
      "mean_token_accuracy": 0.23512990772724152,
      "num_tokens": 90813129.0,
      "step": 40675
    },
    {
      "entropy": 5.29270749092102,
      "epoch": 4.568765092379401,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0003096781860208211,
      "loss": 4.8379,
      "mean_token_accuracy": 0.23040347993373872,
      "num_tokens": 90823445.0,
      "step": 40680
    },
    {
      "entropy": 5.2290245532989506,
      "epoch": 4.569326669287359,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003096385164743314,
      "loss": 4.7834,
      "mean_token_accuracy": 0.2309352770447731,
      "num_tokens": 90834823.0,
      "step": 40685
    },
    {
      "entropy": 5.118114328384399,
      "epoch": 4.569888246195316,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00030959884582493793,
      "loss": 4.6544,
      "mean_token_accuracy": 0.24265053868293762,
      "num_tokens": 90845503.0,
      "step": 40690
    },
    {
      "entropy": 5.288988018035889,
      "epoch": 4.570449823103274,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003095591740739038,
      "loss": 4.8501,
      "mean_token_accuracy": 0.2326929658651352,
      "num_tokens": 90857562.0,
      "step": 40695
    },
    {
      "entropy": 5.276044034957886,
      "epoch": 4.571011400011232,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003095195012224921,
      "loss": 4.866,
      "mean_token_accuracy": 0.23015909940004348,
      "num_tokens": 90868207.0,
      "step": 40700
    },
    {
      "entropy": 5.149024438858032,
      "epoch": 4.571572976919189,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003094798272719661,
      "loss": 4.7668,
      "mean_token_accuracy": 0.23166948556900024,
      "num_tokens": 90879710.0,
      "step": 40705
    },
    {
      "entropy": 5.244139957427978,
      "epoch": 4.572134553827147,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030944015222358906,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23475077450275422,
      "num_tokens": 90890702.0,
      "step": 40710
    },
    {
      "entropy": 5.233556079864502,
      "epoch": 4.572696130735104,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030940047607862415,
      "loss": 4.7254,
      "mean_token_accuracy": 0.2373086020350456,
      "num_tokens": 90902171.0,
      "step": 40715
    },
    {
      "entropy": 5.232194614410401,
      "epoch": 4.573257707643061,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003093607988383348,
      "loss": 4.7622,
      "mean_token_accuracy": 0.2354014590382576,
      "num_tokens": 90913735.0,
      "step": 40720
    },
    {
      "entropy": 5.180649137496948,
      "epoch": 4.5738192845510195,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003093211205039842,
      "loss": 4.7865,
      "mean_token_accuracy": 0.23694392591714858,
      "num_tokens": 90924137.0,
      "step": 40725
    },
    {
      "entropy": 5.143114328384399,
      "epoch": 4.574380861458977,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003092814410768359,
      "loss": 4.7462,
      "mean_token_accuracy": 0.2424958661198616,
      "num_tokens": 90935214.0,
      "step": 40730
    },
    {
      "entropy": 5.205068922042846,
      "epoch": 4.574942438366934,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030924176055815327,
      "loss": 4.7369,
      "mean_token_accuracy": 0.23840753138065338,
      "num_tokens": 90946797.0,
      "step": 40735
    },
    {
      "entropy": 5.223412084579468,
      "epoch": 4.575504015274892,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00030920207894919957,
      "loss": 4.7716,
      "mean_token_accuracy": 0.23694978803396224,
      "num_tokens": 90958055.0,
      "step": 40740
    },
    {
      "entropy": 5.159584379196167,
      "epoch": 4.576065592182849,
      "grad_norm": 1.0,
      "learning_rate": 0.0003091623962512385,
      "loss": 4.7434,
      "mean_token_accuracy": 0.23725191205739976,
      "num_tokens": 90969516.0,
      "step": 40745
    },
    {
      "entropy": 5.131197929382324,
      "epoch": 4.576627169090807,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030912271246553346,
      "loss": 4.7456,
      "mean_token_accuracy": 0.234658944606781,
      "num_tokens": 90980536.0,
      "step": 40750
    },
    {
      "entropy": 5.222881841659546,
      "epoch": 4.577188745998765,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000309083027593348,
      "loss": 4.8345,
      "mean_token_accuracy": 0.23921570032835007,
      "num_tokens": 90992150.0,
      "step": 40755
    },
    {
      "entropy": 5.237351560592652,
      "epoch": 4.577750322906722,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00030904334163594577,
      "loss": 4.7939,
      "mean_token_accuracy": 0.23203352093696594,
      "num_tokens": 91003323.0,
      "step": 40760
    },
    {
      "entropy": 5.192285013198853,
      "epoch": 4.57831189981468,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003090036545945904,
      "loss": 4.6788,
      "mean_token_accuracy": 0.24268206357955932,
      "num_tokens": 91014238.0,
      "step": 40765
    },
    {
      "entropy": 5.198546457290649,
      "epoch": 4.578873476722637,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030896396647054535,
      "loss": 4.8001,
      "mean_token_accuracy": 0.22982438057661056,
      "num_tokens": 91025375.0,
      "step": 40770
    },
    {
      "entropy": 5.195534515380859,
      "epoch": 4.579435053630594,
      "grad_norm": 1.0,
      "learning_rate": 0.0003089242772650744,
      "loss": 4.8351,
      "mean_token_accuracy": 0.2294978305697441,
      "num_tokens": 91036498.0,
      "step": 40775
    },
    {
      "entropy": 5.191793060302734,
      "epoch": 4.579996630538552,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030888458697944144,
      "loss": 4.7495,
      "mean_token_accuracy": 0.24096686989068986,
      "num_tokens": 91046909.0,
      "step": 40780
    },
    {
      "entropy": 5.295937490463257,
      "epoch": 4.58055820744651,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030884489561491005,
      "loss": 4.8737,
      "mean_token_accuracy": 0.23123259097337723,
      "num_tokens": 91056052.0,
      "step": 40785
    },
    {
      "entropy": 5.399011087417603,
      "epoch": 4.581119784354467,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000308805203172744,
      "loss": 4.9788,
      "mean_token_accuracy": 0.2151770070195198,
      "num_tokens": 91067614.0,
      "step": 40790
    },
    {
      "entropy": 5.173900175094604,
      "epoch": 4.581681361262425,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030876550965420725,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24182571321725846,
      "num_tokens": 91078539.0,
      "step": 40795
    },
    {
      "entropy": 5.200805377960205,
      "epoch": 4.582242938170382,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003087258150605635,
      "loss": 4.7782,
      "mean_token_accuracy": 0.2357066109776497,
      "num_tokens": 91088909.0,
      "step": 40800
    },
    {
      "entropy": 5.201543188095092,
      "epoch": 4.58280451507834,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003086861193930767,
      "loss": 4.8041,
      "mean_token_accuracy": 0.23493411839008332,
      "num_tokens": 91101073.0,
      "step": 40805
    },
    {
      "entropy": 5.264812040328979,
      "epoch": 4.5833660919862975,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003086464226530109,
      "loss": 4.7607,
      "mean_token_accuracy": 0.2351444259285927,
      "num_tokens": 91113518.0,
      "step": 40810
    },
    {
      "entropy": 5.270521116256714,
      "epoch": 4.583927668894255,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030860672484162987,
      "loss": 4.7627,
      "mean_token_accuracy": 0.2363172084093094,
      "num_tokens": 91124287.0,
      "step": 40815
    },
    {
      "entropy": 5.276182556152344,
      "epoch": 4.584489245802213,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003085670259601977,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23395420461893082,
      "num_tokens": 91134828.0,
      "step": 40820
    },
    {
      "entropy": 5.260522031784058,
      "epoch": 4.58505082271017,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030852732600997846,
      "loss": 4.8543,
      "mean_token_accuracy": 0.22469398379325867,
      "num_tokens": 91146647.0,
      "step": 40825
    },
    {
      "entropy": 5.214830636978149,
      "epoch": 4.585612399618128,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030848762499223616,
      "loss": 4.8449,
      "mean_token_accuracy": 0.2339801236987114,
      "num_tokens": 91156769.0,
      "step": 40830
    },
    {
      "entropy": 5.25368595123291,
      "epoch": 4.586173976526085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003084479229082348,
      "loss": 4.731,
      "mean_token_accuracy": 0.2395048752427101,
      "num_tokens": 91168876.0,
      "step": 40835
    },
    {
      "entropy": 5.33773455619812,
      "epoch": 4.586735553434043,
      "grad_norm": 1.15625,
      "learning_rate": 0.0003084082197592387,
      "loss": 4.9145,
      "mean_token_accuracy": 0.23041679114103317,
      "num_tokens": 91178587.0,
      "step": 40840
    },
    {
      "entropy": 5.18848729133606,
      "epoch": 4.587297130342001,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030836851554651184,
      "loss": 4.6709,
      "mean_token_accuracy": 0.24821729511022567,
      "num_tokens": 91190210.0,
      "step": 40845
    },
    {
      "entropy": 5.269397497177124,
      "epoch": 4.587858707249958,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030832881027131854,
      "loss": 4.8046,
      "mean_token_accuracy": 0.2358509436249733,
      "num_tokens": 91200921.0,
      "step": 40850
    },
    {
      "entropy": 5.182694673538208,
      "epoch": 4.588420284157915,
      "grad_norm": 1.0,
      "learning_rate": 0.00030828910393492306,
      "loss": 4.7318,
      "mean_token_accuracy": 0.24271580278873445,
      "num_tokens": 91211553.0,
      "step": 40855
    },
    {
      "entropy": 5.247672176361084,
      "epoch": 4.588981861065873,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030824939653858953,
      "loss": 4.807,
      "mean_token_accuracy": 0.2310768574476242,
      "num_tokens": 91222224.0,
      "step": 40860
    },
    {
      "entropy": 5.241736698150635,
      "epoch": 4.5895434379738305,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030820968808358223,
      "loss": 4.7615,
      "mean_token_accuracy": 0.23063301891088486,
      "num_tokens": 91233645.0,
      "step": 40865
    },
    {
      "entropy": 5.2652441501617435,
      "epoch": 4.590105014881788,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030816997857116573,
      "loss": 4.796,
      "mean_token_accuracy": 0.22974125146865845,
      "num_tokens": 91245194.0,
      "step": 40870
    },
    {
      "entropy": 5.208451843261718,
      "epoch": 4.590666591789746,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003081302680026042,
      "loss": 4.7421,
      "mean_token_accuracy": 0.23691069334745407,
      "num_tokens": 91255124.0,
      "step": 40875
    },
    {
      "entropy": 5.275165653228759,
      "epoch": 4.591228168697703,
      "grad_norm": 1.078125,
      "learning_rate": 0.000308090556379162,
      "loss": 4.897,
      "mean_token_accuracy": 0.22506390810012816,
      "num_tokens": 91267129.0,
      "step": 40880
    },
    {
      "entropy": 5.240679740905762,
      "epoch": 4.591789745605661,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030805084370210377,
      "loss": 4.8066,
      "mean_token_accuracy": 0.23099526017904282,
      "num_tokens": 91277614.0,
      "step": 40885
    },
    {
      "entropy": 5.242807531356812,
      "epoch": 4.592351322513618,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030801112997269386,
      "loss": 4.7758,
      "mean_token_accuracy": 0.23015497773885726,
      "num_tokens": 91289856.0,
      "step": 40890
    },
    {
      "entropy": 5.1857222557067875,
      "epoch": 4.5929128994215755,
      "grad_norm": 1.140625,
      "learning_rate": 0.0003079714151921967,
      "loss": 4.8032,
      "mean_token_accuracy": 0.2365572154521942,
      "num_tokens": 91300593.0,
      "step": 40895
    },
    {
      "entropy": 5.234616756439209,
      "epoch": 4.593474476329534,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030793169936187695,
      "loss": 4.7161,
      "mean_token_accuracy": 0.2427835077047348,
      "num_tokens": 91312934.0,
      "step": 40900
    },
    {
      "entropy": 5.194646120071411,
      "epoch": 4.594036053237491,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030789198248299906,
      "loss": 4.7507,
      "mean_token_accuracy": 0.2361937001347542,
      "num_tokens": 91323856.0,
      "step": 40905
    },
    {
      "entropy": 5.245959949493408,
      "epoch": 4.594597630145448,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030785226455682773,
      "loss": 4.8135,
      "mean_token_accuracy": 0.23388624489307402,
      "num_tokens": 91334545.0,
      "step": 40910
    },
    {
      "entropy": 5.2818286418914795,
      "epoch": 4.595159207053406,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003078125455846276,
      "loss": 4.7777,
      "mean_token_accuracy": 0.23260171413421632,
      "num_tokens": 91346006.0,
      "step": 40915
    },
    {
      "entropy": 5.249454832077026,
      "epoch": 4.595720783961363,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030777282556766316,
      "loss": 4.8204,
      "mean_token_accuracy": 0.2241940751671791,
      "num_tokens": 91356710.0,
      "step": 40920
    },
    {
      "entropy": 5.205613327026367,
      "epoch": 4.596282360869321,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030773310450719937,
      "loss": 4.7177,
      "mean_token_accuracy": 0.2373175159096718,
      "num_tokens": 91367747.0,
      "step": 40925
    },
    {
      "entropy": 5.2817662239074705,
      "epoch": 4.596843937777279,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003076933824045008,
      "loss": 4.8353,
      "mean_token_accuracy": 0.2332752004265785,
      "num_tokens": 91379079.0,
      "step": 40930
    },
    {
      "entropy": 5.272145175933838,
      "epoch": 4.597405514685236,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030765365926083214,
      "loss": 4.8652,
      "mean_token_accuracy": 0.23049555867910385,
      "num_tokens": 91390264.0,
      "step": 40935
    },
    {
      "entropy": 5.228577661514282,
      "epoch": 4.597967091593194,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003076139350774584,
      "loss": 4.7269,
      "mean_token_accuracy": 0.24029993861913682,
      "num_tokens": 91401322.0,
      "step": 40940
    },
    {
      "entropy": 5.202799654006958,
      "epoch": 4.598528668501151,
      "grad_norm": 1.0,
      "learning_rate": 0.0003075742098556444,
      "loss": 4.7859,
      "mean_token_accuracy": 0.23176563680171966,
      "num_tokens": 91412277.0,
      "step": 40945
    },
    {
      "entropy": 5.179635047912598,
      "epoch": 4.5990902454091085,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003075344835966548,
      "loss": 4.7186,
      "mean_token_accuracy": 0.23856626600027084,
      "num_tokens": 91422941.0,
      "step": 40950
    },
    {
      "entropy": 5.210261726379395,
      "epoch": 4.599651822317067,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003074947563017547,
      "loss": 4.7552,
      "mean_token_accuracy": 0.24348961114883422,
      "num_tokens": 91434194.0,
      "step": 40955
    },
    {
      "entropy": 5.18354115486145,
      "epoch": 4.600213399225024,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003074550279722089,
      "loss": 4.7408,
      "mean_token_accuracy": 0.233338825404644,
      "num_tokens": 91445330.0,
      "step": 40960
    },
    {
      "entropy": 5.265940380096436,
      "epoch": 4.600774976132981,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003074152986092825,
      "loss": 4.7944,
      "mean_token_accuracy": 0.2333725541830063,
      "num_tokens": 91456762.0,
      "step": 40965
    },
    {
      "entropy": 5.19476203918457,
      "epoch": 4.601336553040939,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003073755682142404,
      "loss": 4.6756,
      "mean_token_accuracy": 0.24181418716907502,
      "num_tokens": 91467272.0,
      "step": 40970
    },
    {
      "entropy": 5.300034666061402,
      "epoch": 4.601898129948896,
      "grad_norm": 1.0,
      "learning_rate": 0.00030733583678834765,
      "loss": 4.9075,
      "mean_token_accuracy": 0.22658019363880158,
      "num_tokens": 91478188.0,
      "step": 40975
    },
    {
      "entropy": 5.213083505630493,
      "epoch": 4.602459706856854,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030729610433286926,
      "loss": 4.7937,
      "mean_token_accuracy": 0.2295680671930313,
      "num_tokens": 91489759.0,
      "step": 40980
    },
    {
      "entropy": 5.234951114654541,
      "epoch": 4.603021283764812,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030725637084907037,
      "loss": 4.7966,
      "mean_token_accuracy": 0.23371512442827225,
      "num_tokens": 91500705.0,
      "step": 40985
    },
    {
      "entropy": 5.225265741348267,
      "epoch": 4.603582860672769,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003072166363382162,
      "loss": 4.8218,
      "mean_token_accuracy": 0.22899653762578964,
      "num_tokens": 91511832.0,
      "step": 40990
    },
    {
      "entropy": 5.2788654327392575,
      "epoch": 4.604144437580727,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003071769008015717,
      "loss": 4.814,
      "mean_token_accuracy": 0.23055777400732042,
      "num_tokens": 91522454.0,
      "step": 40995
    },
    {
      "entropy": 5.280348491668701,
      "epoch": 4.604706014488684,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003071371642404024,
      "loss": 4.8816,
      "mean_token_accuracy": 0.22735129594802855,
      "num_tokens": 91533987.0,
      "step": 41000
    },
    {
      "entropy": 5.218284177780151,
      "epoch": 4.605267591396641,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003070974266559732,
      "loss": 4.778,
      "mean_token_accuracy": 0.23648799806833268,
      "num_tokens": 91545000.0,
      "step": 41005
    },
    {
      "entropy": 5.354546403884887,
      "epoch": 4.6058291683045995,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003070576880495495,
      "loss": 4.854,
      "mean_token_accuracy": 0.2280619204044342,
      "num_tokens": 91556481.0,
      "step": 41010
    },
    {
      "entropy": 5.231566381454468,
      "epoch": 4.606390745212557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003070179484223966,
      "loss": 4.7189,
      "mean_token_accuracy": 0.24017608016729355,
      "num_tokens": 91568668.0,
      "step": 41015
    },
    {
      "entropy": 5.197770452499389,
      "epoch": 4.606952322120515,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003069782077757797,
      "loss": 4.7436,
      "mean_token_accuracy": 0.2361450031399727,
      "num_tokens": 91579884.0,
      "step": 41020
    },
    {
      "entropy": 5.284198951721192,
      "epoch": 4.607513899028472,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030693846611096437,
      "loss": 4.9088,
      "mean_token_accuracy": 0.23572328835725784,
      "num_tokens": 91590488.0,
      "step": 41025
    },
    {
      "entropy": 5.259480953216553,
      "epoch": 4.608075475936429,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030689872342921595,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23412356823682784,
      "num_tokens": 91602163.0,
      "step": 41030
    },
    {
      "entropy": 5.234159755706787,
      "epoch": 4.608637052844387,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030685897973179967,
      "loss": 4.7553,
      "mean_token_accuracy": 0.2436573714017868,
      "num_tokens": 91614251.0,
      "step": 41035
    },
    {
      "entropy": 5.215321731567383,
      "epoch": 4.609198629752345,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030681923501998124,
      "loss": 4.7844,
      "mean_token_accuracy": 0.23399398922920228,
      "num_tokens": 91625102.0,
      "step": 41040
    },
    {
      "entropy": 5.2862693786621096,
      "epoch": 4.609760206660302,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030677948929502596,
      "loss": 4.8434,
      "mean_token_accuracy": 0.2332201912999153,
      "num_tokens": 91636613.0,
      "step": 41045
    },
    {
      "entropy": 5.198167085647583,
      "epoch": 4.61032178356826,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030673974255819944,
      "loss": 4.7415,
      "mean_token_accuracy": 0.23343783020973205,
      "num_tokens": 91647648.0,
      "step": 41050
    },
    {
      "entropy": 5.360820198059082,
      "epoch": 4.610883360476217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003066999948107672,
      "loss": 4.879,
      "mean_token_accuracy": 0.22753241807222366,
      "num_tokens": 91659160.0,
      "step": 41055
    },
    {
      "entropy": 5.215195274353027,
      "epoch": 4.611444937384174,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003066602460539948,
      "loss": 4.7448,
      "mean_token_accuracy": 0.24171554744243623,
      "num_tokens": 91670973.0,
      "step": 41060
    },
    {
      "entropy": 5.292666625976563,
      "epoch": 4.6120065142921325,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030662049628914787,
      "loss": 4.8843,
      "mean_token_accuracy": 0.23148385286331177,
      "num_tokens": 91682710.0,
      "step": 41065
    },
    {
      "entropy": 5.280566883087158,
      "epoch": 4.61256809120009,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003065807455174921,
      "loss": 4.7811,
      "mean_token_accuracy": 0.23499589115381242,
      "num_tokens": 91694008.0,
      "step": 41070
    },
    {
      "entropy": 5.145795392990112,
      "epoch": 4.613129668108048,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0003065409937402932,
      "loss": 4.7562,
      "mean_token_accuracy": 0.2304515063762665,
      "num_tokens": 91705289.0,
      "step": 41075
    },
    {
      "entropy": 5.216486167907715,
      "epoch": 4.613691245016005,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030650124095881674,
      "loss": 4.7777,
      "mean_token_accuracy": 0.22964256554841994,
      "num_tokens": 91717078.0,
      "step": 41080
    },
    {
      "entropy": 5.209720420837402,
      "epoch": 4.614252821923962,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030646148717432855,
      "loss": 4.7971,
      "mean_token_accuracy": 0.23535648733377457,
      "num_tokens": 91728971.0,
      "step": 41085
    },
    {
      "entropy": 5.192715454101562,
      "epoch": 4.61481439883192,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030642173238809453,
      "loss": 4.7177,
      "mean_token_accuracy": 0.24193840324878693,
      "num_tokens": 91739496.0,
      "step": 41090
    },
    {
      "entropy": 5.23811149597168,
      "epoch": 4.615375975739878,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003063819766013802,
      "loss": 4.8084,
      "mean_token_accuracy": 0.22407470047473907,
      "num_tokens": 91749732.0,
      "step": 41095
    },
    {
      "entropy": 5.224412298202514,
      "epoch": 4.615937552647835,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003063422198154516,
      "loss": 4.7159,
      "mean_token_accuracy": 0.23164961785078048,
      "num_tokens": 91758772.0,
      "step": 41100
    },
    {
      "entropy": 5.1947370052337645,
      "epoch": 4.616499129555793,
      "grad_norm": 0.9375,
      "learning_rate": 0.0003063024620315746,
      "loss": 4.7155,
      "mean_token_accuracy": 0.24004772752523423,
      "num_tokens": 91770012.0,
      "step": 41105
    },
    {
      "entropy": 5.221889543533325,
      "epoch": 4.61706070646375,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003062627032510151,
      "loss": 4.8253,
      "mean_token_accuracy": 0.23017769157886506,
      "num_tokens": 91781926.0,
      "step": 41110
    },
    {
      "entropy": 5.307714366912842,
      "epoch": 4.617622283371707,
      "grad_norm": 0.953125,
      "learning_rate": 0.000306222943475039,
      "loss": 4.8862,
      "mean_token_accuracy": 0.23389707654714584,
      "num_tokens": 91794598.0,
      "step": 41115
    },
    {
      "entropy": 5.237533950805664,
      "epoch": 4.618183860279665,
      "grad_norm": 0.96875,
      "learning_rate": 0.0003061831827049123,
      "loss": 4.7922,
      "mean_token_accuracy": 0.22977840304374694,
      "num_tokens": 91805898.0,
      "step": 41120
    },
    {
      "entropy": 5.236455631256104,
      "epoch": 4.618745437187623,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00030614342094190095,
      "loss": 4.8028,
      "mean_token_accuracy": 0.22871690690517427,
      "num_tokens": 91816369.0,
      "step": 41125
    },
    {
      "entropy": 5.148324632644654,
      "epoch": 4.619307014095581,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030610365818727097,
      "loss": 4.7023,
      "mean_token_accuracy": 0.2353366270661354,
      "num_tokens": 91827223.0,
      "step": 41130
    },
    {
      "entropy": 5.276213884353638,
      "epoch": 4.619868591003538,
      "grad_norm": 0.9453125,
      "learning_rate": 0.0003060638944422884,
      "loss": 4.8335,
      "mean_token_accuracy": 0.2295664593577385,
      "num_tokens": 91838898.0,
      "step": 41135
    },
    {
      "entropy": 5.2842494487762455,
      "epoch": 4.620430167911495,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030602412970821944,
      "loss": 4.7814,
      "mean_token_accuracy": 0.23279293328523637,
      "num_tokens": 91850660.0,
      "step": 41140
    },
    {
      "entropy": 5.219230508804321,
      "epoch": 4.620991744819453,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003059843639863301,
      "loss": 4.8086,
      "mean_token_accuracy": 0.23579362779855728,
      "num_tokens": 91860594.0,
      "step": 41145
    },
    {
      "entropy": 5.163915014266967,
      "epoch": 4.6215533217274105,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030594459727788665,
      "loss": 4.7439,
      "mean_token_accuracy": 0.23120025992393495,
      "num_tokens": 91872209.0,
      "step": 41150
    },
    {
      "entropy": 5.251672410964966,
      "epoch": 4.622114898635368,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00030590482958415524,
      "loss": 4.7862,
      "mean_token_accuracy": 0.2374990001320839,
      "num_tokens": 91885241.0,
      "step": 41155
    },
    {
      "entropy": 5.341991853713989,
      "epoch": 4.622676475543326,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030586506090640205,
      "loss": 4.8608,
      "mean_token_accuracy": 0.24058810919523238,
      "num_tokens": 91896999.0,
      "step": 41160
    },
    {
      "entropy": 5.181241750717163,
      "epoch": 4.623238052451283,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003058252912458933,
      "loss": 4.7467,
      "mean_token_accuracy": 0.22643887847661973,
      "num_tokens": 91908554.0,
      "step": 41165
    },
    {
      "entropy": 5.192605876922608,
      "epoch": 4.62379962935924,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003057855206038954,
      "loss": 4.7751,
      "mean_token_accuracy": 0.23605832159519197,
      "num_tokens": 91920013.0,
      "step": 41170
    },
    {
      "entropy": 5.2720057487487795,
      "epoch": 4.624361206267198,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003057457489816745,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23382259160280228,
      "num_tokens": 91930951.0,
      "step": 41175
    },
    {
      "entropy": 5.239209747314453,
      "epoch": 4.624922783175156,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003057059763804971,
      "loss": 4.7901,
      "mean_token_accuracy": 0.23443024307489396,
      "num_tokens": 91941718.0,
      "step": 41180
    },
    {
      "entropy": 5.153269910812378,
      "epoch": 4.625484360083114,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003056662028016295,
      "loss": 4.7154,
      "mean_token_accuracy": 0.23558811396360396,
      "num_tokens": 91952756.0,
      "step": 41185
    },
    {
      "entropy": 5.175510740280151,
      "epoch": 4.626045936991071,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030562642824633805,
      "loss": 4.7714,
      "mean_token_accuracy": 0.2348056510090828,
      "num_tokens": 91964200.0,
      "step": 41190
    },
    {
      "entropy": 5.227747917175293,
      "epoch": 4.626607513899028,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003055866527158893,
      "loss": 4.8275,
      "mean_token_accuracy": 0.2281199336051941,
      "num_tokens": 91975032.0,
      "step": 41195
    },
    {
      "entropy": 5.223046445846558,
      "epoch": 4.627169090806986,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003055468762115497,
      "loss": 4.7333,
      "mean_token_accuracy": 0.22950491756200792,
      "num_tokens": 91986566.0,
      "step": 41200
    },
    {
      "entropy": 5.222451496124267,
      "epoch": 4.6277306677149435,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030550709873458567,
      "loss": 4.73,
      "mean_token_accuracy": 0.24088297486305238,
      "num_tokens": 91996944.0,
      "step": 41205
    },
    {
      "entropy": 5.21050386428833,
      "epoch": 4.628292244622902,
      "grad_norm": 1.0,
      "learning_rate": 0.00030546732028626373,
      "loss": 4.7238,
      "mean_token_accuracy": 0.24200665205717087,
      "num_tokens": 92007692.0,
      "step": 41210
    },
    {
      "entropy": 5.205171394348144,
      "epoch": 4.628853821530859,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030542754086785055,
      "loss": 4.7526,
      "mean_token_accuracy": 0.24043141454458236,
      "num_tokens": 92017514.0,
      "step": 41215
    },
    {
      "entropy": 5.182153654098511,
      "epoch": 4.629415398438816,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003053877604806127,
      "loss": 4.7342,
      "mean_token_accuracy": 0.2348097413778305,
      "num_tokens": 92028311.0,
      "step": 41220
    },
    {
      "entropy": 5.095512437820434,
      "epoch": 4.629976975346774,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030534797912581664,
      "loss": 4.698,
      "mean_token_accuracy": 0.24379759430885314,
      "num_tokens": 92038795.0,
      "step": 41225
    },
    {
      "entropy": 5.135411977767944,
      "epoch": 4.630538552254731,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00030530819680472926,
      "loss": 4.7465,
      "mean_token_accuracy": 0.23425991386175155,
      "num_tokens": 92049578.0,
      "step": 41230
    },
    {
      "entropy": 5.365638351440429,
      "epoch": 4.6311001291626885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030526841351861717,
      "loss": 4.9262,
      "mean_token_accuracy": 0.2284609407186508,
      "num_tokens": 92062868.0,
      "step": 41235
    },
    {
      "entropy": 5.329763889312744,
      "epoch": 4.631661706070647,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00030522862926874697,
      "loss": 4.8155,
      "mean_token_accuracy": 0.2410357266664505,
      "num_tokens": 92073819.0,
      "step": 41240
    },
    {
      "entropy": 5.255297613143921,
      "epoch": 4.632223282978604,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003051888440563855,
      "loss": 4.8408,
      "mean_token_accuracy": 0.2252032995223999,
      "num_tokens": 92085212.0,
      "step": 41245
    },
    {
      "entropy": 5.218042659759521,
      "epoch": 4.632784859886561,
      "grad_norm": 0.921875,
      "learning_rate": 0.0003051490578827996,
      "loss": 4.8054,
      "mean_token_accuracy": 0.2356482282280922,
      "num_tokens": 92096807.0,
      "step": 41250
    },
    {
      "entropy": 5.2095458984375,
      "epoch": 4.633346436794519,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00030510927074925594,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2395027756690979,
      "num_tokens": 92108619.0,
      "step": 41255
    },
    {
      "entropy": 5.257812261581421,
      "epoch": 4.633908013702476,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00030506948265702144,
      "loss": 4.8824,
      "mean_token_accuracy": 0.23202068209648133,
      "num_tokens": 92120815.0,
      "step": 41260
    },
    {
      "entropy": 5.201843118667602,
      "epoch": 4.6344695906104345,
      "grad_norm": 1.03125,
      "learning_rate": 0.000305029693607363,
      "loss": 4.757,
      "mean_token_accuracy": 0.23935923874378204,
      "num_tokens": 92131957.0,
      "step": 41265
    },
    {
      "entropy": 5.185826969146729,
      "epoch": 4.635031167518392,
      "grad_norm": 1.0,
      "learning_rate": 0.00030498990360154746,
      "loss": 4.7455,
      "mean_token_accuracy": 0.2420702174305916,
      "num_tokens": 92143936.0,
      "step": 41270
    },
    {
      "entropy": 5.229315137863159,
      "epoch": 4.635592744426349,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030495011264084165,
      "loss": 4.7839,
      "mean_token_accuracy": 0.23337873220443725,
      "num_tokens": 92155320.0,
      "step": 41275
    },
    {
      "entropy": 5.255706119537353,
      "epoch": 4.636154321334307,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030491032072651266,
      "loss": 4.8062,
      "mean_token_accuracy": 0.23590733855962753,
      "num_tokens": 92166831.0,
      "step": 41280
    },
    {
      "entropy": 5.303618335723877,
      "epoch": 4.636715898242264,
      "grad_norm": 1.0,
      "learning_rate": 0.00030487052785982757,
      "loss": 4.7992,
      "mean_token_accuracy": 0.23098343461751938,
      "num_tokens": 92177927.0,
      "step": 41285
    },
    {
      "entropy": 5.222573375701904,
      "epoch": 4.6372774751502215,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003048307340420532,
      "loss": 4.8017,
      "mean_token_accuracy": 0.2331949830055237,
      "num_tokens": 92188755.0,
      "step": 41290
    },
    {
      "entropy": 5.225586223602295,
      "epoch": 4.63783905205818,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003047909392744566,
      "loss": 4.7209,
      "mean_token_accuracy": 0.2383454069495201,
      "num_tokens": 92199156.0,
      "step": 41295
    },
    {
      "entropy": 5.239971494674682,
      "epoch": 4.638400628966137,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030475114355830513,
      "loss": 4.7219,
      "mean_token_accuracy": 0.24115965515375137,
      "num_tokens": 92211080.0,
      "step": 41300
    },
    {
      "entropy": 5.275614261627197,
      "epoch": 4.638962205874094,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030471134689486563,
      "loss": 4.8455,
      "mean_token_accuracy": 0.22741008549928665,
      "num_tokens": 92220734.0,
      "step": 41305
    },
    {
      "entropy": 5.2972077369689945,
      "epoch": 4.639523782782052,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003046715492854053,
      "loss": 4.8949,
      "mean_token_accuracy": 0.22687780261039733,
      "num_tokens": 92231166.0,
      "step": 41310
    },
    {
      "entropy": 5.261652278900146,
      "epoch": 4.640085359690009,
      "grad_norm": 1.0,
      "learning_rate": 0.00030463175073119133,
      "loss": 4.8129,
      "mean_token_accuracy": 0.2272655725479126,
      "num_tokens": 92241920.0,
      "step": 41315
    },
    {
      "entropy": 5.190199661254883,
      "epoch": 4.6406469365979675,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00030459195123349095,
      "loss": 4.7966,
      "mean_token_accuracy": 0.22653465569019318,
      "num_tokens": 92253745.0,
      "step": 41320
    },
    {
      "entropy": 5.22199125289917,
      "epoch": 4.641208513505925,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003045521507935714,
      "loss": 4.8139,
      "mean_token_accuracy": 0.23276439905166627,
      "num_tokens": 92263672.0,
      "step": 41325
    },
    {
      "entropy": 5.286272239685059,
      "epoch": 4.641770090413882,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003045123494126999,
      "loss": 4.847,
      "mean_token_accuracy": 0.23088071197271348,
      "num_tokens": 92275124.0,
      "step": 41330
    },
    {
      "entropy": 5.3563377380371096,
      "epoch": 4.64233166732184,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003044725470921437,
      "loss": 4.8519,
      "mean_token_accuracy": 0.22960194647312165,
      "num_tokens": 92287077.0,
      "step": 41335
    },
    {
      "entropy": 5.308028316497802,
      "epoch": 4.642893244229797,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003044327438331702,
      "loss": 4.8593,
      "mean_token_accuracy": 0.23059308528900146,
      "num_tokens": 92298649.0,
      "step": 41340
    },
    {
      "entropy": 5.199337768554687,
      "epoch": 4.643454821137754,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003043929396370467,
      "loss": 4.7278,
      "mean_token_accuracy": 0.23696579784154892,
      "num_tokens": 92309589.0,
      "step": 41345
    },
    {
      "entropy": 5.114311361312867,
      "epoch": 4.6440163980457125,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00030435313450504056,
      "loss": 4.7165,
      "mean_token_accuracy": 0.23574910461902618,
      "num_tokens": 92319927.0,
      "step": 41350
    },
    {
      "entropy": 5.230832290649414,
      "epoch": 4.64457797495367,
      "grad_norm": 0.953125,
      "learning_rate": 0.0003043133284384193,
      "loss": 4.8369,
      "mean_token_accuracy": 0.2305160403251648,
      "num_tokens": 92333156.0,
      "step": 41355
    },
    {
      "entropy": 5.1441117286682125,
      "epoch": 4.645139551861627,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0003042735214384503,
      "loss": 4.6269,
      "mean_token_accuracy": 0.24505677521228791,
      "num_tokens": 92345228.0,
      "step": 41360
    },
    {
      "entropy": 5.291011238098145,
      "epoch": 4.645701128769585,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000304233713506401,
      "loss": 4.8563,
      "mean_token_accuracy": 0.23465776443481445,
      "num_tokens": 92355922.0,
      "step": 41365
    },
    {
      "entropy": 5.2225476741790775,
      "epoch": 4.646262705677542,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003041939046435389,
      "loss": 4.7956,
      "mean_token_accuracy": 0.23763725459575652,
      "num_tokens": 92367255.0,
      "step": 41370
    },
    {
      "entropy": 5.163798570632935,
      "epoch": 4.6468242825855,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00030415409485113156,
      "loss": 4.7536,
      "mean_token_accuracy": 0.23606358021497725,
      "num_tokens": 92378577.0,
      "step": 41375
    },
    {
      "entropy": 5.253800773620606,
      "epoch": 4.647385859493458,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030411428413044657,
      "loss": 4.7362,
      "mean_token_accuracy": 0.23443756103515626,
      "num_tokens": 92389822.0,
      "step": 41380
    },
    {
      "entropy": 5.3482969284057615,
      "epoch": 4.647947436401415,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030407447248275137,
      "loss": 4.905,
      "mean_token_accuracy": 0.22479005753993989,
      "num_tokens": 92400682.0,
      "step": 41385
    },
    {
      "entropy": 5.294117403030396,
      "epoch": 4.648509013309373,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003040346599093138,
      "loss": 4.7623,
      "mean_token_accuracy": 0.24050316512584685,
      "num_tokens": 92412194.0,
      "step": 41390
    },
    {
      "entropy": 5.236356782913208,
      "epoch": 4.64907059021733,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003039948464114014,
      "loss": 4.7822,
      "mean_token_accuracy": 0.2363414466381073,
      "num_tokens": 92423311.0,
      "step": 41395
    },
    {
      "entropy": 5.201730155944825,
      "epoch": 4.649632167125288,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030395503199028186,
      "loss": 4.8865,
      "mean_token_accuracy": 0.2290159523487091,
      "num_tokens": 92434008.0,
      "step": 41400
    },
    {
      "entropy": 5.1859677791595455,
      "epoch": 4.6501937440332455,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003039152166472228,
      "loss": 4.7307,
      "mean_token_accuracy": 0.2415481001138687,
      "num_tokens": 92445887.0,
      "step": 41405
    },
    {
      "entropy": 5.3133941173553465,
      "epoch": 4.650755320941203,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003038754003834921,
      "loss": 4.8636,
      "mean_token_accuracy": 0.24355973303318024,
      "num_tokens": 92457856.0,
      "step": 41410
    },
    {
      "entropy": 5.313298654556275,
      "epoch": 4.651316897849161,
      "grad_norm": 0.984375,
      "learning_rate": 0.00030383558320035744,
      "loss": 4.8427,
      "mean_token_accuracy": 0.2375480830669403,
      "num_tokens": 92469976.0,
      "step": 41415
    },
    {
      "entropy": 5.195514678955078,
      "epoch": 4.651878474757118,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030379576509908656,
      "loss": 4.7208,
      "mean_token_accuracy": 0.23539242148399353,
      "num_tokens": 92481035.0,
      "step": 41420
    },
    {
      "entropy": 5.171347618103027,
      "epoch": 4.652440051665075,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00030375594608094737,
      "loss": 4.8086,
      "mean_token_accuracy": 0.2340395599603653,
      "num_tokens": 92492145.0,
      "step": 41425
    },
    {
      "entropy": 5.238928747177124,
      "epoch": 4.653001628573033,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003037161261472078,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2339181900024414,
      "num_tokens": 92504541.0,
      "step": 41430
    },
    {
      "entropy": 5.308482456207275,
      "epoch": 4.653563205480991,
      "grad_norm": 0.9375,
      "learning_rate": 0.00030367630529913554,
      "loss": 4.863,
      "mean_token_accuracy": 0.22993022203445435,
      "num_tokens": 92516911.0,
      "step": 41435
    },
    {
      "entropy": 5.3048543453216555,
      "epoch": 4.654124782388948,
      "grad_norm": 1.078125,
      "learning_rate": 0.00030363648353799857,
      "loss": 4.8256,
      "mean_token_accuracy": 0.23041145652532577,
      "num_tokens": 92526619.0,
      "step": 41440
    },
    {
      "entropy": 5.328931903839111,
      "epoch": 4.654686359296906,
      "grad_norm": 1.0625,
      "learning_rate": 0.00030359666086506493,
      "loss": 4.8336,
      "mean_token_accuracy": 0.22970838993787765,
      "num_tokens": 92536579.0,
      "step": 41445
    },
    {
      "entropy": 5.231435680389405,
      "epoch": 4.655247936204863,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030355683728160247,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23943205922842026,
      "num_tokens": 92548147.0,
      "step": 41450
    },
    {
      "entropy": 5.216999864578247,
      "epoch": 4.655809513112821,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003035170127888793,
      "loss": 4.8043,
      "mean_token_accuracy": 0.2413004830479622,
      "num_tokens": 92559335.0,
      "step": 41455
    },
    {
      "entropy": 5.214371538162231,
      "epoch": 4.656371090020778,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030347718738816333,
      "loss": 4.7044,
      "mean_token_accuracy": 0.23514473885297776,
      "num_tokens": 92571527.0,
      "step": 41460
    },
    {
      "entropy": 5.245526123046875,
      "epoch": 4.656932666928736,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00030343736108072266,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23667004704475403,
      "num_tokens": 92581996.0,
      "step": 41465
    },
    {
      "entropy": 5.202060747146606,
      "epoch": 4.657494243836694,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003033975338678254,
      "loss": 4.7604,
      "mean_token_accuracy": 0.2394145667552948,
      "num_tokens": 92593134.0,
      "step": 41470
    },
    {
      "entropy": 5.192568969726563,
      "epoch": 4.658055820744651,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00030335770575073967,
      "loss": 4.7634,
      "mean_token_accuracy": 0.23392338901758195,
      "num_tokens": 92605196.0,
      "step": 41475
    },
    {
      "entropy": 5.218348503112793,
      "epoch": 4.658617397652608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0003033178767307336,
      "loss": 4.7923,
      "mean_token_accuracy": 0.23194611966609954,
      "num_tokens": 92615151.0,
      "step": 41480
    },
    {
      "entropy": 5.2732898712158205,
      "epoch": 4.659178974560566,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003032780468090753,
      "loss": 4.8387,
      "mean_token_accuracy": 0.23247817903757095,
      "num_tokens": 92625723.0,
      "step": 41485
    },
    {
      "entropy": 5.253317975997925,
      "epoch": 4.6597405514685235,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003032382159870331,
      "loss": 4.8762,
      "mean_token_accuracy": 0.23354447185993193,
      "num_tokens": 92636699.0,
      "step": 41490
    },
    {
      "entropy": 5.247439336776734,
      "epoch": 4.660302128376481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030319838426587505,
      "loss": 4.8129,
      "mean_token_accuracy": 0.22827425897121428,
      "num_tokens": 92646977.0,
      "step": 41495
    },
    {
      "entropy": 5.181144380569458,
      "epoch": 4.660863705284439,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030315855164686956,
      "loss": 4.7107,
      "mean_token_accuracy": 0.23945663273334503,
      "num_tokens": 92658021.0,
      "step": 41500
    },
    {
      "entropy": 5.173824071884155,
      "epoch": 4.661425282192396,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003031187181312848,
      "loss": 4.6784,
      "mean_token_accuracy": 0.245303612947464,
      "num_tokens": 92668573.0,
      "step": 41505
    },
    {
      "entropy": 5.235267496109008,
      "epoch": 4.661986859100354,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030307888372038927,
      "loss": 4.8323,
      "mean_token_accuracy": 0.2399773269891739,
      "num_tokens": 92680286.0,
      "step": 41510
    },
    {
      "entropy": 5.214830541610718,
      "epoch": 4.662548436008311,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00030303904841545113,
      "loss": 4.8157,
      "mean_token_accuracy": 0.23340262472629547,
      "num_tokens": 92692309.0,
      "step": 41515
    },
    {
      "entropy": 5.174638938903809,
      "epoch": 4.663110012916269,
      "grad_norm": 1.078125,
      "learning_rate": 0.0003029992122177388,
      "loss": 4.7441,
      "mean_token_accuracy": 0.2428589329123497,
      "num_tokens": 92703194.0,
      "step": 41520
    },
    {
      "entropy": 5.3814427852630615,
      "epoch": 4.663671589824227,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030295937512852075,
      "loss": 4.9124,
      "mean_token_accuracy": 0.22865038365125656,
      "num_tokens": 92714218.0,
      "step": 41525
    },
    {
      "entropy": 5.236824703216553,
      "epoch": 4.664233166732184,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0003029195371490652,
      "loss": 4.7413,
      "mean_token_accuracy": 0.24344320595264435,
      "num_tokens": 92726422.0,
      "step": 41530
    },
    {
      "entropy": 5.225294303894043,
      "epoch": 4.664794743640142,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030287969828064087,
      "loss": 4.8115,
      "mean_token_accuracy": 0.23117559403181076,
      "num_tokens": 92736835.0,
      "step": 41535
    },
    {
      "entropy": 5.223731994628906,
      "epoch": 4.665356320548099,
      "grad_norm": 1.125,
      "learning_rate": 0.00030283985852451624,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24257729649543763,
      "num_tokens": 92746665.0,
      "step": 41540
    },
    {
      "entropy": 5.25496187210083,
      "epoch": 4.6659178974560565,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003028000178819596,
      "loss": 4.7896,
      "mean_token_accuracy": 0.23721262514591218,
      "num_tokens": 92758059.0,
      "step": 41545
    },
    {
      "entropy": 5.13163537979126,
      "epoch": 4.666479474364014,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030276017635423965,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24068332314491273,
      "num_tokens": 92768982.0,
      "step": 41550
    },
    {
      "entropy": 5.12239408493042,
      "epoch": 4.667041051271972,
      "grad_norm": 1.125,
      "learning_rate": 0.00030272033394262483,
      "loss": 4.6677,
      "mean_token_accuracy": 0.24168223291635513,
      "num_tokens": 92779117.0,
      "step": 41555
    },
    {
      "entropy": 5.232770013809204,
      "epoch": 4.667602628179929,
      "grad_norm": 0.921875,
      "learning_rate": 0.00030268049064838395,
      "loss": 4.7926,
      "mean_token_accuracy": 0.22828803211450577,
      "num_tokens": 92790595.0,
      "step": 41560
    },
    {
      "entropy": 5.270312881469726,
      "epoch": 4.668164205087887,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00030264064647278537,
      "loss": 4.824,
      "mean_token_accuracy": 0.23701032102108002,
      "num_tokens": 92801661.0,
      "step": 41565
    },
    {
      "entropy": 5.137563228607178,
      "epoch": 4.668725781995844,
      "grad_norm": 1.015625,
      "learning_rate": 0.000302600801417098,
      "loss": 4.6624,
      "mean_token_accuracy": 0.242216257750988,
      "num_tokens": 92813223.0,
      "step": 41570
    },
    {
      "entropy": 5.165786647796631,
      "epoch": 4.6692873589038015,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00030256095548259035,
      "loss": 4.8144,
      "mean_token_accuracy": 0.23689187467098236,
      "num_tokens": 92823839.0,
      "step": 41575
    },
    {
      "entropy": 5.098864507675171,
      "epoch": 4.66984893581176,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0003025211086705312,
      "loss": 4.6716,
      "mean_token_accuracy": 0.2397522509098053,
      "num_tokens": 92834739.0,
      "step": 41580
    },
    {
      "entropy": 5.24866886138916,
      "epoch": 4.670410512719717,
      "grad_norm": 1.09375,
      "learning_rate": 0.0003024812609821893,
      "loss": 4.8031,
      "mean_token_accuracy": 0.24014270603656768,
      "num_tokens": 92846369.0,
      "step": 41585
    },
    {
      "entropy": 5.186387443542481,
      "epoch": 4.670972089627675,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003024414124188333,
      "loss": 4.7716,
      "mean_token_accuracy": 0.2368997663259506,
      "num_tokens": 92857963.0,
      "step": 41590
    },
    {
      "entropy": 5.243175840377807,
      "epoch": 4.671533666535632,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0003024015629817322,
      "loss": 4.7368,
      "mean_token_accuracy": 0.2362124055624008,
      "num_tokens": 92868922.0,
      "step": 41595
    },
    {
      "entropy": 5.248807096481324,
      "epoch": 4.672095243443589,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003023617126721547,
      "loss": 4.7773,
      "mean_token_accuracy": 0.23386409431695937,
      "num_tokens": 92881107.0,
      "step": 41600
    },
    {
      "entropy": 5.2621866226196286,
      "epoch": 4.6726568203515475,
      "grad_norm": 1.125,
      "learning_rate": 0.0003023218614913696,
      "loss": 4.7867,
      "mean_token_accuracy": 0.23448487967252732,
      "num_tokens": 92891538.0,
      "step": 41605
    },
    {
      "entropy": 5.208858966827393,
      "epoch": 4.673218397259505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003022820094406458,
      "loss": 4.8103,
      "mean_token_accuracy": 0.2300663024187088,
      "num_tokens": 92903015.0,
      "step": 41610
    },
    {
      "entropy": 5.176233959197998,
      "epoch": 4.673779974167462,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0003022421565212524,
      "loss": 4.7179,
      "mean_token_accuracy": 0.23628106117248535,
      "num_tokens": 92914237.0,
      "step": 41615
    },
    {
      "entropy": 5.254446744918823,
      "epoch": 4.67434155107542,
      "grad_norm": 1.0,
      "learning_rate": 0.00030220230273445806,
      "loss": 4.7233,
      "mean_token_accuracy": 0.23803755193948745,
      "num_tokens": 92925351.0,
      "step": 41620
    },
    {
      "entropy": 5.230694150924682,
      "epoch": 4.674903127983377,
      "grad_norm": 1.046875,
      "learning_rate": 0.00030216244808153186,
      "loss": 4.7815,
      "mean_token_accuracy": 0.2409300535917282,
      "num_tokens": 92935472.0,
      "step": 41625
    },
    {
      "entropy": 5.272508049011231,
      "epoch": 4.6754647048913345,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003021225925637428,
      "loss": 4.89,
      "mean_token_accuracy": 0.2237304538488388,
      "num_tokens": 92948328.0,
      "step": 41630
    },
    {
      "entropy": 5.202515363693237,
      "epoch": 4.676026281799293,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003020827361823598,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24182479679584504,
      "num_tokens": 92958458.0,
      "step": 41635
    },
    {
      "entropy": 5.2303999900817875,
      "epoch": 4.67658785870725,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00030204287893865204,
      "loss": 4.8061,
      "mean_token_accuracy": 0.22882190942764283,
      "num_tokens": 92969870.0,
      "step": 41640
    },
    {
      "entropy": 5.1674212455749515,
      "epoch": 4.677149435615208,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030200302083388856,
      "loss": 4.7741,
      "mean_token_accuracy": 0.2382907509803772,
      "num_tokens": 92981597.0,
      "step": 41645
    },
    {
      "entropy": 5.299702167510986,
      "epoch": 4.677711012523165,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030196316186933834,
      "loss": 4.8314,
      "mean_token_accuracy": 0.2415623351931572,
      "num_tokens": 92992525.0,
      "step": 41650
    },
    {
      "entropy": 5.295407390594482,
      "epoch": 4.678272589431122,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030192330204627054,
      "loss": 4.7902,
      "mean_token_accuracy": 0.23357938081026078,
      "num_tokens": 93003652.0,
      "step": 41655
    },
    {
      "entropy": 5.322975730895996,
      "epoch": 4.6788341663390804,
      "grad_norm": 1.046875,
      "learning_rate": 0.0003018834413659545,
      "loss": 4.8763,
      "mean_token_accuracy": 0.23228095471858978,
      "num_tokens": 93015794.0,
      "step": 41660
    },
    {
      "entropy": 5.1520891189575195,
      "epoch": 4.679395743247038,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003018435798296592,
      "loss": 4.6952,
      "mean_token_accuracy": 0.24073555171489716,
      "num_tokens": 93026454.0,
      "step": 41665
    },
    {
      "entropy": 5.171286296844483,
      "epoch": 4.679957320154995,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003018037174386539,
      "loss": 4.7501,
      "mean_token_accuracy": 0.230028435587883,
      "num_tokens": 93039263.0,
      "step": 41670
    },
    {
      "entropy": 5.25229549407959,
      "epoch": 4.680518897062953,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0003017638541942078,
      "loss": 4.8079,
      "mean_token_accuracy": 0.2394163802266121,
      "num_tokens": 93050437.0,
      "step": 41675
    },
    {
      "entropy": 5.275503730773925,
      "epoch": 4.68108047397091,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030172399009759034,
      "loss": 4.8352,
      "mean_token_accuracy": 0.23201545923948289,
      "num_tokens": 93060602.0,
      "step": 41680
    },
    {
      "entropy": 5.2991866111755375,
      "epoch": 4.681642050878867,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0003016841251500705,
      "loss": 4.8365,
      "mean_token_accuracy": 0.22997706532478332,
      "num_tokens": 93073062.0,
      "step": 41685
    },
    {
      "entropy": 5.239197826385498,
      "epoch": 4.6822036277868255,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030164425935291794,
      "loss": 4.8211,
      "mean_token_accuracy": 0.23897149562835693,
      "num_tokens": 93083150.0,
      "step": 41690
    },
    {
      "entropy": 5.2625054836273195,
      "epoch": 4.682765204694783,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0003016043927074018,
      "loss": 4.8435,
      "mean_token_accuracy": 0.22979916334152223,
      "num_tokens": 93094506.0,
      "step": 41695
    },
    {
      "entropy": 5.142901659011841,
      "epoch": 4.683326781602741,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0003015645252147914,
      "loss": 4.671,
      "mean_token_accuracy": 0.23341472148895265,
      "num_tokens": 93105148.0,
      "step": 41700
    },
    {
      "entropy": 5.235538339614868,
      "epoch": 4.683888358510698,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003015246568763562,
      "loss": 4.8237,
      "mean_token_accuracy": 0.23221111297607422,
      "num_tokens": 93116470.0,
      "step": 41705
    },
    {
      "entropy": 5.214530897140503,
      "epoch": 4.684449935418655,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030148478769336564,
      "loss": 4.7716,
      "mean_token_accuracy": 0.23138223588466644,
      "num_tokens": 93126662.0,
      "step": 41710
    },
    {
      "entropy": 5.214367246627807,
      "epoch": 4.685011512326613,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003014449176670892,
      "loss": 4.8085,
      "mean_token_accuracy": 0.2302285075187683,
      "num_tokens": 93137690.0,
      "step": 41715
    },
    {
      "entropy": 5.124921464920044,
      "epoch": 4.685573089234571,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00030140504679879637,
      "loss": 4.741,
      "mean_token_accuracy": 0.2350754126906395,
      "num_tokens": 93149760.0,
      "step": 41720
    },
    {
      "entropy": 5.272023677825928,
      "epoch": 4.686134666142529,
      "grad_norm": 1.328125,
      "learning_rate": 0.0003013651750897566,
      "loss": 4.8327,
      "mean_token_accuracy": 0.2448715090751648,
      "num_tokens": 93162046.0,
      "step": 41725
    },
    {
      "entropy": 5.255337858200074,
      "epoch": 4.686696243050486,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003013253025412393,
      "loss": 4.7376,
      "mean_token_accuracy": 0.24093971997499466,
      "num_tokens": 93172912.0,
      "step": 41730
    },
    {
      "entropy": 5.243921852111816,
      "epoch": 4.687257819958443,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030128542915451427,
      "loss": 4.8438,
      "mean_token_accuracy": 0.22690722346305847,
      "num_tokens": 93184595.0,
      "step": 41735
    },
    {
      "entropy": 5.181595182418823,
      "epoch": 4.6878193968664,
      "grad_norm": 1.015625,
      "learning_rate": 0.00030124555493085095,
      "loss": 4.7271,
      "mean_token_accuracy": 0.23654299974441528,
      "num_tokens": 93194753.0,
      "step": 41740
    },
    {
      "entropy": 5.199721956253052,
      "epoch": 4.6883809737743585,
      "grad_norm": 1.109375,
      "learning_rate": 0.0003012056798715189,
      "loss": 4.7715,
      "mean_token_accuracy": 0.2349541202187538,
      "num_tokens": 93204259.0,
      "step": 41745
    },
    {
      "entropy": 5.192100381851196,
      "epoch": 4.688942550682316,
      "grad_norm": 1.046875,
      "learning_rate": 0.000301165803977788,
      "loss": 4.7,
      "mean_token_accuracy": 0.2470775678753853,
      "num_tokens": 93214984.0,
      "step": 41750
    },
    {
      "entropy": 5.219284200668335,
      "epoch": 4.689504127590274,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003011259272509277,
      "loss": 4.8361,
      "mean_token_accuracy": 0.22938413619995118,
      "num_tokens": 93227004.0,
      "step": 41755
    },
    {
      "entropy": 5.342329216003418,
      "epoch": 4.690065704498231,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0003010860496922077,
      "loss": 4.899,
      "mean_token_accuracy": 0.22543821930885316,
      "num_tokens": 93238485.0,
      "step": 41760
    },
    {
      "entropy": 5.2829302787780765,
      "epoch": 4.690627281406188,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030104617130289783,
      "loss": 4.8108,
      "mean_token_accuracy": 0.23593654483556747,
      "num_tokens": 93249318.0,
      "step": 41765
    },
    {
      "entropy": 5.27196741104126,
      "epoch": 4.691188858314146,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0003010062920842677,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23846006989479065,
      "num_tokens": 93262040.0,
      "step": 41770
    },
    {
      "entropy": 5.230993127822876,
      "epoch": 4.6917504352221036,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00030096641203758716,
      "loss": 4.8525,
      "mean_token_accuracy": 0.23220339119434358,
      "num_tokens": 93272766.0,
      "step": 41775
    },
    {
      "entropy": 5.334011793136597,
      "epoch": 4.692312012130062,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00030092653116412605,
      "loss": 4.8401,
      "mean_token_accuracy": 0.23820247650146484,
      "num_tokens": 93283179.0,
      "step": 41780
    },
    {
      "entropy": 5.2280439853668215,
      "epoch": 4.692873589038019,
      "grad_norm": 1.03125,
      "learning_rate": 0.000300886649465154,
      "loss": 4.7532,
      "mean_token_accuracy": 0.23126286119222642,
      "num_tokens": 93293484.0,
      "step": 41785
    },
    {
      "entropy": 5.195008420944214,
      "epoch": 4.693435165945976,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0003008467669419411,
      "loss": 4.743,
      "mean_token_accuracy": 0.2346260815858841,
      "num_tokens": 93305611.0,
      "step": 41790
    },
    {
      "entropy": 5.240504550933838,
      "epoch": 4.693996742853934,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003008068835957571,
      "loss": 4.7417,
      "mean_token_accuracy": 0.23871322125196456,
      "num_tokens": 93316495.0,
      "step": 41795
    },
    {
      "entropy": 5.169501924514771,
      "epoch": 4.694558319761891,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0003007669994278719,
      "loss": 4.6966,
      "mean_token_accuracy": 0.2406464472413063,
      "num_tokens": 93328484.0,
      "step": 41800
    },
    {
      "entropy": 5.245008754730224,
      "epoch": 4.695119896669849,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003007271144395554,
      "loss": 4.8549,
      "mean_token_accuracy": 0.2297447219491005,
      "num_tokens": 93339826.0,
      "step": 41805
    },
    {
      "entropy": 5.234109687805176,
      "epoch": 4.695681473577807,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0003006872286320777,
      "loss": 4.8191,
      "mean_token_accuracy": 0.23987438678741455,
      "num_tokens": 93351923.0,
      "step": 41810
    },
    {
      "entropy": 5.178082132339478,
      "epoch": 4.696243050485764,
      "grad_norm": 1.03125,
      "learning_rate": 0.0003006473420067086,
      "loss": 4.7219,
      "mean_token_accuracy": 0.236831434071064,
      "num_tokens": 93363813.0,
      "step": 41815
    },
    {
      "entropy": 5.2540364265441895,
      "epoch": 4.696804627393721,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0003006074545647182,
      "loss": 4.8699,
      "mean_token_accuracy": 0.24114856123924255,
      "num_tokens": 93374905.0,
      "step": 41820
    },
    {
      "entropy": 5.287430000305176,
      "epoch": 4.697366204301679,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030056756630737655,
      "loss": 4.8488,
      "mean_token_accuracy": 0.23485321998596193,
      "num_tokens": 93386473.0,
      "step": 41825
    },
    {
      "entropy": 5.2589105606079105,
      "epoch": 4.6979277812096365,
      "grad_norm": 1.09375,
      "learning_rate": 0.00030052767723595365,
      "loss": 4.8176,
      "mean_token_accuracy": 0.23398317694664,
      "num_tokens": 93398010.0,
      "step": 41830
    },
    {
      "entropy": 5.257199239730835,
      "epoch": 4.698489358117595,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0003004877873517196,
      "loss": 4.8551,
      "mean_token_accuracy": 0.2297453686594963,
      "num_tokens": 93410692.0,
      "step": 41835
    },
    {
      "entropy": 5.22028226852417,
      "epoch": 4.699050935025552,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00030044789665594444,
      "loss": 4.7524,
      "mean_token_accuracy": 0.2311755448579788,
      "num_tokens": 93421830.0,
      "step": 41840
    },
    {
      "entropy": 5.307054138183593,
      "epoch": 4.699612511933509,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0003004080051498985,
      "loss": 4.9318,
      "mean_token_accuracy": 0.2289809763431549,
      "num_tokens": 93434136.0,
      "step": 41845
    },
    {
      "entropy": 5.19147629737854,
      "epoch": 4.700174088841467,
      "grad_norm": 1.109375,
      "learning_rate": 0.00030036811283485184,
      "loss": 4.7084,
      "mean_token_accuracy": 0.24014998525381087,
      "num_tokens": 93444559.0,
      "step": 41850
    },
    {
      "entropy": 5.237454414367676,
      "epoch": 4.700735665749424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00030032821971207453,
      "loss": 4.7888,
      "mean_token_accuracy": 0.23095411211252212,
      "num_tokens": 93456258.0,
      "step": 41855
    },
    {
      "entropy": 5.223675346374511,
      "epoch": 4.701297242657382,
      "grad_norm": 1.015625,
      "learning_rate": 0.0003002883257828369,
      "loss": 4.7674,
      "mean_token_accuracy": 0.23892131298780442,
      "num_tokens": 93467402.0,
      "step": 41860
    },
    {
      "entropy": 5.24248480796814,
      "epoch": 4.70185881956534,
      "grad_norm": 0.96875,
      "learning_rate": 0.00030024843104840924,
      "loss": 4.7682,
      "mean_token_accuracy": 0.2326974019408226,
      "num_tokens": 93477515.0,
      "step": 41865
    },
    {
      "entropy": 5.274873113632202,
      "epoch": 4.702420396473297,
      "grad_norm": 1.125,
      "learning_rate": 0.00030020853551006164,
      "loss": 4.8513,
      "mean_token_accuracy": 0.2264925003051758,
      "num_tokens": 93488173.0,
      "step": 41870
    },
    {
      "entropy": 5.23474531173706,
      "epoch": 4.702981973381254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0003001686391690645,
      "loss": 4.669,
      "mean_token_accuracy": 0.24542700201272966,
      "num_tokens": 93498236.0,
      "step": 41875
    },
    {
      "entropy": 5.205762720108032,
      "epoch": 4.703543550289212,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00030012874202668823,
      "loss": 4.7463,
      "mean_token_accuracy": 0.24051887094974517,
      "num_tokens": 93510162.0,
      "step": 41880
    },
    {
      "entropy": 5.2313916206359865,
      "epoch": 4.7041051271971694,
      "grad_norm": 1.03125,
      "learning_rate": 0.00030008884408420293,
      "loss": 4.824,
      "mean_token_accuracy": 0.23719654232263565,
      "num_tokens": 93521344.0,
      "step": 41885
    },
    {
      "entropy": 5.258843183517456,
      "epoch": 4.7046667041051276,
      "grad_norm": 1.0625,
      "learning_rate": 0.0003000489453428792,
      "loss": 4.8015,
      "mean_token_accuracy": 0.23094963580369948,
      "num_tokens": 93533116.0,
      "step": 41890
    },
    {
      "entropy": 5.238783454895019,
      "epoch": 4.705228281013085,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0003000090458039873,
      "loss": 4.8418,
      "mean_token_accuracy": 0.23115420043468476,
      "num_tokens": 93544612.0,
      "step": 41895
    },
    {
      "entropy": 5.288690519332886,
      "epoch": 4.705789857921042,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029996914546879773,
      "loss": 4.7944,
      "mean_token_accuracy": 0.23468899577856064,
      "num_tokens": 93556033.0,
      "step": 41900
    },
    {
      "entropy": 5.335260009765625,
      "epoch": 4.706351434829,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002999292443385809,
      "loss": 4.9251,
      "mean_token_accuracy": 0.22639843374490737,
      "num_tokens": 93567250.0,
      "step": 41905
    },
    {
      "entropy": 5.291127777099609,
      "epoch": 4.706913011736957,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002998893424146072,
      "loss": 4.8416,
      "mean_token_accuracy": 0.22958435416221618,
      "num_tokens": 93578521.0,
      "step": 41910
    },
    {
      "entropy": 5.29158763885498,
      "epoch": 4.707474588644915,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002998494396981472,
      "loss": 4.8255,
      "mean_token_accuracy": 0.237677900493145,
      "num_tokens": 93589809.0,
      "step": 41915
    },
    {
      "entropy": 5.18988823890686,
      "epoch": 4.708036165552873,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002998095361904714,
      "loss": 4.6842,
      "mean_token_accuracy": 0.24750614464282988,
      "num_tokens": 93599688.0,
      "step": 41920
    },
    {
      "entropy": 5.185416316986084,
      "epoch": 4.70859774246083,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002997696318928503,
      "loss": 4.7614,
      "mean_token_accuracy": 0.24446642845869065,
      "num_tokens": 93610575.0,
      "step": 41925
    },
    {
      "entropy": 5.152307987213135,
      "epoch": 4.709159319368787,
      "grad_norm": 1.0,
      "learning_rate": 0.0002997297268065546,
      "loss": 4.7267,
      "mean_token_accuracy": 0.24521965831518172,
      "num_tokens": 93622329.0,
      "step": 41930
    },
    {
      "entropy": 5.184309816360473,
      "epoch": 4.709720896276745,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029968982093285475,
      "loss": 4.7549,
      "mean_token_accuracy": 0.235873019695282,
      "num_tokens": 93632686.0,
      "step": 41935
    },
    {
      "entropy": 5.267481899261474,
      "epoch": 4.710282473184702,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029964991427302147,
      "loss": 4.7918,
      "mean_token_accuracy": 0.23714353889226913,
      "num_tokens": 93643086.0,
      "step": 41940
    },
    {
      "entropy": 5.298587131500244,
      "epoch": 4.7108440500926605,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002996100068283253,
      "loss": 4.8164,
      "mean_token_accuracy": 0.23395881354808806,
      "num_tokens": 93654457.0,
      "step": 41945
    },
    {
      "entropy": 5.236347341537476,
      "epoch": 4.711405627000618,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029957009860003697,
      "loss": 4.8445,
      "mean_token_accuracy": 0.2296712264418602,
      "num_tokens": 93665806.0,
      "step": 41950
    },
    {
      "entropy": 5.207467889785766,
      "epoch": 4.711967203908575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029953018958942717,
      "loss": 4.8227,
      "mean_token_accuracy": 0.22824996709823608,
      "num_tokens": 93676685.0,
      "step": 41955
    },
    {
      "entropy": 5.343743133544922,
      "epoch": 4.712528780816533,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002994902797977667,
      "loss": 4.9065,
      "mean_token_accuracy": 0.22920321971178054,
      "num_tokens": 93688111.0,
      "step": 41960
    },
    {
      "entropy": 5.246121406555176,
      "epoch": 4.71309035772449,
      "grad_norm": 1.0,
      "learning_rate": 0.00029945036922632615,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23133155703544617,
      "num_tokens": 93698919.0,
      "step": 41965
    },
    {
      "entropy": 5.253857851028442,
      "epoch": 4.713651934632448,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029941045787637636,
      "loss": 4.8221,
      "mean_token_accuracy": 0.2298457682132721,
      "num_tokens": 93711367.0,
      "step": 41970
    },
    {
      "entropy": 5.216466188430786,
      "epoch": 4.714213511540406,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002993705457491882,
      "loss": 4.7663,
      "mean_token_accuracy": 0.2384597972035408,
      "num_tokens": 93722649.0,
      "step": 41975
    },
    {
      "entropy": 5.294773530960083,
      "epoch": 4.714775088448363,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029933063284603235,
      "loss": 4.8615,
      "mean_token_accuracy": 0.2285796582698822,
      "num_tokens": 93733873.0,
      "step": 41980
    },
    {
      "entropy": 5.159249210357666,
      "epoch": 4.715336665356321,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002992907191681797,
      "loss": 4.619,
      "mean_token_accuracy": 0.25541278421878816,
      "num_tokens": 93744954.0,
      "step": 41985
    },
    {
      "entropy": 5.223907423019409,
      "epoch": 4.715898242264278,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029925080471690117,
      "loss": 4.7705,
      "mean_token_accuracy": 0.2343378946185112,
      "num_tokens": 93756607.0,
      "step": 41990
    },
    {
      "entropy": 5.141498041152954,
      "epoch": 4.716459819172235,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029921088949346765,
      "loss": 4.729,
      "mean_token_accuracy": 0.2321431040763855,
      "num_tokens": 93767496.0,
      "step": 41995
    },
    {
      "entropy": 5.146331739425659,
      "epoch": 4.717021396080193,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002991709734991499,
      "loss": 4.704,
      "mean_token_accuracy": 0.23449007868766786,
      "num_tokens": 93778432.0,
      "step": 42000
    },
    {
      "epoch": 4.717021396080193,
      "eval_entropy": 5.030434479933595,
      "eval_loss": 4.9764933586120605,
      "eval_mean_token_accuracy": 0.23224899788315775,
      "eval_num_tokens": 93778432.0,
      "eval_runtime": 24.1592,
      "eval_samples_per_second": 1283.692,
      "eval_steps_per_second": 160.477,
      "step": 42000
    },
    {
      "entropy": 5.192931365966797,
      "epoch": 4.717582972988151,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029913105673521907,
      "loss": 4.7691,
      "mean_token_accuracy": 0.2332636073231697,
      "num_tokens": 93789652.0,
      "step": 42005
    },
    {
      "entropy": 5.259381246566773,
      "epoch": 4.718144549896108,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000299091139202946,
      "loss": 4.7465,
      "mean_token_accuracy": 0.23954323083162307,
      "num_tokens": 93800386.0,
      "step": 42010
    },
    {
      "entropy": 5.266865205764771,
      "epoch": 4.718706126804066,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029905122090360167,
      "loss": 4.7989,
      "mean_token_accuracy": 0.2357688695192337,
      "num_tokens": 93810184.0,
      "step": 42015
    },
    {
      "entropy": 5.207478475570679,
      "epoch": 4.719267703712023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002990113018384572,
      "loss": 4.7155,
      "mean_token_accuracy": 0.23684285283088685,
      "num_tokens": 93820139.0,
      "step": 42020
    },
    {
      "entropy": 5.213775777816773,
      "epoch": 4.719829280619981,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002989713820087836,
      "loss": 4.6752,
      "mean_token_accuracy": 0.24353758096694947,
      "num_tokens": 93831323.0,
      "step": 42025
    },
    {
      "entropy": 5.278851699829102,
      "epoch": 4.7203908575279385,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029893146141585187,
      "loss": 4.7883,
      "mean_token_accuracy": 0.23842610269784928,
      "num_tokens": 93841160.0,
      "step": 42030
    },
    {
      "entropy": 5.295783519744873,
      "epoch": 4.720952434435896,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029889154006093306,
      "loss": 4.9111,
      "mean_token_accuracy": 0.2268192708492279,
      "num_tokens": 93851530.0,
      "step": 42035
    },
    {
      "entropy": 5.314613151550293,
      "epoch": 4.721514011343854,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002988516179452984,
      "loss": 4.8222,
      "mean_token_accuracy": 0.24241721034049987,
      "num_tokens": 93862466.0,
      "step": 42040
    },
    {
      "entropy": 5.313533687591553,
      "epoch": 4.722075588251811,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029881169507021895,
      "loss": 4.7551,
      "mean_token_accuracy": 0.2484622597694397,
      "num_tokens": 93874399.0,
      "step": 42045
    },
    {
      "entropy": 5.216810464859009,
      "epoch": 4.722637165159768,
      "grad_norm": 1.03125,
      "learning_rate": 0.000298771771436966,
      "loss": 4.7579,
      "mean_token_accuracy": 0.23851071447134017,
      "num_tokens": 93885054.0,
      "step": 42050
    },
    {
      "entropy": 5.168850612640381,
      "epoch": 4.723198742067726,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002987318470468105,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24878327250480653,
      "num_tokens": 93894813.0,
      "step": 42055
    },
    {
      "entropy": 5.16182165145874,
      "epoch": 4.723760318975684,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029869192190102385,
      "loss": 4.7601,
      "mean_token_accuracy": 0.24036237597465515,
      "num_tokens": 93905625.0,
      "step": 42060
    },
    {
      "entropy": 5.20502667427063,
      "epoch": 4.724321895883641,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029865199600087724,
      "loss": 4.7397,
      "mean_token_accuracy": 0.24443283379077912,
      "num_tokens": 93917539.0,
      "step": 42065
    },
    {
      "entropy": 5.258980131149292,
      "epoch": 4.724883472791599,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002986120693476418,
      "loss": 4.751,
      "mean_token_accuracy": 0.23700616061687468,
      "num_tokens": 93927307.0,
      "step": 42070
    },
    {
      "entropy": 5.226017332077026,
      "epoch": 4.725445049699556,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029857214194258896,
      "loss": 4.7518,
      "mean_token_accuracy": 0.2439068540930748,
      "num_tokens": 93938252.0,
      "step": 42075
    },
    {
      "entropy": 5.237222671508789,
      "epoch": 4.726006626607514,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029853221378699,
      "loss": 4.7446,
      "mean_token_accuracy": 0.2404535785317421,
      "num_tokens": 93947762.0,
      "step": 42080
    },
    {
      "entropy": 5.256132364273071,
      "epoch": 4.7265682035154715,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002984922848821162,
      "loss": 4.8808,
      "mean_token_accuracy": 0.22894509583711625,
      "num_tokens": 93958410.0,
      "step": 42085
    },
    {
      "entropy": 5.226931524276734,
      "epoch": 4.727129780423429,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029845235522923903,
      "loss": 4.7756,
      "mean_token_accuracy": 0.238748237490654,
      "num_tokens": 93969610.0,
      "step": 42090
    },
    {
      "entropy": 5.262172269821167,
      "epoch": 4.727691357331387,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029841242482962966,
      "loss": 4.7727,
      "mean_token_accuracy": 0.23273730278015137,
      "num_tokens": 93978404.0,
      "step": 42095
    },
    {
      "entropy": 5.159990072250366,
      "epoch": 4.728252934239344,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029837249368455963,
      "loss": 4.7299,
      "mean_token_accuracy": 0.23225004822015763,
      "num_tokens": 93989578.0,
      "step": 42100
    },
    {
      "entropy": 5.270604944229126,
      "epoch": 4.728814511147302,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002983325617953004,
      "loss": 4.8647,
      "mean_token_accuracy": 0.22616825997829437,
      "num_tokens": 94000010.0,
      "step": 42105
    },
    {
      "entropy": 5.260491037368775,
      "epoch": 4.729376088055259,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029829262916312326,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2449204817414284,
      "num_tokens": 94009386.0,
      "step": 42110
    },
    {
      "entropy": 5.193319463729859,
      "epoch": 4.7299376649632165,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029825269578929986,
      "loss": 4.6696,
      "mean_token_accuracy": 0.2437155529856682,
      "num_tokens": 94021594.0,
      "step": 42115
    },
    {
      "entropy": 5.263955879211426,
      "epoch": 4.730499241871174,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029821276167510153,
      "loss": 4.8148,
      "mean_token_accuracy": 0.2300688400864601,
      "num_tokens": 94033518.0,
      "step": 42120
    },
    {
      "entropy": 5.198920345306396,
      "epoch": 4.731060818779132,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002981728268217999,
      "loss": 4.7802,
      "mean_token_accuracy": 0.22827684581279756,
      "num_tokens": 94044174.0,
      "step": 42125
    },
    {
      "entropy": 5.191985082626343,
      "epoch": 4.731622395687089,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002981328912306665,
      "loss": 4.7407,
      "mean_token_accuracy": 0.23633689880371095,
      "num_tokens": 94056000.0,
      "step": 42130
    },
    {
      "entropy": 5.264649391174316,
      "epoch": 4.732183972595047,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029809295490297285,
      "loss": 4.8766,
      "mean_token_accuracy": 0.23594100922346115,
      "num_tokens": 94067274.0,
      "step": 42135
    },
    {
      "entropy": 5.193397903442383,
      "epoch": 4.732745549503004,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002980530178399906,
      "loss": 4.7298,
      "mean_token_accuracy": 0.2339920088648796,
      "num_tokens": 94079113.0,
      "step": 42140
    },
    {
      "entropy": 5.19591121673584,
      "epoch": 4.733307126410962,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002980130800429912,
      "loss": 4.7815,
      "mean_token_accuracy": 0.2360069677233696,
      "num_tokens": 94090438.0,
      "step": 42145
    },
    {
      "entropy": 5.242797327041626,
      "epoch": 4.73386870331892,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029797314151324653,
      "loss": 4.7706,
      "mean_token_accuracy": 0.23271309584379196,
      "num_tokens": 94101153.0,
      "step": 42150
    },
    {
      "entropy": 5.374098587036133,
      "epoch": 4.734430280226877,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029793320225202807,
      "loss": 4.9228,
      "mean_token_accuracy": 0.23392973244190216,
      "num_tokens": 94112366.0,
      "step": 42155
    },
    {
      "entropy": 5.212232398986816,
      "epoch": 4.734991857134835,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002978932622606075,
      "loss": 4.7774,
      "mean_token_accuracy": 0.23617660105228425,
      "num_tokens": 94122919.0,
      "step": 42160
    },
    {
      "entropy": 5.2428154945373535,
      "epoch": 4.735553434042792,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029785332154025676,
      "loss": 4.8821,
      "mean_token_accuracy": 0.23369577378034592,
      "num_tokens": 94135158.0,
      "step": 42165
    },
    {
      "entropy": 5.257854700088501,
      "epoch": 4.7361150109507495,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002978133800922473,
      "loss": 4.8122,
      "mean_token_accuracy": 0.22518398761749267,
      "num_tokens": 94145773.0,
      "step": 42170
    },
    {
      "entropy": 5.190692567825318,
      "epoch": 4.736676587858708,
      "grad_norm": 1.0,
      "learning_rate": 0.0002977734379178509,
      "loss": 4.7203,
      "mean_token_accuracy": 0.23794980198144913,
      "num_tokens": 94157260.0,
      "step": 42175
    },
    {
      "entropy": 5.215803003311157,
      "epoch": 4.737238164766665,
      "grad_norm": 1.0,
      "learning_rate": 0.0002977334950183394,
      "loss": 4.7619,
      "mean_token_accuracy": 0.24020307809114455,
      "num_tokens": 94166513.0,
      "step": 42180
    },
    {
      "entropy": 5.306399154663086,
      "epoch": 4.737799741674622,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029769355139498465,
      "loss": 4.8922,
      "mean_token_accuracy": 0.2168360859155655,
      "num_tokens": 94178565.0,
      "step": 42185
    },
    {
      "entropy": 5.218471384048462,
      "epoch": 4.73836131858258,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002976536070490585,
      "loss": 4.7351,
      "mean_token_accuracy": 0.23795252442359924,
      "num_tokens": 94189319.0,
      "step": 42190
    },
    {
      "entropy": 5.281133842468262,
      "epoch": 4.738922895490537,
      "grad_norm": 0.96875,
      "learning_rate": 0.00029761366198183264,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23070551604032516,
      "num_tokens": 94203030.0,
      "step": 42195
    },
    {
      "entropy": 5.264018869400024,
      "epoch": 4.739484472398495,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00029757371619457904,
      "loss": 4.8371,
      "mean_token_accuracy": 0.2306304469704628,
      "num_tokens": 94214901.0,
      "step": 42200
    },
    {
      "entropy": 5.204097414016724,
      "epoch": 4.740046049306453,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029753376968856955,
      "loss": 4.7498,
      "mean_token_accuracy": 0.2368601903319359,
      "num_tokens": 94226833.0,
      "step": 42205
    },
    {
      "entropy": 5.292758274078369,
      "epoch": 4.74060762621441,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029749382246507604,
      "loss": 4.7783,
      "mean_token_accuracy": 0.2409559264779091,
      "num_tokens": 94239464.0,
      "step": 42210
    },
    {
      "entropy": 5.222079849243164,
      "epoch": 4.741169203122368,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029745387452537056,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2436130464076996,
      "num_tokens": 94250692.0,
      "step": 42215
    },
    {
      "entropy": 5.181179571151733,
      "epoch": 4.741730780030325,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00029741392587072504,
      "loss": 4.8028,
      "mean_token_accuracy": 0.23320293724536895,
      "num_tokens": 94261130.0,
      "step": 42220
    },
    {
      "entropy": 5.254964923858642,
      "epoch": 4.742292356938282,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029737397650241135,
      "loss": 4.8166,
      "mean_token_accuracy": 0.23261569887399675,
      "num_tokens": 94271581.0,
      "step": 42225
    },
    {
      "entropy": 5.282184267044068,
      "epoch": 4.7428539338462405,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002973340264217016,
      "loss": 4.793,
      "mean_token_accuracy": 0.23165538012981415,
      "num_tokens": 94283353.0,
      "step": 42230
    },
    {
      "entropy": 5.266208791732788,
      "epoch": 4.743415510754198,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029729407562986774,
      "loss": 4.8536,
      "mean_token_accuracy": 0.2312355011701584,
      "num_tokens": 94294669.0,
      "step": 42235
    },
    {
      "entropy": 5.234904432296753,
      "epoch": 4.743977087662155,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00029725412412818196,
      "loss": 4.7692,
      "mean_token_accuracy": 0.23974690437316895,
      "num_tokens": 94306127.0,
      "step": 42240
    },
    {
      "entropy": 5.218931007385254,
      "epoch": 4.744538664570113,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029721417191791617,
      "loss": 4.7065,
      "mean_token_accuracy": 0.2396002009510994,
      "num_tokens": 94316621.0,
      "step": 42245
    },
    {
      "entropy": 5.2526146411895756,
      "epoch": 4.74510024147807,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029717421900034253,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23303119093179703,
      "num_tokens": 94327715.0,
      "step": 42250
    },
    {
      "entropy": 5.210161733627319,
      "epoch": 4.7456618183860275,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002971342653767332,
      "loss": 4.7357,
      "mean_token_accuracy": 0.24235699325799942,
      "num_tokens": 94339452.0,
      "step": 42255
    },
    {
      "entropy": 5.22100510597229,
      "epoch": 4.746223395293986,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029709431104836027,
      "loss": 4.7587,
      "mean_token_accuracy": 0.24264533519744874,
      "num_tokens": 94350309.0,
      "step": 42260
    },
    {
      "entropy": 5.2836751461029055,
      "epoch": 4.746784972201943,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029705435601649584,
      "loss": 4.8755,
      "mean_token_accuracy": 0.22682124972343445,
      "num_tokens": 94361665.0,
      "step": 42265
    },
    {
      "entropy": 5.3561841487884525,
      "epoch": 4.747346549109901,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002970144002824122,
      "loss": 4.8778,
      "mean_token_accuracy": 0.22638976126909255,
      "num_tokens": 94374141.0,
      "step": 42270
    },
    {
      "entropy": 5.226428270339966,
      "epoch": 4.747908126017858,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002969744438473815,
      "loss": 4.7131,
      "mean_token_accuracy": 0.227239228785038,
      "num_tokens": 94384379.0,
      "step": 42275
    },
    {
      "entropy": 5.191330051422119,
      "epoch": 4.748469702925815,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002969344867126761,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24202504754066467,
      "num_tokens": 94394747.0,
      "step": 42280
    },
    {
      "entropy": 5.244580888748169,
      "epoch": 4.7490312798337735,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029689452887956804,
      "loss": 4.8623,
      "mean_token_accuracy": 0.2280452936887741,
      "num_tokens": 94407530.0,
      "step": 42285
    },
    {
      "entropy": 5.267723178863525,
      "epoch": 4.749592856741731,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029685457034932966,
      "loss": 4.8514,
      "mean_token_accuracy": 0.23003534227609634,
      "num_tokens": 94420031.0,
      "step": 42290
    },
    {
      "entropy": 5.265851306915283,
      "epoch": 4.750154433649689,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002968146111232333,
      "loss": 4.7446,
      "mean_token_accuracy": 0.2312387153506279,
      "num_tokens": 94429634.0,
      "step": 42295
    },
    {
      "entropy": 5.267775678634644,
      "epoch": 4.750716010557646,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029677465120255133,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23080833852291108,
      "num_tokens": 94440310.0,
      "step": 42300
    },
    {
      "entropy": 5.257569789886475,
      "epoch": 4.751277587465603,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002967346905885559,
      "loss": 4.8578,
      "mean_token_accuracy": 0.23041652739048005,
      "num_tokens": 94451050.0,
      "step": 42305
    },
    {
      "entropy": 5.306243133544922,
      "epoch": 4.7518391643735605,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029669472928251954,
      "loss": 4.8448,
      "mean_token_accuracy": 0.24068802893161773,
      "num_tokens": 94462210.0,
      "step": 42310
    },
    {
      "entropy": 5.138492679595947,
      "epoch": 4.752400741281519,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029665476728571466,
      "loss": 4.6436,
      "mean_token_accuracy": 0.2377065747976303,
      "num_tokens": 94472716.0,
      "step": 42315
    },
    {
      "entropy": 5.163623094558716,
      "epoch": 4.752962318189476,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002966148045994135,
      "loss": 4.6772,
      "mean_token_accuracy": 0.2397520959377289,
      "num_tokens": 94482446.0,
      "step": 42320
    },
    {
      "entropy": 5.2145256996154785,
      "epoch": 4.753523895097434,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002965748412248886,
      "loss": 4.7845,
      "mean_token_accuracy": 0.2362396538257599,
      "num_tokens": 94492476.0,
      "step": 42325
    },
    {
      "entropy": 5.181947183609009,
      "epoch": 4.754085472005391,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002965348771634124,
      "loss": 4.6967,
      "mean_token_accuracy": 0.24279128313064574,
      "num_tokens": 94502607.0,
      "step": 42330
    },
    {
      "entropy": 5.161437129974365,
      "epoch": 4.754647048913348,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029649491241625736,
      "loss": 4.6564,
      "mean_token_accuracy": 0.23864296227693557,
      "num_tokens": 94512528.0,
      "step": 42335
    },
    {
      "entropy": 5.414765882492065,
      "epoch": 4.755208625821306,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029645494698469604,
      "loss": 4.9879,
      "mean_token_accuracy": 0.22580883800983428,
      "num_tokens": 94524890.0,
      "step": 42340
    },
    {
      "entropy": 5.197875165939331,
      "epoch": 4.755770202729264,
      "grad_norm": 0.90625,
      "learning_rate": 0.00029641498087000083,
      "loss": 4.6566,
      "mean_token_accuracy": 0.2489125147461891,
      "num_tokens": 94536483.0,
      "step": 42345
    },
    {
      "entropy": 5.281005096435547,
      "epoch": 4.756331779637222,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002963750140734443,
      "loss": 4.9357,
      "mean_token_accuracy": 0.22342754602432252,
      "num_tokens": 94547235.0,
      "step": 42350
    },
    {
      "entropy": 5.292998743057251,
      "epoch": 4.756893356545179,
      "grad_norm": 1.125,
      "learning_rate": 0.0002963350465962991,
      "loss": 4.8097,
      "mean_token_accuracy": 0.2383578673005104,
      "num_tokens": 94558511.0,
      "step": 42355
    },
    {
      "entropy": 5.223335790634155,
      "epoch": 4.757454933453136,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002962950784398376,
      "loss": 4.8022,
      "mean_token_accuracy": 0.22662146985530854,
      "num_tokens": 94569949.0,
      "step": 42360
    },
    {
      "entropy": 5.226363945007324,
      "epoch": 4.758016510361094,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0002962551096053327,
      "loss": 4.7604,
      "mean_token_accuracy": 0.23708774149417877,
      "num_tokens": 94581353.0,
      "step": 42365
    },
    {
      "entropy": 5.178072834014893,
      "epoch": 4.7585780872690515,
      "grad_norm": 0.9375,
      "learning_rate": 0.00029621514009405675,
      "loss": 4.6811,
      "mean_token_accuracy": 0.23961562663316727,
      "num_tokens": 94591653.0,
      "step": 42370
    },
    {
      "entropy": 5.189004278182983,
      "epoch": 4.759139664177009,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002961751699072826,
      "loss": 4.7925,
      "mean_token_accuracy": 0.23448618203401567,
      "num_tokens": 94603615.0,
      "step": 42375
    },
    {
      "entropy": 5.248154258728027,
      "epoch": 4.759701241084967,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002961351990462827,
      "loss": 4.8882,
      "mean_token_accuracy": 0.22828168720006942,
      "num_tokens": 94614813.0,
      "step": 42380
    },
    {
      "entropy": 5.225432634353638,
      "epoch": 4.760262817992924,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029609522751232996,
      "loss": 4.6771,
      "mean_token_accuracy": 0.23888061344623565,
      "num_tokens": 94624483.0,
      "step": 42385
    },
    {
      "entropy": 5.259465742111206,
      "epoch": 4.760824394900881,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00029605525530669705,
      "loss": 4.7881,
      "mean_token_accuracy": 0.23874033838510514,
      "num_tokens": 94634332.0,
      "step": 42390
    },
    {
      "entropy": 5.178995084762573,
      "epoch": 4.761385971808839,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029601528243065656,
      "loss": 4.7425,
      "mean_token_accuracy": 0.24052809178829193,
      "num_tokens": 94645019.0,
      "step": 42395
    },
    {
      "entropy": 5.28426570892334,
      "epoch": 4.761947548716797,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029597530888548135,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23372431248426437,
      "num_tokens": 94656393.0,
      "step": 42400
    },
    {
      "entropy": 5.220977020263672,
      "epoch": 4.762509125624755,
      "grad_norm": 1.140625,
      "learning_rate": 0.00029593533467244415,
      "loss": 4.7214,
      "mean_token_accuracy": 0.23920051604509354,
      "num_tokens": 94666730.0,
      "step": 42405
    },
    {
      "entropy": 5.204805660247803,
      "epoch": 4.763070702532712,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002958953597928178,
      "loss": 4.8169,
      "mean_token_accuracy": 0.23146586418151854,
      "num_tokens": 94677937.0,
      "step": 42410
    },
    {
      "entropy": 5.231938982009888,
      "epoch": 4.763632279440669,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002958553842478751,
      "loss": 4.7789,
      "mean_token_accuracy": 0.23708059638738632,
      "num_tokens": 94688141.0,
      "step": 42415
    },
    {
      "entropy": 5.236642074584961,
      "epoch": 4.764193856348627,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029581540803888886,
      "loss": 4.7611,
      "mean_token_accuracy": 0.24150611609220504,
      "num_tokens": 94700055.0,
      "step": 42420
    },
    {
      "entropy": 5.2684314250946045,
      "epoch": 4.7647554332565845,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00029577543116713197,
      "loss": 4.8485,
      "mean_token_accuracy": 0.2315422534942627,
      "num_tokens": 94711858.0,
      "step": 42425
    },
    {
      "entropy": 5.252051162719726,
      "epoch": 4.765317010164542,
      "grad_norm": 1.21875,
      "learning_rate": 0.00029573545363387725,
      "loss": 4.839,
      "mean_token_accuracy": 0.23064298629760743,
      "num_tokens": 94722691.0,
      "step": 42430
    },
    {
      "entropy": 5.248066759109497,
      "epoch": 4.7658785870725,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002956954754403977,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23765638172626496,
      "num_tokens": 94733570.0,
      "step": 42435
    },
    {
      "entropy": 5.2293048858642575,
      "epoch": 4.766440163980457,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002956554965879662,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24183008223772048,
      "num_tokens": 94744719.0,
      "step": 42440
    },
    {
      "entropy": 5.286810493469238,
      "epoch": 4.767001740888414,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002956155170778556,
      "loss": 4.8814,
      "mean_token_accuracy": 0.22911386787891388,
      "num_tokens": 94755272.0,
      "step": 42445
    },
    {
      "entropy": 5.31816554069519,
      "epoch": 4.767563317796372,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002955755369113389,
      "loss": 4.9014,
      "mean_token_accuracy": 0.2303791433572769,
      "num_tokens": 94767178.0,
      "step": 42450
    },
    {
      "entropy": 5.225532150268554,
      "epoch": 4.7681248947043295,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029553555608968925,
      "loss": 4.759,
      "mean_token_accuracy": 0.23396682739257812,
      "num_tokens": 94777088.0,
      "step": 42455
    },
    {
      "entropy": 5.307748222351075,
      "epoch": 4.768686471612288,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002954955746141795,
      "loss": 4.8579,
      "mean_token_accuracy": 0.23330367505550384,
      "num_tokens": 94787995.0,
      "step": 42460
    },
    {
      "entropy": 5.292849063873291,
      "epoch": 4.769248048520245,
      "grad_norm": 1.0,
      "learning_rate": 0.0002954555924860827,
      "loss": 4.7874,
      "mean_token_accuracy": 0.23343326300382614,
      "num_tokens": 94797810.0,
      "step": 42465
    },
    {
      "entropy": 5.29623851776123,
      "epoch": 4.769809625428202,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002954156097066718,
      "loss": 4.7895,
      "mean_token_accuracy": 0.23675475716590882,
      "num_tokens": 94809725.0,
      "step": 42470
    },
    {
      "entropy": 5.3051341533660885,
      "epoch": 4.77037120233616,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002953756262772201,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2375253066420555,
      "num_tokens": 94819740.0,
      "step": 42475
    },
    {
      "entropy": 5.307251644134522,
      "epoch": 4.770932779244117,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002953356421990005,
      "loss": 4.8663,
      "mean_token_accuracy": 0.22695779651403428,
      "num_tokens": 94830419.0,
      "step": 42480
    },
    {
      "entropy": 5.252999258041382,
      "epoch": 4.7714943561520755,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002952956574732862,
      "loss": 4.8343,
      "mean_token_accuracy": 0.22632877230644227,
      "num_tokens": 94842796.0,
      "step": 42485
    },
    {
      "entropy": 5.291872978210449,
      "epoch": 4.772055933060033,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002952556721013503,
      "loss": 4.8735,
      "mean_token_accuracy": 0.23081649839878082,
      "num_tokens": 94854065.0,
      "step": 42490
    },
    {
      "entropy": 5.285281181335449,
      "epoch": 4.77261750996799,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002952156860844659,
      "loss": 4.7252,
      "mean_token_accuracy": 0.24556298553943634,
      "num_tokens": 94866264.0,
      "step": 42495
    },
    {
      "entropy": 5.253991031646729,
      "epoch": 4.773179086875947,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002951756994239063,
      "loss": 4.8233,
      "mean_token_accuracy": 0.22829024642705917,
      "num_tokens": 94878549.0,
      "step": 42500
    },
    {
      "entropy": 5.266448354721069,
      "epoch": 4.773740663783905,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00029513571212094443,
      "loss": 4.7921,
      "mean_token_accuracy": 0.23724705576896668,
      "num_tokens": 94890527.0,
      "step": 42505
    },
    {
      "entropy": 5.202773904800415,
      "epoch": 4.7743022406918625,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002950957241768539,
      "loss": 4.7439,
      "mean_token_accuracy": 0.23202210217714309,
      "num_tokens": 94901424.0,
      "step": 42510
    },
    {
      "entropy": 5.204929685592651,
      "epoch": 4.774863817599821,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029505573559290756,
      "loss": 4.7293,
      "mean_token_accuracy": 0.24417496025562285,
      "num_tokens": 94912153.0,
      "step": 42515
    },
    {
      "entropy": 5.260928297042847,
      "epoch": 4.775425394507778,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029501574637037884,
      "loss": 4.7929,
      "mean_token_accuracy": 0.2357147678732872,
      "num_tokens": 94923912.0,
      "step": 42520
    },
    {
      "entropy": 5.206602621078491,
      "epoch": 4.775986971415735,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000294975756510541,
      "loss": 4.6675,
      "mean_token_accuracy": 0.24568266570568084,
      "num_tokens": 94934768.0,
      "step": 42525
    },
    {
      "entropy": 5.250728273391724,
      "epoch": 4.776548548323693,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002949357660146673,
      "loss": 4.8791,
      "mean_token_accuracy": 0.2248656377196312,
      "num_tokens": 94946617.0,
      "step": 42530
    },
    {
      "entropy": 5.3464579582214355,
      "epoch": 4.77711012523165,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002948957748840311,
      "loss": 4.8653,
      "mean_token_accuracy": 0.22316263169050216,
      "num_tokens": 94958863.0,
      "step": 42535
    },
    {
      "entropy": 5.252085208892822,
      "epoch": 4.7776717021396085,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029485578311990574,
      "loss": 4.794,
      "mean_token_accuracy": 0.2376030534505844,
      "num_tokens": 94969930.0,
      "step": 42540
    },
    {
      "entropy": 5.2510394096374515,
      "epoch": 4.778233279047566,
      "grad_norm": 1.109375,
      "learning_rate": 0.00029481579072356456,
      "loss": 4.8194,
      "mean_token_accuracy": 0.2317234233021736,
      "num_tokens": 94980858.0,
      "step": 42545
    },
    {
      "entropy": 5.382183313369751,
      "epoch": 4.778794855955523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002947757976962809,
      "loss": 4.9242,
      "mean_token_accuracy": 0.22647121846675872,
      "num_tokens": 94993075.0,
      "step": 42550
    },
    {
      "entropy": 5.207949304580689,
      "epoch": 4.779356432863481,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029473580403932814,
      "loss": 4.751,
      "mean_token_accuracy": 0.23393159210681916,
      "num_tokens": 95003849.0,
      "step": 42555
    },
    {
      "entropy": 5.243566370010376,
      "epoch": 4.779918009771438,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002946958097539798,
      "loss": 4.7917,
      "mean_token_accuracy": 0.23492118269205092,
      "num_tokens": 95015766.0,
      "step": 42560
    },
    {
      "entropy": 5.2101490020751955,
      "epoch": 4.780479586679395,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002946558148415092,
      "loss": 4.838,
      "mean_token_accuracy": 0.23247105926275252,
      "num_tokens": 95026690.0,
      "step": 42565
    },
    {
      "entropy": 5.171800041198731,
      "epoch": 4.7810411635873535,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002946158193031898,
      "loss": 4.7414,
      "mean_token_accuracy": 0.24284403175115585,
      "num_tokens": 95038591.0,
      "step": 42570
    },
    {
      "entropy": 5.341688346862793,
      "epoch": 4.781602740495311,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029457582314029507,
      "loss": 4.8818,
      "mean_token_accuracy": 0.2293292537331581,
      "num_tokens": 95050268.0,
      "step": 42575
    },
    {
      "entropy": 5.357887411117554,
      "epoch": 4.782164317403268,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00029453582635409866,
      "loss": 4.7735,
      "mean_token_accuracy": 0.23464259654283523,
      "num_tokens": 95061448.0,
      "step": 42580
    },
    {
      "entropy": 5.2368494987487795,
      "epoch": 4.782725894311226,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002944958289458739,
      "loss": 4.7776,
      "mean_token_accuracy": 0.23223871141672134,
      "num_tokens": 95072331.0,
      "step": 42585
    },
    {
      "entropy": 5.220898723602295,
      "epoch": 4.783287471219183,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029445583091689437,
      "loss": 4.8082,
      "mean_token_accuracy": 0.2312780275940895,
      "num_tokens": 95082536.0,
      "step": 42590
    },
    {
      "entropy": 5.148130798339844,
      "epoch": 4.783849048127141,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002944158322684336,
      "loss": 4.7209,
      "mean_token_accuracy": 0.2440629705786705,
      "num_tokens": 95093877.0,
      "step": 42595
    },
    {
      "entropy": 5.287763786315918,
      "epoch": 4.784410625035099,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029437583300176526,
      "loss": 4.7974,
      "mean_token_accuracy": 0.23389589935541152,
      "num_tokens": 95104672.0,
      "step": 42600
    },
    {
      "entropy": 5.256943321228027,
      "epoch": 4.784972201943056,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029433583311816284,
      "loss": 4.7899,
      "mean_token_accuracy": 0.23278333842754365,
      "num_tokens": 95115206.0,
      "step": 42605
    },
    {
      "entropy": 5.193232297897339,
      "epoch": 4.785533778851014,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002942958326189,
      "loss": 4.7307,
      "mean_token_accuracy": 0.23676421940326692,
      "num_tokens": 95125142.0,
      "step": 42610
    },
    {
      "entropy": 5.220086908340454,
      "epoch": 4.786095355758971,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029425583150525044,
      "loss": 4.7903,
      "mean_token_accuracy": 0.23561375141143798,
      "num_tokens": 95136108.0,
      "step": 42615
    },
    {
      "entropy": 5.2435469150543215,
      "epoch": 4.786656932666928,
      "grad_norm": 1.125,
      "learning_rate": 0.00029421582977848763,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23231323361396788,
      "num_tokens": 95146452.0,
      "step": 42620
    },
    {
      "entropy": 5.120134687423706,
      "epoch": 4.7872185095748865,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029417582743988547,
      "loss": 4.5743,
      "mean_token_accuracy": 0.24304490834474562,
      "num_tokens": 95156665.0,
      "step": 42625
    },
    {
      "entropy": 5.201931428909302,
      "epoch": 4.787780086482844,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029413582449071745,
      "loss": 4.7513,
      "mean_token_accuracy": 0.23857809007167816,
      "num_tokens": 95167092.0,
      "step": 42630
    },
    {
      "entropy": 5.23696002960205,
      "epoch": 4.788341663390801,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002940958209322574,
      "loss": 4.7357,
      "mean_token_accuracy": 0.2366462767124176,
      "num_tokens": 95178158.0,
      "step": 42635
    },
    {
      "entropy": 5.198457908630371,
      "epoch": 4.788903240298759,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00029405581676577897,
      "loss": 4.7548,
      "mean_token_accuracy": 0.23106028288602828,
      "num_tokens": 95189432.0,
      "step": 42640
    },
    {
      "entropy": 5.21173586845398,
      "epoch": 4.789464817206716,
      "grad_norm": 1.1484375,
      "learning_rate": 0.000294015811992556,
      "loss": 4.7624,
      "mean_token_accuracy": 0.2344503492116928,
      "num_tokens": 95200682.0,
      "step": 42645
    },
    {
      "entropy": 5.215321683883667,
      "epoch": 4.790026394114674,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002939758066138621,
      "loss": 4.7158,
      "mean_token_accuracy": 0.23878752142190934,
      "num_tokens": 95212643.0,
      "step": 42650
    },
    {
      "entropy": 5.2337251663208,
      "epoch": 4.790587971022632,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002939358006309712,
      "loss": 4.7856,
      "mean_token_accuracy": 0.2380930721759796,
      "num_tokens": 95223098.0,
      "step": 42655
    },
    {
      "entropy": 5.1864031791687015,
      "epoch": 4.791149547930589,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002938957940451571,
      "loss": 4.7535,
      "mean_token_accuracy": 0.23587131649255752,
      "num_tokens": 95232946.0,
      "step": 42660
    },
    {
      "entropy": 5.265413045883179,
      "epoch": 4.791711124838547,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00029385578685769346,
      "loss": 4.8193,
      "mean_token_accuracy": 0.23537904471158982,
      "num_tokens": 95243966.0,
      "step": 42665
    },
    {
      "entropy": 5.219954490661621,
      "epoch": 4.792272701746504,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002938157790698544,
      "loss": 4.8081,
      "mean_token_accuracy": 0.2377510815858841,
      "num_tokens": 95255251.0,
      "step": 42670
    },
    {
      "entropy": 5.1802818775177,
      "epoch": 4.792834278654462,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002937757706829136,
      "loss": 4.7023,
      "mean_token_accuracy": 0.24633622914552689,
      "num_tokens": 95267517.0,
      "step": 42675
    },
    {
      "entropy": 5.229265785217285,
      "epoch": 4.793395855562419,
      "grad_norm": 1.140625,
      "learning_rate": 0.000293735761698145,
      "loss": 4.8319,
      "mean_token_accuracy": 0.23043248504400254,
      "num_tokens": 95277508.0,
      "step": 42680
    },
    {
      "entropy": 5.202279329299927,
      "epoch": 4.793957432470377,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002936957521168224,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23640658259391784,
      "num_tokens": 95289757.0,
      "step": 42685
    },
    {
      "entropy": 5.275124740600586,
      "epoch": 4.794519009378335,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002936557419402198,
      "loss": 4.8252,
      "mean_token_accuracy": 0.23289871364831924,
      "num_tokens": 95301541.0,
      "step": 42690
    },
    {
      "entropy": 5.186788034439087,
      "epoch": 4.795080586286292,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002936157311696112,
      "loss": 4.7475,
      "mean_token_accuracy": 0.23686839789152145,
      "num_tokens": 95312840.0,
      "step": 42695
    },
    {
      "entropy": 5.1860438823699955,
      "epoch": 4.795642163194249,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002935757198062705,
      "loss": 4.6897,
      "mean_token_accuracy": 0.2427582785487175,
      "num_tokens": 95322909.0,
      "step": 42700
    },
    {
      "entropy": 5.089188718795777,
      "epoch": 4.796203740102207,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002935357078514717,
      "loss": 4.6414,
      "mean_token_accuracy": 0.23978898972272872,
      "num_tokens": 95333437.0,
      "step": 42705
    },
    {
      "entropy": 5.20587124824524,
      "epoch": 4.7967653170101645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029349569530648877,
      "loss": 4.7589,
      "mean_token_accuracy": 0.23832244873046876,
      "num_tokens": 95344477.0,
      "step": 42710
    },
    {
      "entropy": 5.27377142906189,
      "epoch": 4.797326893918122,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002934556821725957,
      "loss": 4.8227,
      "mean_token_accuracy": 0.23054609149694444,
      "num_tokens": 95354354.0,
      "step": 42715
    },
    {
      "entropy": 5.231694078445434,
      "epoch": 4.79788847082608,
      "grad_norm": 1.09375,
      "learning_rate": 0.00029341566845106647,
      "loss": 4.7686,
      "mean_token_accuracy": 0.2352826938033104,
      "num_tokens": 95365026.0,
      "step": 42720
    },
    {
      "entropy": 5.304851102828979,
      "epoch": 4.798450047734037,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002933756541431753,
      "loss": 4.8423,
      "mean_token_accuracy": 0.23436466306447984,
      "num_tokens": 95376625.0,
      "step": 42725
    },
    {
      "entropy": 5.195061540603637,
      "epoch": 4.799011624641995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029333563925019616,
      "loss": 4.7921,
      "mean_token_accuracy": 0.2328742891550064,
      "num_tokens": 95387066.0,
      "step": 42730
    },
    {
      "entropy": 5.153316450119019,
      "epoch": 4.799573201549952,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029329562377340327,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23498481512069702,
      "num_tokens": 95398623.0,
      "step": 42735
    },
    {
      "entropy": 5.340720176696777,
      "epoch": 4.80013477845791,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002932556077140704,
      "loss": 4.8546,
      "mean_token_accuracy": 0.23290211260318755,
      "num_tokens": 95412514.0,
      "step": 42740
    },
    {
      "entropy": 5.331302785873413,
      "epoch": 4.800696355365868,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029321559107347206,
      "loss": 4.8375,
      "mean_token_accuracy": 0.22562118470668793,
      "num_tokens": 95422885.0,
      "step": 42745
    },
    {
      "entropy": 5.2619373321533205,
      "epoch": 4.801257932273825,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029317557385288216,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23565358072519302,
      "num_tokens": 95433659.0,
      "step": 42750
    },
    {
      "entropy": 5.29003643989563,
      "epoch": 4.801819509181782,
      "grad_norm": 1.1328125,
      "learning_rate": 0.000293135556053575,
      "loss": 4.8306,
      "mean_token_accuracy": 0.23060289174318313,
      "num_tokens": 95444161.0,
      "step": 42755
    },
    {
      "entropy": 5.244157409667968,
      "epoch": 4.80238108608974,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029309553767682467,
      "loss": 4.7711,
      "mean_token_accuracy": 0.2313929855823517,
      "num_tokens": 95454840.0,
      "step": 42760
    },
    {
      "entropy": 5.307953357696533,
      "epoch": 4.8029426629976975,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002930555187239054,
      "loss": 4.7927,
      "mean_token_accuracy": 0.23490502834320068,
      "num_tokens": 95466766.0,
      "step": 42765
    },
    {
      "entropy": 5.241179656982422,
      "epoch": 4.803504239905655,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029301549919609146,
      "loss": 4.7977,
      "mean_token_accuracy": 0.23925425112247467,
      "num_tokens": 95477767.0,
      "step": 42770
    },
    {
      "entropy": 5.200474834442138,
      "epoch": 4.804065816813613,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00029297547909465697,
      "loss": 4.7495,
      "mean_token_accuracy": 0.2286310762166977,
      "num_tokens": 95489428.0,
      "step": 42775
    },
    {
      "entropy": 5.254314994812011,
      "epoch": 4.80462739372157,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002929354584208763,
      "loss": 4.8908,
      "mean_token_accuracy": 0.2297083005309105,
      "num_tokens": 95500844.0,
      "step": 42780
    },
    {
      "entropy": 5.219866943359375,
      "epoch": 4.805188970629528,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002928954371760237,
      "loss": 4.6952,
      "mean_token_accuracy": 0.24961926490068437,
      "num_tokens": 95511829.0,
      "step": 42785
    },
    {
      "entropy": 5.335515689849854,
      "epoch": 4.805750547537485,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002928554153613734,
      "loss": 4.8629,
      "mean_token_accuracy": 0.23048376590013503,
      "num_tokens": 95523366.0,
      "step": 42790
    },
    {
      "entropy": 5.14068374633789,
      "epoch": 4.8063121244454425,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002928153929781997,
      "loss": 4.6427,
      "mean_token_accuracy": 0.2479366272687912,
      "num_tokens": 95533495.0,
      "step": 42795
    },
    {
      "entropy": 5.165835905075073,
      "epoch": 4.806873701353401,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029277537002777703,
      "loss": 4.7715,
      "mean_token_accuracy": 0.23506071120500566,
      "num_tokens": 95545375.0,
      "step": 42800
    },
    {
      "entropy": 5.120263004302979,
      "epoch": 4.807435278261358,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00029273534651137967,
      "loss": 4.6961,
      "mean_token_accuracy": 0.2412147507071495,
      "num_tokens": 95557212.0,
      "step": 42805
    },
    {
      "entropy": 5.365332317352295,
      "epoch": 4.807996855169315,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029269532243028195,
      "loss": 4.935,
      "mean_token_accuracy": 0.23399805277585983,
      "num_tokens": 95568608.0,
      "step": 42810
    },
    {
      "entropy": 5.141621828079224,
      "epoch": 4.808558432077273,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002926552977857584,
      "loss": 4.6913,
      "mean_token_accuracy": 0.24036397486925126,
      "num_tokens": 95579069.0,
      "step": 42815
    },
    {
      "entropy": 5.2327783584594725,
      "epoch": 4.80912000898523,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029261527257908325,
      "loss": 4.8181,
      "mean_token_accuracy": 0.23513815850019454,
      "num_tokens": 95589244.0,
      "step": 42820
    },
    {
      "entropy": 5.188214731216431,
      "epoch": 4.809681585893188,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00029257524681153106,
      "loss": 4.8213,
      "mean_token_accuracy": 0.24168289750814437,
      "num_tokens": 95600277.0,
      "step": 42825
    },
    {
      "entropy": 5.29790096282959,
      "epoch": 4.810243162801146,
      "grad_norm": 1.0,
      "learning_rate": 0.0002925352204843761,
      "loss": 4.8032,
      "mean_token_accuracy": 0.23424663841724397,
      "num_tokens": 95609913.0,
      "step": 42830
    },
    {
      "entropy": 5.378079557418824,
      "epoch": 4.810804739709103,
      "grad_norm": 1.0,
      "learning_rate": 0.000292495193598893,
      "loss": 4.8919,
      "mean_token_accuracy": 0.23054115623235702,
      "num_tokens": 95623142.0,
      "step": 42835
    },
    {
      "entropy": 5.216630601882935,
      "epoch": 4.811366316617061,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029245516615635615,
      "loss": 4.8177,
      "mean_token_accuracy": 0.22951046228408814,
      "num_tokens": 95633711.0,
      "step": 42840
    },
    {
      "entropy": 5.231844568252564,
      "epoch": 4.811927893525018,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00029241513815804,
      "loss": 4.7551,
      "mean_token_accuracy": 0.24110592007637024,
      "num_tokens": 95646956.0,
      "step": 42845
    },
    {
      "entropy": 5.333809518814087,
      "epoch": 4.8124894704329755,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029237510960521904,
      "loss": 4.8429,
      "mean_token_accuracy": 0.22442401945590973,
      "num_tokens": 95659260.0,
      "step": 42850
    },
    {
      "entropy": 5.326037454605102,
      "epoch": 4.813051047340934,
      "grad_norm": 0.95703125,
      "learning_rate": 0.000292335080499168,
      "loss": 4.8705,
      "mean_token_accuracy": 0.23163036555051802,
      "num_tokens": 95671122.0,
      "step": 42855
    },
    {
      "entropy": 5.223679780960083,
      "epoch": 4.813612624248891,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029229505084116123,
      "loss": 4.7668,
      "mean_token_accuracy": 0.2359406664967537,
      "num_tokens": 95681513.0,
      "step": 42860
    },
    {
      "entropy": 5.231516790390015,
      "epoch": 4.814174201156849,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002922550206324732,
      "loss": 4.7574,
      "mean_token_accuracy": 0.2382620871067047,
      "num_tokens": 95691670.0,
      "step": 42865
    },
    {
      "entropy": 5.204552030563354,
      "epoch": 4.814735778064806,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002922149898743787,
      "loss": 4.7778,
      "mean_token_accuracy": 0.2380501225590706,
      "num_tokens": 95702359.0,
      "step": 42870
    },
    {
      "entropy": 5.314701652526855,
      "epoch": 4.815297354972763,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002921749585681523,
      "loss": 4.8078,
      "mean_token_accuracy": 0.24402069598436354,
      "num_tokens": 95714069.0,
      "step": 42875
    },
    {
      "entropy": 5.243848991394043,
      "epoch": 4.8158589318807214,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029213492671506855,
      "loss": 4.7645,
      "mean_token_accuracy": 0.23253798484802246,
      "num_tokens": 95724420.0,
      "step": 42880
    },
    {
      "entropy": 5.242564296722412,
      "epoch": 4.816420508788679,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029209489431640203,
      "loss": 4.7654,
      "mean_token_accuracy": 0.23861873596906663,
      "num_tokens": 95735229.0,
      "step": 42885
    },
    {
      "entropy": 5.125169992446899,
      "epoch": 4.816982085696636,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002920548613734275,
      "loss": 4.6303,
      "mean_token_accuracy": 0.24617010205984116,
      "num_tokens": 95745246.0,
      "step": 42890
    },
    {
      "entropy": 5.159977579116822,
      "epoch": 4.817543662604594,
      "grad_norm": 1.0,
      "learning_rate": 0.00029201482788741953,
      "loss": 4.79,
      "mean_token_accuracy": 0.23271361440420152,
      "num_tokens": 95755752.0,
      "step": 42895
    },
    {
      "entropy": 5.277925300598144,
      "epoch": 4.818105239512551,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002919747938596528,
      "loss": 4.857,
      "mean_token_accuracy": 0.23032985478639603,
      "num_tokens": 95766476.0,
      "step": 42900
    },
    {
      "entropy": 5.3268153190612795,
      "epoch": 4.818666816420508,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002919347592914021,
      "loss": 4.8534,
      "mean_token_accuracy": 0.22711195945739746,
      "num_tokens": 95776862.0,
      "step": 42905
    },
    {
      "entropy": 5.260144805908203,
      "epoch": 4.8192283933284665,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00029189472418394215,
      "loss": 4.7648,
      "mean_token_accuracy": 0.23737579584121704,
      "num_tokens": 95788216.0,
      "step": 42910
    },
    {
      "entropy": 5.265138149261475,
      "epoch": 4.819789970236424,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029185468853854755,
      "loss": 4.7875,
      "mean_token_accuracy": 0.24217973053455352,
      "num_tokens": 95799050.0,
      "step": 42915
    },
    {
      "entropy": 5.26116452217102,
      "epoch": 4.820351547144382,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002918146523564932,
      "loss": 4.7779,
      "mean_token_accuracy": 0.23420479893684387,
      "num_tokens": 95810763.0,
      "step": 42920
    },
    {
      "entropy": 5.275738954544067,
      "epoch": 4.820913124052339,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00029177461563905375,
      "loss": 4.7946,
      "mean_token_accuracy": 0.23673703223466874,
      "num_tokens": 95821730.0,
      "step": 42925
    },
    {
      "entropy": 5.307149028778076,
      "epoch": 4.821474700960296,
      "grad_norm": 0.984375,
      "learning_rate": 0.00029173457838750405,
      "loss": 4.8511,
      "mean_token_accuracy": 0.22964649349451066,
      "num_tokens": 95834037.0,
      "step": 42930
    },
    {
      "entropy": 5.182266187667847,
      "epoch": 4.822036277868254,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00029169454060311886,
      "loss": 4.741,
      "mean_token_accuracy": 0.23332736343145372,
      "num_tokens": 95845074.0,
      "step": 42935
    },
    {
      "entropy": 5.309709930419922,
      "epoch": 4.822597854776212,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000291654502287173,
      "loss": 4.8421,
      "mean_token_accuracy": 0.23579255044460296,
      "num_tokens": 95855439.0,
      "step": 42940
    },
    {
      "entropy": 5.357291841506958,
      "epoch": 4.823159431684169,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002916144634409414,
      "loss": 4.849,
      "mean_token_accuracy": 0.235209558904171,
      "num_tokens": 95866511.0,
      "step": 42945
    },
    {
      "entropy": 5.272783088684082,
      "epoch": 4.823721008592127,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029157442406569885,
      "loss": 4.7842,
      "mean_token_accuracy": 0.23485724478960038,
      "num_tokens": 95877795.0,
      "step": 42950
    },
    {
      "entropy": 5.20100359916687,
      "epoch": 4.824282585500084,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002915343841627201,
      "loss": 4.8046,
      "mean_token_accuracy": 0.23641516715288163,
      "num_tokens": 95889039.0,
      "step": 42955
    },
    {
      "entropy": 5.2075506210327145,
      "epoch": 4.824844162408041,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002914943437332802,
      "loss": 4.7004,
      "mean_token_accuracy": 0.2450213223695755,
      "num_tokens": 95900101.0,
      "step": 42960
    },
    {
      "entropy": 5.192027616500854,
      "epoch": 4.8254057393159995,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029145430277865406,
      "loss": 4.7103,
      "mean_token_accuracy": 0.23974850475788118,
      "num_tokens": 95910023.0,
      "step": 42965
    },
    {
      "entropy": 5.20181245803833,
      "epoch": 4.825967316223957,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002914142613001165,
      "loss": 4.7029,
      "mean_token_accuracy": 0.24464835822582245,
      "num_tokens": 95920610.0,
      "step": 42970
    },
    {
      "entropy": 5.1692663669586185,
      "epoch": 4.826528893131915,
      "grad_norm": 1.015625,
      "learning_rate": 0.00029137421929894235,
      "loss": 4.6976,
      "mean_token_accuracy": 0.2438136473298073,
      "num_tokens": 95931521.0,
      "step": 42975
    },
    {
      "entropy": 5.345513153076172,
      "epoch": 4.827090470039872,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002913341767764069,
      "loss": 4.8429,
      "mean_token_accuracy": 0.24151468724012376,
      "num_tokens": 95943179.0,
      "step": 42980
    },
    {
      "entropy": 5.2918072700500485,
      "epoch": 4.827652046947829,
      "grad_norm": 1.0,
      "learning_rate": 0.00029129413373378484,
      "loss": 4.7833,
      "mean_token_accuracy": 0.23539597988128663,
      "num_tokens": 95955282.0,
      "step": 42985
    },
    {
      "entropy": 5.207897520065307,
      "epoch": 4.828213623855787,
      "grad_norm": 1.0625,
      "learning_rate": 0.00029125409017235126,
      "loss": 4.6804,
      "mean_token_accuracy": 0.23883818387985228,
      "num_tokens": 95966296.0,
      "step": 42990
    },
    {
      "entropy": 5.288874769210816,
      "epoch": 4.828775200763745,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002912140460933812,
      "loss": 4.7999,
      "mean_token_accuracy": 0.23287111818790435,
      "num_tokens": 95977039.0,
      "step": 42995
    },
    {
      "entropy": 5.359220504760742,
      "epoch": 4.829336777671702,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029117400149814953,
      "loss": 4.934,
      "mean_token_accuracy": 0.22414402216672896,
      "num_tokens": 95988041.0,
      "step": 43000
    },
    {
      "entropy": 5.2011377811431885,
      "epoch": 4.82989835457966,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002911339563879315,
      "loss": 4.6851,
      "mean_token_accuracy": 0.24904188066720961,
      "num_tokens": 95998950.0,
      "step": 43005
    },
    {
      "entropy": 5.219040441513061,
      "epoch": 4.830459931487617,
      "grad_norm": 1.0,
      "learning_rate": 0.0002910939107640019,
      "loss": 4.7545,
      "mean_token_accuracy": 0.23187455534934998,
      "num_tokens": 96010380.0,
      "step": 43010
    },
    {
      "entropy": 5.209896564483643,
      "epoch": 4.831021508395574,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000291053864627636,
      "loss": 4.7904,
      "mean_token_accuracy": 0.23090133517980577,
      "num_tokens": 96021860.0,
      "step": 43015
    },
    {
      "entropy": 5.278688955307007,
      "epoch": 4.831583085303532,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00029101381798010885,
      "loss": 4.7623,
      "mean_token_accuracy": 0.23570869415998458,
      "num_tokens": 96034127.0,
      "step": 43020
    },
    {
      "entropy": 5.208369255065918,
      "epoch": 4.83214466221149,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002909737708226954,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23247303813695908,
      "num_tokens": 96045029.0,
      "step": 43025
    },
    {
      "entropy": 5.25722918510437,
      "epoch": 4.832706239119448,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029093372315667105,
      "loss": 4.7471,
      "mean_token_accuracy": 0.2373448744416237,
      "num_tokens": 96055409.0,
      "step": 43030
    },
    {
      "entropy": 5.278338003158569,
      "epoch": 4.833267816027405,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029089367498331075,
      "loss": 4.7102,
      "mean_token_accuracy": 0.23695811927318572,
      "num_tokens": 96064928.0,
      "step": 43035
    },
    {
      "entropy": 5.30109429359436,
      "epoch": 4.833829392935362,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00029085362630388965,
      "loss": 4.8638,
      "mean_token_accuracy": 0.2374212622642517,
      "num_tokens": 96077111.0,
      "step": 43040
    },
    {
      "entropy": 5.188708734512329,
      "epoch": 4.83439096984332,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002908135771196829,
      "loss": 4.6815,
      "mean_token_accuracy": 0.2476551666855812,
      "num_tokens": 96088850.0,
      "step": 43045
    },
    {
      "entropy": 5.23883843421936,
      "epoch": 4.8349525467512775,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002907735274319657,
      "loss": 4.7708,
      "mean_token_accuracy": 0.2390112817287445,
      "num_tokens": 96099273.0,
      "step": 43050
    },
    {
      "entropy": 5.239950180053711,
      "epoch": 4.835514123659236,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002907334772420134,
      "loss": 4.8498,
      "mean_token_accuracy": 0.22925185561180114,
      "num_tokens": 96111332.0,
      "step": 43055
    },
    {
      "entropy": 5.240479135513306,
      "epoch": 4.836075700567193,
      "grad_norm": 1.015625,
      "learning_rate": 0.000290693426551101,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23392106741666793,
      "num_tokens": 96122225.0,
      "step": 43060
    },
    {
      "entropy": 5.23611159324646,
      "epoch": 4.83663727747515,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029065337536050384,
      "loss": 4.7608,
      "mean_token_accuracy": 0.24600140899419784,
      "num_tokens": 96133465.0,
      "step": 43065
    },
    {
      "entropy": 5.195169496536255,
      "epoch": 4.837198854383108,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029061332367149705,
      "loss": 4.7039,
      "mean_token_accuracy": 0.24217026233673095,
      "num_tokens": 96145036.0,
      "step": 43070
    },
    {
      "entropy": 5.113138151168823,
      "epoch": 4.837760431291065,
      "grad_norm": 0.953125,
      "learning_rate": 0.00029057327148535614,
      "loss": 4.6258,
      "mean_token_accuracy": 0.24867565631866456,
      "num_tokens": 96154986.0,
      "step": 43075
    },
    {
      "entropy": 5.265556287765503,
      "epoch": 4.838322008199023,
      "grad_norm": 0.96484375,
      "learning_rate": 0.000290533218803356,
      "loss": 4.8716,
      "mean_token_accuracy": 0.22989531457424164,
      "num_tokens": 96166271.0,
      "step": 43080
    },
    {
      "entropy": 5.277633714675903,
      "epoch": 4.838883585106981,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00029049316562677235,
      "loss": 4.8661,
      "mean_token_accuracy": 0.2393007308244705,
      "num_tokens": 96177203.0,
      "step": 43085
    },
    {
      "entropy": 5.3136022090911865,
      "epoch": 4.839445162014938,
      "grad_norm": 0.921875,
      "learning_rate": 0.00029045311195688016,
      "loss": 4.8095,
      "mean_token_accuracy": 0.23536953032016755,
      "num_tokens": 96188749.0,
      "step": 43090
    },
    {
      "entropy": 5.286208963394165,
      "epoch": 4.840006738922895,
      "grad_norm": 1.09375,
      "learning_rate": 0.000290413057794955,
      "loss": 4.8024,
      "mean_token_accuracy": 0.23577251136302949,
      "num_tokens": 96198701.0,
      "step": 43095
    },
    {
      "entropy": 5.35704870223999,
      "epoch": 4.840568315830853,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00029037300314227205,
      "loss": 4.9426,
      "mean_token_accuracy": 0.23325861096382142,
      "num_tokens": 96211016.0,
      "step": 43100
    },
    {
      "entropy": 5.28086667060852,
      "epoch": 4.8411298927388104,
      "grad_norm": 1.046875,
      "learning_rate": 0.00029033294800010666,
      "loss": 4.8324,
      "mean_token_accuracy": 0.23183769583702088,
      "num_tokens": 96222449.0,
      "step": 43105
    },
    {
      "entropy": 5.176928758621216,
      "epoch": 4.8416914696467686,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00029029289236973436,
      "loss": 4.7607,
      "mean_token_accuracy": 0.2376435339450836,
      "num_tokens": 96234434.0,
      "step": 43110
    },
    {
      "entropy": 5.213658857345581,
      "epoch": 4.842253046554726,
      "grad_norm": 1.03125,
      "learning_rate": 0.00029025283625243043,
      "loss": 4.7699,
      "mean_token_accuracy": 0.2336507633328438,
      "num_tokens": 96245612.0,
      "step": 43115
    },
    {
      "entropy": 5.245091104507447,
      "epoch": 4.842814623462683,
      "grad_norm": 1.078125,
      "learning_rate": 0.00029021277964947023,
      "loss": 4.667,
      "mean_token_accuracy": 0.24134237170219422,
      "num_tokens": 96256790.0,
      "step": 43120
    },
    {
      "entropy": 5.298769092559814,
      "epoch": 4.843376200370641,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00029017272256212927,
      "loss": 4.8267,
      "mean_token_accuracy": 0.23258245438337327,
      "num_tokens": 96267562.0,
      "step": 43125
    },
    {
      "entropy": 5.214736986160278,
      "epoch": 4.843937777278598,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002901326649916829,
      "loss": 4.7136,
      "mean_token_accuracy": 0.2384059876203537,
      "num_tokens": 96278469.0,
      "step": 43130
    },
    {
      "entropy": 5.314862442016602,
      "epoch": 4.8444993541865555,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00029009260693940665,
      "loss": 4.8939,
      "mean_token_accuracy": 0.23403248190879822,
      "num_tokens": 96289460.0,
      "step": 43135
    },
    {
      "entropy": 5.2976805686950685,
      "epoch": 4.845060931094514,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002900525484065759,
      "loss": 4.8251,
      "mean_token_accuracy": 0.23264504820108414,
      "num_tokens": 96300963.0,
      "step": 43140
    },
    {
      "entropy": 5.2076362609863285,
      "epoch": 4.845622508002471,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00029001248939446617,
      "loss": 4.6722,
      "mean_token_accuracy": 0.24140622913837434,
      "num_tokens": 96311521.0,
      "step": 43145
    },
    {
      "entropy": 5.228193473815918,
      "epoch": 4.846184084910428,
      "grad_norm": 1.03125,
      "learning_rate": 0.000289972429904353,
      "loss": 4.8271,
      "mean_token_accuracy": 0.23142098486423493,
      "num_tokens": 96321959.0,
      "step": 43150
    },
    {
      "entropy": 5.372592306137085,
      "epoch": 4.846745661818386,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028993236993751174,
      "loss": 4.9988,
      "mean_token_accuracy": 0.22707874327898026,
      "num_tokens": 96333916.0,
      "step": 43155
    },
    {
      "entropy": 5.302463674545288,
      "epoch": 4.847307238726343,
      "grad_norm": 1.140625,
      "learning_rate": 0.00028989230949521817,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23554083108901977,
      "num_tokens": 96345081.0,
      "step": 43160
    },
    {
      "entropy": 5.177349281311035,
      "epoch": 4.8478688156343015,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028985224857874765,
      "loss": 4.738,
      "mean_token_accuracy": 0.23044031411409377,
      "num_tokens": 96356394.0,
      "step": 43165
    },
    {
      "entropy": 5.121951580047607,
      "epoch": 4.848430392542259,
      "grad_norm": 1.0,
      "learning_rate": 0.00028981218718937573,
      "loss": 4.7053,
      "mean_token_accuracy": 0.23492421656847,
      "num_tokens": 96366952.0,
      "step": 43170
    },
    {
      "entropy": 5.254810571670532,
      "epoch": 4.848991969450216,
      "grad_norm": 0.921875,
      "learning_rate": 0.00028977212532837804,
      "loss": 4.8481,
      "mean_token_accuracy": 0.22561053037643433,
      "num_tokens": 96378165.0,
      "step": 43175
    },
    {
      "entropy": 5.363896369934082,
      "epoch": 4.849553546358174,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002897320629970302,
      "loss": 4.8163,
      "mean_token_accuracy": 0.24588999152183533,
      "num_tokens": 96390606.0,
      "step": 43180
    },
    {
      "entropy": 5.201605749130249,
      "epoch": 4.850115123266131,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028969200019660774,
      "loss": 4.7041,
      "mean_token_accuracy": 0.23959688395261763,
      "num_tokens": 96401089.0,
      "step": 43185
    },
    {
      "entropy": 5.219020175933838,
      "epoch": 4.8506767001740885,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002896519369283862,
      "loss": 4.8783,
      "mean_token_accuracy": 0.23425260484218596,
      "num_tokens": 96411908.0,
      "step": 43190
    },
    {
      "entropy": 5.203458213806153,
      "epoch": 4.851238277082047,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002896118731936414,
      "loss": 4.7311,
      "mean_token_accuracy": 0.23703682869672776,
      "num_tokens": 96422859.0,
      "step": 43195
    },
    {
      "entropy": 5.277085018157959,
      "epoch": 4.851799853990004,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002895718089936489,
      "loss": 4.7835,
      "mean_token_accuracy": 0.23770980089902877,
      "num_tokens": 96435122.0,
      "step": 43200
    },
    {
      "entropy": 5.310477304458618,
      "epoch": 4.852361430897961,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002895317443296843,
      "loss": 4.8331,
      "mean_token_accuracy": 0.22862143069505692,
      "num_tokens": 96446916.0,
      "step": 43205
    },
    {
      "entropy": 5.226219129562378,
      "epoch": 4.852923007805919,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028949167920302337,
      "loss": 4.7718,
      "mean_token_accuracy": 0.2282156378030777,
      "num_tokens": 96458241.0,
      "step": 43210
    },
    {
      "entropy": 5.271644735336304,
      "epoch": 4.853484584713876,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00028945161361494177,
      "loss": 4.7906,
      "mean_token_accuracy": 0.23047894835472107,
      "num_tokens": 96468581.0,
      "step": 43215
    },
    {
      "entropy": 5.234185218811035,
      "epoch": 4.854046161621834,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002894115475667151,
      "loss": 4.7451,
      "mean_token_accuracy": 0.24118465632200242,
      "num_tokens": 96479119.0,
      "step": 43220
    },
    {
      "entropy": 5.266829967498779,
      "epoch": 4.854607738529792,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028937148105961924,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22701949924230574,
      "num_tokens": 96490306.0,
      "step": 43225
    },
    {
      "entropy": 5.15960545539856,
      "epoch": 4.855169315437749,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002893314140949298,
      "loss": 4.7407,
      "mean_token_accuracy": 0.24435730278491974,
      "num_tokens": 96501222.0,
      "step": 43230
    },
    {
      "entropy": 5.243821096420288,
      "epoch": 4.855730892345707,
      "grad_norm": 1.0,
      "learning_rate": 0.00028929134667392265,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2310175985097885,
      "num_tokens": 96513283.0,
      "step": 43235
    },
    {
      "entropy": 5.277310848236084,
      "epoch": 4.856292469253664,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028925127879787343,
      "loss": 4.8074,
      "mean_token_accuracy": 0.23801844120025634,
      "num_tokens": 96523546.0,
      "step": 43240
    },
    {
      "entropy": 5.256101512908936,
      "epoch": 4.856854046161622,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002892112104680579,
      "loss": 4.7953,
      "mean_token_accuracy": 0.24065771251916884,
      "num_tokens": 96534906.0,
      "step": 43245
    },
    {
      "entropy": 5.18895378112793,
      "epoch": 4.8574156230695795,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028917114168575196,
      "loss": 4.7405,
      "mean_token_accuracy": 0.23670633435249328,
      "num_tokens": 96547029.0,
      "step": 43250
    },
    {
      "entropy": 5.257447195053101,
      "epoch": 4.857977199977537,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002891310724522314,
      "loss": 4.7134,
      "mean_token_accuracy": 0.24289303421974182,
      "num_tokens": 96558632.0,
      "step": 43255
    },
    {
      "entropy": 5.24356050491333,
      "epoch": 4.858538776885495,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028909100276877203,
      "loss": 4.8118,
      "mean_token_accuracy": 0.2426719680428505,
      "num_tokens": 96568644.0,
      "step": 43260
    },
    {
      "entropy": 5.187562608718872,
      "epoch": 4.859100353793452,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028905093263664966,
      "loss": 4.7292,
      "mean_token_accuracy": 0.23954353779554366,
      "num_tokens": 96578844.0,
      "step": 43265
    },
    {
      "entropy": 5.2348321914672855,
      "epoch": 4.859661930701409,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028901086205714016,
      "loss": 4.7708,
      "mean_token_accuracy": 0.2323597714304924,
      "num_tokens": 96590715.0,
      "step": 43270
    },
    {
      "entropy": 5.2316090106964115,
      "epoch": 4.860223507609367,
      "grad_norm": 1.125,
      "learning_rate": 0.00028897079103151936,
      "loss": 4.7614,
      "mean_token_accuracy": 0.23769118040800094,
      "num_tokens": 96601001.0,
      "step": 43275
    },
    {
      "entropy": 5.209568881988526,
      "epoch": 4.860785084517325,
      "grad_norm": 1.0,
      "learning_rate": 0.00028893071956106315,
      "loss": 4.7109,
      "mean_token_accuracy": 0.24194251745939255,
      "num_tokens": 96612591.0,
      "step": 43280
    },
    {
      "entropy": 5.159500646591186,
      "epoch": 4.861346661425282,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028889064764704735,
      "loss": 4.7827,
      "mean_token_accuracy": 0.23705492913722992,
      "num_tokens": 96624353.0,
      "step": 43285
    },
    {
      "entropy": 5.2539325714111325,
      "epoch": 4.86190823833324,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028885057529074804,
      "loss": 4.7425,
      "mean_token_accuracy": 0.23493810445070268,
      "num_tokens": 96634293.0,
      "step": 43290
    },
    {
      "entropy": 5.34385929107666,
      "epoch": 4.862469815241197,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028881050249344103,
      "loss": 4.8673,
      "mean_token_accuracy": 0.22613053470849992,
      "num_tokens": 96645308.0,
      "step": 43295
    },
    {
      "entropy": 5.252020692825317,
      "epoch": 4.863031392149155,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002887704292564022,
      "loss": 4.8305,
      "mean_token_accuracy": 0.23137160241603852,
      "num_tokens": 96655198.0,
      "step": 43300
    },
    {
      "entropy": 5.1854010105133055,
      "epoch": 4.8635929690571125,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028873035558090766,
      "loss": 4.7695,
      "mean_token_accuracy": 0.23809419572353363,
      "num_tokens": 96665753.0,
      "step": 43305
    },
    {
      "entropy": 5.2687455177307125,
      "epoch": 4.86415454596507,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002886902814682332,
      "loss": 4.7809,
      "mean_token_accuracy": 0.237637896835804,
      "num_tokens": 96676038.0,
      "step": 43310
    },
    {
      "entropy": 5.309824657440186,
      "epoch": 4.864716122873028,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028865020691965485,
      "loss": 4.7296,
      "mean_token_accuracy": 0.23640351444482804,
      "num_tokens": 96687893.0,
      "step": 43315
    },
    {
      "entropy": 5.306502628326416,
      "epoch": 4.865277699780985,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028861013193644865,
      "loss": 4.8697,
      "mean_token_accuracy": 0.2295312151312828,
      "num_tokens": 96698937.0,
      "step": 43320
    },
    {
      "entropy": 5.225679111480713,
      "epoch": 4.865839276688942,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002885700565198906,
      "loss": 4.7526,
      "mean_token_accuracy": 0.23506538718938827,
      "num_tokens": 96710117.0,
      "step": 43325
    },
    {
      "entropy": 5.176984739303589,
      "epoch": 4.8664008535969,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028852998067125675,
      "loss": 4.7077,
      "mean_token_accuracy": 0.24178897887468337,
      "num_tokens": 96719831.0,
      "step": 43330
    },
    {
      "entropy": 5.222787714004516,
      "epoch": 4.8669624305048576,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002884899043918229,
      "loss": 4.8962,
      "mean_token_accuracy": 0.2279340460896492,
      "num_tokens": 96732127.0,
      "step": 43335
    },
    {
      "entropy": 5.317231178283691,
      "epoch": 4.867524007412815,
      "grad_norm": 1.0,
      "learning_rate": 0.0002884498276828653,
      "loss": 4.9054,
      "mean_token_accuracy": 0.22546141296625138,
      "num_tokens": 96744406.0,
      "step": 43340
    },
    {
      "entropy": 5.2930686473846436,
      "epoch": 4.868085584320773,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028840975054566007,
      "loss": 4.7637,
      "mean_token_accuracy": 0.24602360874414445,
      "num_tokens": 96754810.0,
      "step": 43345
    },
    {
      "entropy": 5.214267635345459,
      "epoch": 4.86864716122873,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028836967298148317,
      "loss": 4.7058,
      "mean_token_accuracy": 0.2380739137530327,
      "num_tokens": 96765520.0,
      "step": 43350
    },
    {
      "entropy": 5.313134288787841,
      "epoch": 4.869208738136688,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028832959499161065,
      "loss": 4.9208,
      "mean_token_accuracy": 0.2300558254122734,
      "num_tokens": 96777262.0,
      "step": 43355
    },
    {
      "entropy": 5.252005624771118,
      "epoch": 4.869770315044645,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002882895165773187,
      "loss": 4.8021,
      "mean_token_accuracy": 0.23271918296813965,
      "num_tokens": 96788452.0,
      "step": 43360
    },
    {
      "entropy": 5.2065071105957035,
      "epoch": 4.870331891952603,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002882494377398833,
      "loss": 4.7584,
      "mean_token_accuracy": 0.23804692625999452,
      "num_tokens": 96799402.0,
      "step": 43365
    },
    {
      "entropy": 5.291556549072266,
      "epoch": 4.870893468860561,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028820935848058066,
      "loss": 4.8779,
      "mean_token_accuracy": 0.23590587228536605,
      "num_tokens": 96810467.0,
      "step": 43370
    },
    {
      "entropy": 5.2170844078063965,
      "epoch": 4.871455045768518,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000288169278800687,
      "loss": 4.7038,
      "mean_token_accuracy": 0.2427579715847969,
      "num_tokens": 96821480.0,
      "step": 43375
    },
    {
      "entropy": 5.1930571556091305,
      "epoch": 4.872016622676475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028812919870147837,
      "loss": 4.6993,
      "mean_token_accuracy": 0.2491835355758667,
      "num_tokens": 96832473.0,
      "step": 43380
    },
    {
      "entropy": 5.242822980880737,
      "epoch": 4.872578199584433,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028808911818423096,
      "loss": 4.7749,
      "mean_token_accuracy": 0.235091832280159,
      "num_tokens": 96844019.0,
      "step": 43385
    },
    {
      "entropy": 5.19097490310669,
      "epoch": 4.8731397764923905,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028804903725022094,
      "loss": 4.7712,
      "mean_token_accuracy": 0.2346346750855446,
      "num_tokens": 96853902.0,
      "step": 43390
    },
    {
      "entropy": 5.287706565856934,
      "epoch": 4.873701353400348,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002880089559007245,
      "loss": 4.7997,
      "mean_token_accuracy": 0.23829675763845443,
      "num_tokens": 96865066.0,
      "step": 43395
    },
    {
      "entropy": 5.287779378890991,
      "epoch": 4.874262930308306,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002879688741370179,
      "loss": 4.8594,
      "mean_token_accuracy": 0.23133019506931304,
      "num_tokens": 96876329.0,
      "step": 43400
    },
    {
      "entropy": 5.255696582794189,
      "epoch": 4.874824507216263,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002879287919603773,
      "loss": 4.7531,
      "mean_token_accuracy": 0.23611065447330476,
      "num_tokens": 96887108.0,
      "step": 43405
    },
    {
      "entropy": 5.23940076828003,
      "epoch": 4.875386084124221,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028788870937207896,
      "loss": 4.791,
      "mean_token_accuracy": 0.2342168375849724,
      "num_tokens": 96898956.0,
      "step": 43410
    },
    {
      "entropy": 5.246927547454834,
      "epoch": 4.875947661032178,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002878486263733991,
      "loss": 4.7602,
      "mean_token_accuracy": 0.23647944033145904,
      "num_tokens": 96910889.0,
      "step": 43415
    },
    {
      "entropy": 5.219026947021485,
      "epoch": 4.876509237940136,
      "grad_norm": 0.9609375,
      "learning_rate": 0.000287808542965614,
      "loss": 4.6991,
      "mean_token_accuracy": 0.23823823779821396,
      "num_tokens": 96921682.0,
      "step": 43420
    },
    {
      "entropy": 5.230688428878784,
      "epoch": 4.877070814848094,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028776845914999996,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24247609972953796,
      "num_tokens": 96933502.0,
      "step": 43425
    },
    {
      "entropy": 5.128164339065552,
      "epoch": 4.877632391756051,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002877283749278332,
      "loss": 4.6677,
      "mean_token_accuracy": 0.24352139830589295,
      "num_tokens": 96945621.0,
      "step": 43430
    },
    {
      "entropy": 5.316097640991211,
      "epoch": 4.878193968664009,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028768829030039,
      "loss": 4.9056,
      "mean_token_accuracy": 0.23072812259197234,
      "num_tokens": 96957164.0,
      "step": 43435
    },
    {
      "entropy": 5.305681419372559,
      "epoch": 4.878755545571966,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028764820526894675,
      "loss": 4.8175,
      "mean_token_accuracy": 0.2294481486082077,
      "num_tokens": 96968646.0,
      "step": 43440
    },
    {
      "entropy": 5.275383186340332,
      "epoch": 4.879317122479923,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028760811983477976,
      "loss": 4.7343,
      "mean_token_accuracy": 0.23949117958545685,
      "num_tokens": 96979855.0,
      "step": 43445
    },
    {
      "entropy": 5.214121389389038,
      "epoch": 4.8798786993878815,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028756803399916534,
      "loss": 4.7244,
      "mean_token_accuracy": 0.24909638017416,
      "num_tokens": 96989603.0,
      "step": 43450
    },
    {
      "entropy": 5.209624481201172,
      "epoch": 4.880440276295839,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002875279477633798,
      "loss": 4.7491,
      "mean_token_accuracy": 0.24111688882112503,
      "num_tokens": 96999511.0,
      "step": 43455
    },
    {
      "entropy": 5.251860618591309,
      "epoch": 4.881001853203796,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002874878611286995,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23861589133739472,
      "num_tokens": 97011210.0,
      "step": 43460
    },
    {
      "entropy": 5.2446129322052,
      "epoch": 4.881563430111754,
      "grad_norm": 1.234375,
      "learning_rate": 0.000287447774096401,
      "loss": 4.8091,
      "mean_token_accuracy": 0.244111005961895,
      "num_tokens": 97022351.0,
      "step": 43465
    },
    {
      "entropy": 5.2593152046203615,
      "epoch": 4.882125007019711,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028740768666776046,
      "loss": 4.773,
      "mean_token_accuracy": 0.2352570801973343,
      "num_tokens": 97033269.0,
      "step": 43470
    },
    {
      "entropy": 5.254382514953614,
      "epoch": 4.8826865839276685,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028736759884405433,
      "loss": 4.816,
      "mean_token_accuracy": 0.23192360252141953,
      "num_tokens": 97045278.0,
      "step": 43475
    },
    {
      "entropy": 5.216646814346314,
      "epoch": 4.883248160835627,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002873275106265591,
      "loss": 4.7071,
      "mean_token_accuracy": 0.2469874680042267,
      "num_tokens": 97055617.0,
      "step": 43480
    },
    {
      "entropy": 5.3040906429290775,
      "epoch": 4.883809737743584,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028728742201655116,
      "loss": 4.8885,
      "mean_token_accuracy": 0.236313833296299,
      "num_tokens": 97068815.0,
      "step": 43485
    },
    {
      "entropy": 5.264676141738891,
      "epoch": 4.884371314651542,
      "grad_norm": 1.109375,
      "learning_rate": 0.000287247333015307,
      "loss": 4.8902,
      "mean_token_accuracy": 0.22733744233846664,
      "num_tokens": 97078812.0,
      "step": 43490
    },
    {
      "entropy": 5.171149778366089,
      "epoch": 4.884932891559499,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002872072436241029,
      "loss": 4.6797,
      "mean_token_accuracy": 0.25477910339832305,
      "num_tokens": 97090423.0,
      "step": 43495
    },
    {
      "entropy": 5.167397356033325,
      "epoch": 4.885494468467456,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002871671538442155,
      "loss": 4.6366,
      "mean_token_accuracy": 0.24528705775737764,
      "num_tokens": 97100756.0,
      "step": 43500
    },
    {
      "entropy": 5.150864601135254,
      "epoch": 4.8860560453754145,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002871270636769212,
      "loss": 4.7361,
      "mean_token_accuracy": 0.24350023716688157,
      "num_tokens": 97111016.0,
      "step": 43505
    },
    {
      "entropy": 5.267702436447143,
      "epoch": 4.886617622283372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002870869731234965,
      "loss": 4.7643,
      "mean_token_accuracy": 0.23264165818691254,
      "num_tokens": 97122148.0,
      "step": 43510
    },
    {
      "entropy": 5.252947664260864,
      "epoch": 4.887179199191329,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002870468821852178,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2342359259724617,
      "num_tokens": 97133301.0,
      "step": 43515
    },
    {
      "entropy": 5.227630615234375,
      "epoch": 4.887740776099287,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028700679086336175,
      "loss": 4.7665,
      "mean_token_accuracy": 0.23371610343456267,
      "num_tokens": 97144637.0,
      "step": 43520
    },
    {
      "entropy": 5.168023872375488,
      "epoch": 4.888302353007244,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002869666991592048,
      "loss": 4.6048,
      "mean_token_accuracy": 0.2457760527729988,
      "num_tokens": 97156047.0,
      "step": 43525
    },
    {
      "entropy": 5.201550817489624,
      "epoch": 4.8888639299152015,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002869266070740235,
      "loss": 4.7559,
      "mean_token_accuracy": 0.23339523375034332,
      "num_tokens": 97167767.0,
      "step": 43530
    },
    {
      "entropy": 5.1376869678497314,
      "epoch": 4.88942550682316,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002868865146090944,
      "loss": 4.7297,
      "mean_token_accuracy": 0.23497493714094161,
      "num_tokens": 97179326.0,
      "step": 43535
    },
    {
      "entropy": 5.229129600524902,
      "epoch": 4.889987083731117,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00028684642176569414,
      "loss": 4.8225,
      "mean_token_accuracy": 0.23999102115631105,
      "num_tokens": 97190437.0,
      "step": 43540
    },
    {
      "entropy": 5.180620765686035,
      "epoch": 4.890548660639075,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002868063285450992,
      "loss": 4.663,
      "mean_token_accuracy": 0.24555125534534455,
      "num_tokens": 97200913.0,
      "step": 43545
    },
    {
      "entropy": 5.1841107368469235,
      "epoch": 4.891110237547032,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002867662349485861,
      "loss": 4.6904,
      "mean_token_accuracy": 0.240912526845932,
      "num_tokens": 97210935.0,
      "step": 43550
    },
    {
      "entropy": 5.214673852920532,
      "epoch": 4.891671814454989,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028672614097743156,
      "loss": 4.7127,
      "mean_token_accuracy": 0.23965777307748795,
      "num_tokens": 97221568.0,
      "step": 43555
    },
    {
      "entropy": 5.226635217666626,
      "epoch": 4.892233391362947,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028668604663291213,
      "loss": 4.7582,
      "mean_token_accuracy": 0.23441214114427567,
      "num_tokens": 97232149.0,
      "step": 43560
    },
    {
      "entropy": 5.246486377716065,
      "epoch": 4.892794968270905,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028664595191630455,
      "loss": 4.8244,
      "mean_token_accuracy": 0.231528902053833,
      "num_tokens": 97242800.0,
      "step": 43565
    },
    {
      "entropy": 5.229920482635498,
      "epoch": 4.893356545178862,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028660585682888524,
      "loss": 4.8059,
      "mean_token_accuracy": 0.23874802887439728,
      "num_tokens": 97253681.0,
      "step": 43570
    },
    {
      "entropy": 5.2969334602355955,
      "epoch": 4.89391812208682,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00028656576137193096,
      "loss": 4.8212,
      "mean_token_accuracy": 0.23187200278043746,
      "num_tokens": 97264949.0,
      "step": 43575
    },
    {
      "entropy": 5.283009719848633,
      "epoch": 4.894479698994777,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002865256655467184,
      "loss": 4.7773,
      "mean_token_accuracy": 0.24136568009853362,
      "num_tokens": 97276670.0,
      "step": 43580
    },
    {
      "entropy": 5.1681030750274655,
      "epoch": 4.895041275902734,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002864855693545241,
      "loss": 4.6503,
      "mean_token_accuracy": 0.24184062331914902,
      "num_tokens": 97287134.0,
      "step": 43585
    },
    {
      "entropy": 5.187997579574585,
      "epoch": 4.8956028528106925,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028644547279662485,
      "loss": 4.8017,
      "mean_token_accuracy": 0.23711936175823212,
      "num_tokens": 97298250.0,
      "step": 43590
    },
    {
      "entropy": 5.1645957946777346,
      "epoch": 4.89616442971865,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028640537587429734,
      "loss": 4.7031,
      "mean_token_accuracy": 0.23433919697999955,
      "num_tokens": 97309144.0,
      "step": 43595
    },
    {
      "entropy": 5.258561801910401,
      "epoch": 4.896726006626608,
      "grad_norm": 1.0,
      "learning_rate": 0.0002863652785888182,
      "loss": 4.7996,
      "mean_token_accuracy": 0.23285835683345796,
      "num_tokens": 97320586.0,
      "step": 43600
    },
    {
      "entropy": 5.120062351226807,
      "epoch": 4.897287583534565,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002863251809414642,
      "loss": 4.693,
      "mean_token_accuracy": 0.24315644353628157,
      "num_tokens": 97330924.0,
      "step": 43605
    },
    {
      "entropy": 5.271547555923462,
      "epoch": 4.897849160442522,
      "grad_norm": 1.046875,
      "learning_rate": 0.000286285082933512,
      "loss": 4.8481,
      "mean_token_accuracy": 0.23620101660490037,
      "num_tokens": 97341926.0,
      "step": 43610
    },
    {
      "entropy": 5.282849073410034,
      "epoch": 4.89841073735048,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002862449845662385,
      "loss": 4.8626,
      "mean_token_accuracy": 0.22765442579984665,
      "num_tokens": 97353809.0,
      "step": 43615
    },
    {
      "entropy": 5.274544525146484,
      "epoch": 4.898972314258438,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002862048858409201,
      "loss": 4.8378,
      "mean_token_accuracy": 0.23503336012363435,
      "num_tokens": 97365317.0,
      "step": 43620
    },
    {
      "entropy": 5.280163860321045,
      "epoch": 4.899533891166396,
      "grad_norm": 1.21875,
      "learning_rate": 0.000286164786758834,
      "loss": 4.7531,
      "mean_token_accuracy": 0.2360938861966133,
      "num_tokens": 97375581.0,
      "step": 43625
    },
    {
      "entropy": 5.416835117340088,
      "epoch": 4.900095468074353,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028612468732125667,
      "loss": 4.915,
      "mean_token_accuracy": 0.236959046125412,
      "num_tokens": 97387371.0,
      "step": 43630
    },
    {
      "entropy": 5.200968980789185,
      "epoch": 4.90065704498231,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002860845875294651,
      "loss": 4.7257,
      "mean_token_accuracy": 0.24403670877218248,
      "num_tokens": 97398929.0,
      "step": 43635
    },
    {
      "entropy": 5.353480291366577,
      "epoch": 4.901218621890268,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002860444873847358,
      "loss": 4.9573,
      "mean_token_accuracy": 0.22603852599859237,
      "num_tokens": 97410469.0,
      "step": 43640
    },
    {
      "entropy": 5.202486848831176,
      "epoch": 4.9017801987982255,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028600438688834587,
      "loss": 4.7104,
      "mean_token_accuracy": 0.23910336792469025,
      "num_tokens": 97422253.0,
      "step": 43645
    },
    {
      "entropy": 5.294445610046386,
      "epoch": 4.902341775706183,
      "grad_norm": 0.98046875,
      "learning_rate": 0.000285964286041572,
      "loss": 4.7946,
      "mean_token_accuracy": 0.22731930017471313,
      "num_tokens": 97433384.0,
      "step": 43650
    },
    {
      "entropy": 5.26914792060852,
      "epoch": 4.902903352614141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028592418484569095,
      "loss": 4.7799,
      "mean_token_accuracy": 0.23553137630224227,
      "num_tokens": 97444755.0,
      "step": 43655
    },
    {
      "entropy": 5.114477491378784,
      "epoch": 4.903464929522098,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002858840833019795,
      "loss": 4.6199,
      "mean_token_accuracy": 0.24280341863632202,
      "num_tokens": 97456156.0,
      "step": 43660
    },
    {
      "entropy": 5.187962436676026,
      "epoch": 4.904026506430055,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002858439814117148,
      "loss": 4.7325,
      "mean_token_accuracy": 0.24281354993581772,
      "num_tokens": 97467494.0,
      "step": 43665
    },
    {
      "entropy": 5.178075742721558,
      "epoch": 4.904588083338013,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028580387917617343,
      "loss": 4.6948,
      "mean_token_accuracy": 0.2439817428588867,
      "num_tokens": 97478317.0,
      "step": 43670
    },
    {
      "entropy": 5.222591733932495,
      "epoch": 4.9051496602459705,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002857637765966324,
      "loss": 4.7501,
      "mean_token_accuracy": 0.24476616233587264,
      "num_tokens": 97488530.0,
      "step": 43675
    },
    {
      "entropy": 5.19831109046936,
      "epoch": 4.905711237153929,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002857236736743685,
      "loss": 4.7022,
      "mean_token_accuracy": 0.2452893152832985,
      "num_tokens": 97501148.0,
      "step": 43680
    },
    {
      "entropy": 5.2316046237945555,
      "epoch": 4.906272814061886,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002856835704106587,
      "loss": 4.7728,
      "mean_token_accuracy": 0.2372315600514412,
      "num_tokens": 97512136.0,
      "step": 43685
    },
    {
      "entropy": 5.251141929626465,
      "epoch": 4.906834390969843,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028564346680677995,
      "loss": 4.7436,
      "mean_token_accuracy": 0.24378256797790526,
      "num_tokens": 97523787.0,
      "step": 43690
    },
    {
      "entropy": 5.27055253982544,
      "epoch": 4.907395967877801,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002856033628640089,
      "loss": 4.7952,
      "mean_token_accuracy": 0.23299800455570222,
      "num_tokens": 97534851.0,
      "step": 43695
    },
    {
      "entropy": 5.191018533706665,
      "epoch": 4.907957544785758,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002855632585836228,
      "loss": 4.7088,
      "mean_token_accuracy": 0.23615113347768785,
      "num_tokens": 97545160.0,
      "step": 43700
    },
    {
      "entropy": 5.244176292419434,
      "epoch": 4.908519121693716,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028552315396689846,
      "loss": 4.8276,
      "mean_token_accuracy": 0.2322790414094925,
      "num_tokens": 97556321.0,
      "step": 43705
    },
    {
      "entropy": 5.274408197402954,
      "epoch": 4.909080698601674,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002854830490151128,
      "loss": 4.841,
      "mean_token_accuracy": 0.23017602562904357,
      "num_tokens": 97567406.0,
      "step": 43710
    },
    {
      "entropy": 5.241971969604492,
      "epoch": 4.909642275509631,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028544294372954276,
      "loss": 4.6975,
      "mean_token_accuracy": 0.24280092120170593,
      "num_tokens": 97577998.0,
      "step": 43715
    },
    {
      "entropy": 5.213908624649048,
      "epoch": 4.910203852417588,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002854028381114654,
      "loss": 4.7429,
      "mean_token_accuracy": 0.23965901881456375,
      "num_tokens": 97588262.0,
      "step": 43720
    },
    {
      "entropy": 5.161943531036377,
      "epoch": 4.910765429325546,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028536273216215756,
      "loss": 4.6597,
      "mean_token_accuracy": 0.24180734902620316,
      "num_tokens": 97600011.0,
      "step": 43725
    },
    {
      "entropy": 5.380242013931275,
      "epoch": 4.9113270062335035,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028532262588289636,
      "loss": 4.9289,
      "mean_token_accuracy": 0.23954717367887496,
      "num_tokens": 97612037.0,
      "step": 43730
    },
    {
      "entropy": 5.231942844390869,
      "epoch": 4.911888583141462,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028528251927495874,
      "loss": 4.678,
      "mean_token_accuracy": 0.243729867041111,
      "num_tokens": 97622643.0,
      "step": 43735
    },
    {
      "entropy": 5.205245447158814,
      "epoch": 4.912450160049419,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028524241233962176,
      "loss": 4.8337,
      "mean_token_accuracy": 0.2337718352675438,
      "num_tokens": 97634720.0,
      "step": 43740
    },
    {
      "entropy": 5.3258264541625975,
      "epoch": 4.913011736957376,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028520230507816236,
      "loss": 4.9519,
      "mean_token_accuracy": 0.22302308678627014,
      "num_tokens": 97646314.0,
      "step": 43745
    },
    {
      "entropy": 5.288826179504395,
      "epoch": 4.913573313865334,
      "grad_norm": 1.03125,
      "learning_rate": 0.00028516219749185755,
      "loss": 4.7539,
      "mean_token_accuracy": 0.24100474119186402,
      "num_tokens": 97657478.0,
      "step": 43750
    },
    {
      "entropy": 5.287932395935059,
      "epoch": 4.914134890773291,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002851220895819845,
      "loss": 4.7244,
      "mean_token_accuracy": 0.23965900540351867,
      "num_tokens": 97668084.0,
      "step": 43755
    },
    {
      "entropy": 5.223093700408936,
      "epoch": 4.914696467681249,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028508198134982016,
      "loss": 4.7628,
      "mean_token_accuracy": 0.23360619693994522,
      "num_tokens": 97678769.0,
      "step": 43760
    },
    {
      "entropy": 5.213558340072632,
      "epoch": 4.915258044589207,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028504187279664164,
      "loss": 4.8039,
      "mean_token_accuracy": 0.23101097345352173,
      "num_tokens": 97689587.0,
      "step": 43765
    },
    {
      "entropy": 5.237526273727417,
      "epoch": 4.915819621497164,
      "grad_norm": 1.0,
      "learning_rate": 0.000285001763923726,
      "loss": 4.8106,
      "mean_token_accuracy": 0.2380768969655037,
      "num_tokens": 97700334.0,
      "step": 43770
    },
    {
      "entropy": 5.2300373077392575,
      "epoch": 4.916381198405121,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002849616547323503,
      "loss": 4.715,
      "mean_token_accuracy": 0.24091078639030455,
      "num_tokens": 97713038.0,
      "step": 43775
    },
    {
      "entropy": 5.242444753646851,
      "epoch": 4.916942775313079,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028492154522379167,
      "loss": 4.8002,
      "mean_token_accuracy": 0.2358182355761528,
      "num_tokens": 97724651.0,
      "step": 43780
    },
    {
      "entropy": 5.205949831008911,
      "epoch": 4.917504352221036,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002848814353993273,
      "loss": 4.8386,
      "mean_token_accuracy": 0.23285646736621857,
      "num_tokens": 97735862.0,
      "step": 43785
    },
    {
      "entropy": 5.246805238723755,
      "epoch": 4.9180659291289945,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028484132526023404,
      "loss": 4.7778,
      "mean_token_accuracy": 0.23850328028202056,
      "num_tokens": 97747356.0,
      "step": 43790
    },
    {
      "entropy": 5.234715843200684,
      "epoch": 4.918627506036952,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002848012148077892,
      "loss": 4.7897,
      "mean_token_accuracy": 0.2327198565006256,
      "num_tokens": 97758670.0,
      "step": 43795
    },
    {
      "entropy": 5.276385879516601,
      "epoch": 4.919189082944909,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028476110404326984,
      "loss": 4.8446,
      "mean_token_accuracy": 0.23556251525878907,
      "num_tokens": 97770291.0,
      "step": 43800
    },
    {
      "entropy": 5.233238983154297,
      "epoch": 4.919750659852867,
      "grad_norm": 1.0,
      "learning_rate": 0.0002847209929679532,
      "loss": 4.7562,
      "mean_token_accuracy": 0.23777887970209122,
      "num_tokens": 97781567.0,
      "step": 43805
    },
    {
      "entropy": 5.195800685882569,
      "epoch": 4.920312236760824,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002846808815831164,
      "loss": 4.7253,
      "mean_token_accuracy": 0.2407629057765007,
      "num_tokens": 97792281.0,
      "step": 43810
    },
    {
      "entropy": 5.1013247013092045,
      "epoch": 4.920873813668782,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002846407698900365,
      "loss": 4.655,
      "mean_token_accuracy": 0.24087126106023787,
      "num_tokens": 97803606.0,
      "step": 43815
    },
    {
      "entropy": 5.188695001602173,
      "epoch": 4.92143539057674,
      "grad_norm": 0.953125,
      "learning_rate": 0.00028460065788999073,
      "loss": 4.7093,
      "mean_token_accuracy": 0.24201874136924745,
      "num_tokens": 97816024.0,
      "step": 43820
    },
    {
      "entropy": 5.226176309585571,
      "epoch": 4.921996967484697,
      "grad_norm": 1.0,
      "learning_rate": 0.0002845605455842564,
      "loss": 4.75,
      "mean_token_accuracy": 0.24148050993680953,
      "num_tokens": 97827121.0,
      "step": 43825
    },
    {
      "entropy": 5.22331862449646,
      "epoch": 4.922558544392655,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002845204329741106,
      "loss": 4.7699,
      "mean_token_accuracy": 0.2389939546585083,
      "num_tokens": 97838758.0,
      "step": 43830
    },
    {
      "entropy": 5.1336345195770265,
      "epoch": 4.923120121300612,
      "grad_norm": 1.0,
      "learning_rate": 0.0002844803200608304,
      "loss": 4.621,
      "mean_token_accuracy": 0.2502728372812271,
      "num_tokens": 97850208.0,
      "step": 43835
    },
    {
      "entropy": 5.1073096752166744,
      "epoch": 4.923681698208569,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002844402068456931,
      "loss": 4.6343,
      "mean_token_accuracy": 0.23942764848470688,
      "num_tokens": 97860210.0,
      "step": 43840
    },
    {
      "entropy": 5.204695081710815,
      "epoch": 4.9242432751165275,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002844000933299761,
      "loss": 4.7442,
      "mean_token_accuracy": 0.23649261593818666,
      "num_tokens": 97871348.0,
      "step": 43845
    },
    {
      "entropy": 5.244145154953003,
      "epoch": 4.924804852024485,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028435997951495643,
      "loss": 4.7875,
      "mean_token_accuracy": 0.23828830122947692,
      "num_tokens": 97882767.0,
      "step": 43850
    },
    {
      "entropy": 5.260208225250244,
      "epoch": 4.925366428932442,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002843198654019114,
      "loss": 4.8096,
      "mean_token_accuracy": 0.2359371468424797,
      "num_tokens": 97895690.0,
      "step": 43855
    },
    {
      "entropy": 5.238337230682373,
      "epoch": 4.9259280058404,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002842797509921182,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23964402228593826,
      "num_tokens": 97906951.0,
      "step": 43860
    },
    {
      "entropy": 5.2599921226501465,
      "epoch": 4.926489582748357,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028423963628685414,
      "loss": 4.8078,
      "mean_token_accuracy": 0.23406265676021576,
      "num_tokens": 97918538.0,
      "step": 43865
    },
    {
      "entropy": 5.205487680435181,
      "epoch": 4.927051159656315,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002841995212873965,
      "loss": 4.6877,
      "mean_token_accuracy": 0.2428694948554039,
      "num_tokens": 97930175.0,
      "step": 43870
    },
    {
      "entropy": 5.237417888641358,
      "epoch": 4.927612736564273,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002841594059950225,
      "loss": 4.813,
      "mean_token_accuracy": 0.23448692858219147,
      "num_tokens": 97940870.0,
      "step": 43875
    },
    {
      "entropy": 5.2173662185668945,
      "epoch": 4.92817431347223,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028411929041100947,
      "loss": 4.7463,
      "mean_token_accuracy": 0.23625297099351883,
      "num_tokens": 97951492.0,
      "step": 43880
    },
    {
      "entropy": 5.190489196777344,
      "epoch": 4.928735890380188,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002840791745366347,
      "loss": 4.7268,
      "mean_token_accuracy": 0.24281008988618852,
      "num_tokens": 97962243.0,
      "step": 43885
    },
    {
      "entropy": 5.21227011680603,
      "epoch": 4.929297467288145,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028403905837317547,
      "loss": 4.7144,
      "mean_token_accuracy": 0.24460788816213608,
      "num_tokens": 97973150.0,
      "step": 43890
    },
    {
      "entropy": 5.201372480392456,
      "epoch": 4.929859044196102,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00028399894192190916,
      "loss": 4.7704,
      "mean_token_accuracy": 0.23370704054832458,
      "num_tokens": 97985314.0,
      "step": 43895
    },
    {
      "entropy": 5.212586069107056,
      "epoch": 4.93042062110406,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028395882518411305,
      "loss": 4.776,
      "mean_token_accuracy": 0.2379367396235466,
      "num_tokens": 97997065.0,
      "step": 43900
    },
    {
      "entropy": 5.1990941047668455,
      "epoch": 4.930982198012018,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002839187081610645,
      "loss": 4.7594,
      "mean_token_accuracy": 0.23721840977668762,
      "num_tokens": 98007576.0,
      "step": 43905
    },
    {
      "entropy": 5.175595331192016,
      "epoch": 4.931543774919975,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002838785908540407,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24085288792848586,
      "num_tokens": 98017870.0,
      "step": 43910
    },
    {
      "entropy": 5.255601644515991,
      "epoch": 4.932105351827933,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028383847326431924,
      "loss": 4.8523,
      "mean_token_accuracy": 0.22913147509098053,
      "num_tokens": 98029076.0,
      "step": 43915
    },
    {
      "entropy": 5.272110223770142,
      "epoch": 4.93266692873589,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028379835539317727,
      "loss": 4.7197,
      "mean_token_accuracy": 0.24296425580978392,
      "num_tokens": 98042099.0,
      "step": 43920
    },
    {
      "entropy": 5.183430337905884,
      "epoch": 4.933228505643848,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002837582372418923,
      "loss": 4.7316,
      "mean_token_accuracy": 0.23968379944562912,
      "num_tokens": 98053335.0,
      "step": 43925
    },
    {
      "entropy": 5.246534776687622,
      "epoch": 4.9337900825518055,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002837181188117417,
      "loss": 4.8044,
      "mean_token_accuracy": 0.23265934437513353,
      "num_tokens": 98063723.0,
      "step": 43930
    },
    {
      "entropy": 5.200674152374267,
      "epoch": 4.934351659459763,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002836780001040028,
      "loss": 4.7353,
      "mean_token_accuracy": 0.23733187913894654,
      "num_tokens": 98074756.0,
      "step": 43935
    },
    {
      "entropy": 5.19329342842102,
      "epoch": 4.934913236367721,
      "grad_norm": 1.0625,
      "learning_rate": 0.000283637881119953,
      "loss": 4.742,
      "mean_token_accuracy": 0.24207340478897094,
      "num_tokens": 98086584.0,
      "step": 43940
    },
    {
      "entropy": 5.272948598861694,
      "epoch": 4.935474813275678,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028359776186086965,
      "loss": 4.9572,
      "mean_token_accuracy": 0.2298854559659958,
      "num_tokens": 98099996.0,
      "step": 43945
    },
    {
      "entropy": 5.215657711029053,
      "epoch": 4.936036390183635,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002835576423280303,
      "loss": 4.7298,
      "mean_token_accuracy": 0.23656198084354402,
      "num_tokens": 98111740.0,
      "step": 43950
    },
    {
      "entropy": 5.269463634490966,
      "epoch": 4.936597967091593,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002835175225227123,
      "loss": 4.7128,
      "mean_token_accuracy": 0.2393142431974411,
      "num_tokens": 98123946.0,
      "step": 43955
    },
    {
      "entropy": 5.234606218338013,
      "epoch": 4.937159543999551,
      "grad_norm": 1.0,
      "learning_rate": 0.0002834774024461931,
      "loss": 4.7514,
      "mean_token_accuracy": 0.23567394763231278,
      "num_tokens": 98134980.0,
      "step": 43960
    },
    {
      "entropy": 5.236455249786377,
      "epoch": 4.937721120907508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00028343728209975,
      "loss": 4.8151,
      "mean_token_accuracy": 0.23614969253540039,
      "num_tokens": 98145639.0,
      "step": 43965
    },
    {
      "entropy": 5.251062631607056,
      "epoch": 4.938282697815466,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028339716148466067,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2385004222393036,
      "num_tokens": 98156078.0,
      "step": 43970
    },
    {
      "entropy": 5.283962917327881,
      "epoch": 4.938844274723423,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00028335704060220246,
      "loss": 4.8976,
      "mean_token_accuracy": 0.23209537863731383,
      "num_tokens": 98168486.0,
      "step": 43975
    },
    {
      "entropy": 5.333389806747436,
      "epoch": 4.939405851631381,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028331691945365274,
      "loss": 4.8249,
      "mean_token_accuracy": 0.23366974741220475,
      "num_tokens": 98178196.0,
      "step": 43980
    },
    {
      "entropy": 5.248131656646729,
      "epoch": 4.9399674285393385,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002832767980402892,
      "loss": 4.7754,
      "mean_token_accuracy": 0.23406852632761002,
      "num_tokens": 98188912.0,
      "step": 43985
    },
    {
      "entropy": 5.2323251247406,
      "epoch": 4.940529005447296,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002832366763633892,
      "loss": 4.8821,
      "mean_token_accuracy": 0.22770891040563584,
      "num_tokens": 98200207.0,
      "step": 43990
    },
    {
      "entropy": 5.125894069671631,
      "epoch": 4.941090582355254,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028319655442423015,
      "loss": 4.6659,
      "mean_token_accuracy": 0.2446729302406311,
      "num_tokens": 98210846.0,
      "step": 43995
    },
    {
      "entropy": 5.323276996612549,
      "epoch": 4.941652159263211,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028315643222408965,
      "loss": 4.8801,
      "mean_token_accuracy": 0.22580666989088058,
      "num_tokens": 98222303.0,
      "step": 44000
    },
    {
      "entropy": 5.168296432495117,
      "epoch": 4.942213736171169,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00028311630976424526,
      "loss": 4.7176,
      "mean_token_accuracy": 0.24341639131307602,
      "num_tokens": 98233374.0,
      "step": 44005
    },
    {
      "entropy": 5.342137479782105,
      "epoch": 4.942775313079126,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002830761870459743,
      "loss": 4.916,
      "mean_token_accuracy": 0.23302103877067565,
      "num_tokens": 98244840.0,
      "step": 44010
    },
    {
      "entropy": 5.197004413604736,
      "epoch": 4.9433368899870835,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002830360640705545,
      "loss": 4.7324,
      "mean_token_accuracy": 0.23952704519033433,
      "num_tokens": 98254829.0,
      "step": 44015
    },
    {
      "entropy": 5.251994562149048,
      "epoch": 4.943898466895042,
      "grad_norm": 1.0625,
      "learning_rate": 0.00028299594083926324,
      "loss": 4.8175,
      "mean_token_accuracy": 0.23506294190883636,
      "num_tokens": 98266296.0,
      "step": 44020
    },
    {
      "entropy": 5.184359169006347,
      "epoch": 4.944460043802999,
      "grad_norm": 1.109375,
      "learning_rate": 0.00028295581735337817,
      "loss": 4.7611,
      "mean_token_accuracy": 0.24783453792333604,
      "num_tokens": 98277723.0,
      "step": 44025
    },
    {
      "entropy": 5.261891555786133,
      "epoch": 4.945021620710956,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028291569361417674,
      "loss": 4.7965,
      "mean_token_accuracy": 0.24749888181686402,
      "num_tokens": 98288799.0,
      "step": 44030
    },
    {
      "entropy": 5.259262943267823,
      "epoch": 4.945583197618914,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002828755696229366,
      "loss": 4.8215,
      "mean_token_accuracy": 0.23232014030218123,
      "num_tokens": 98299731.0,
      "step": 44035
    },
    {
      "entropy": 5.240669870376587,
      "epoch": 4.946144774526871,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002828354453809352,
      "loss": 4.6924,
      "mean_token_accuracy": 0.23643240928649903,
      "num_tokens": 98310607.0,
      "step": 44040
    },
    {
      "entropy": 5.201419925689697,
      "epoch": 4.946706351434829,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002827953208894503,
      "loss": 4.7243,
      "mean_token_accuracy": 0.23718061298131943,
      "num_tokens": 98320149.0,
      "step": 44045
    },
    {
      "entropy": 5.207759952545166,
      "epoch": 4.947267928342787,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002827551961497593,
      "loss": 4.7244,
      "mean_token_accuracy": 0.2361192911863327,
      "num_tokens": 98330816.0,
      "step": 44050
    },
    {
      "entropy": 5.137837934494018,
      "epoch": 4.947829505250744,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002827150711631398,
      "loss": 4.6345,
      "mean_token_accuracy": 0.2416153609752655,
      "num_tokens": 98342041.0,
      "step": 44055
    },
    {
      "entropy": 5.205510997772217,
      "epoch": 4.948391082158702,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00028267494593086945,
      "loss": 4.8573,
      "mean_token_accuracy": 0.23232554793357849,
      "num_tokens": 98353891.0,
      "step": 44060
    },
    {
      "entropy": 5.292726373672485,
      "epoch": 4.948952659066659,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028263482045422597,
      "loss": 4.8497,
      "mean_token_accuracy": 0.23188333064317704,
      "num_tokens": 98364931.0,
      "step": 44065
    },
    {
      "entropy": 5.3134759902954105,
      "epoch": 4.9495142359746165,
      "grad_norm": 0.9375,
      "learning_rate": 0.0002825946947344867,
      "loss": 4.7937,
      "mean_token_accuracy": 0.23692181557416916,
      "num_tokens": 98376787.0,
      "step": 44070
    },
    {
      "entropy": 5.263869142532348,
      "epoch": 4.950075812882575,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002825545687729295,
      "loss": 4.7796,
      "mean_token_accuracy": 0.2334692433476448,
      "num_tokens": 98387750.0,
      "step": 44075
    },
    {
      "entropy": 5.153075075149536,
      "epoch": 4.950637389790532,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028251444257083186,
      "loss": 4.734,
      "mean_token_accuracy": 0.24027980715036393,
      "num_tokens": 98399309.0,
      "step": 44080
    },
    {
      "entropy": 5.16041145324707,
      "epoch": 4.951198966698489,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00028247431612947146,
      "loss": 4.6817,
      "mean_token_accuracy": 0.23720501661300658,
      "num_tokens": 98410850.0,
      "step": 44085
    },
    {
      "entropy": 5.261443281173706,
      "epoch": 4.951760543606447,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002824341894501259,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23445961326360704,
      "num_tokens": 98422166.0,
      "step": 44090
    },
    {
      "entropy": 5.241806268692017,
      "epoch": 4.952322120514404,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028239406253407286,
      "loss": 4.8101,
      "mean_token_accuracy": 0.2341987058520317,
      "num_tokens": 98432885.0,
      "step": 44095
    },
    {
      "entropy": 5.281293964385986,
      "epoch": 4.952883697422362,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00028235393538259007,
      "loss": 4.7914,
      "mean_token_accuracy": 0.23445586115121841,
      "num_tokens": 98444513.0,
      "step": 44100
    },
    {
      "entropy": 5.265384769439697,
      "epoch": 4.95344527433032,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002823138079969551,
      "loss": 4.7909,
      "mean_token_accuracy": 0.23606136590242385,
      "num_tokens": 98456009.0,
      "step": 44105
    },
    {
      "entropy": 5.341421270370484,
      "epoch": 4.954006851238277,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028227368037844557,
      "loss": 4.8623,
      "mean_token_accuracy": 0.23545572608709336,
      "num_tokens": 98466282.0,
      "step": 44110
    },
    {
      "entropy": 5.27578821182251,
      "epoch": 4.954568428146235,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028223355252833933,
      "loss": 4.7879,
      "mean_token_accuracy": 0.2366115301847458,
      "num_tokens": 98478334.0,
      "step": 44115
    },
    {
      "entropy": 5.261091232299805,
      "epoch": 4.955130005054192,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028219342444791396,
      "loss": 4.7981,
      "mean_token_accuracy": 0.2371227577328682,
      "num_tokens": 98488446.0,
      "step": 44120
    },
    {
      "entropy": 5.175391483306885,
      "epoch": 4.955691581962149,
      "grad_norm": 0.96875,
      "learning_rate": 0.00028215329613844715,
      "loss": 4.6359,
      "mean_token_accuracy": 0.2458663374185562,
      "num_tokens": 98499698.0,
      "step": 44125
    },
    {
      "entropy": 5.247283887863159,
      "epoch": 4.9562531588701075,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028211316760121656,
      "loss": 4.8115,
      "mean_token_accuracy": 0.24179720133543015,
      "num_tokens": 98509980.0,
      "step": 44130
    },
    {
      "entropy": 5.332046270370483,
      "epoch": 4.956814735778065,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002820730388375001,
      "loss": 4.8442,
      "mean_token_accuracy": 0.23987049758434295,
      "num_tokens": 98521406.0,
      "step": 44135
    },
    {
      "entropy": 5.342965412139892,
      "epoch": 4.957376312686022,
      "grad_norm": 1.203125,
      "learning_rate": 0.0002820329098485752,
      "loss": 4.9334,
      "mean_token_accuracy": 0.2344888687133789,
      "num_tokens": 98532845.0,
      "step": 44140
    },
    {
      "entropy": 5.16468505859375,
      "epoch": 4.95793788959398,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028199278063571976,
      "loss": 4.6564,
      "mean_token_accuracy": 0.24483602344989777,
      "num_tokens": 98543639.0,
      "step": 44145
    },
    {
      "entropy": 5.229601621627808,
      "epoch": 4.958499466501937,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00028195265120021157,
      "loss": 4.7564,
      "mean_token_accuracy": 0.23717784583568574,
      "num_tokens": 98554604.0,
      "step": 44150
    },
    {
      "entropy": 5.269136381149292,
      "epoch": 4.9590610434098945,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00028191252154332816,
      "loss": 4.7941,
      "mean_token_accuracy": 0.23161568939685823,
      "num_tokens": 98566388.0,
      "step": 44155
    },
    {
      "entropy": 5.2833911895751955,
      "epoch": 4.959622620317853,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028187239166634737,
      "loss": 4.8599,
      "mean_token_accuracy": 0.2367495447397232,
      "num_tokens": 98577215.0,
      "step": 44160
    },
    {
      "entropy": 5.376021003723144,
      "epoch": 4.96018419722581,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028183226157054705,
      "loss": 4.9676,
      "mean_token_accuracy": 0.22531538903713227,
      "num_tokens": 98588640.0,
      "step": 44165
    },
    {
      "entropy": 5.269132375717163,
      "epoch": 4.960745774133768,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002817921312572048,
      "loss": 4.7398,
      "mean_token_accuracy": 0.24325349032878876,
      "num_tokens": 98599710.0,
      "step": 44170
    },
    {
      "entropy": 5.236153316497803,
      "epoch": 4.961307351041725,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002817520007275985,
      "loss": 4.8613,
      "mean_token_accuracy": 0.23346639573574066,
      "num_tokens": 98611510.0,
      "step": 44175
    },
    {
      "entropy": 5.201397180557251,
      "epoch": 4.961868927949682,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002817118699830058,
      "loss": 4.741,
      "mean_token_accuracy": 0.23767634630203247,
      "num_tokens": 98622982.0,
      "step": 44180
    },
    {
      "entropy": 5.337669754028321,
      "epoch": 4.9624305048576405,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002816717390247046,
      "loss": 4.9223,
      "mean_token_accuracy": 0.2355792388319969,
      "num_tokens": 98634216.0,
      "step": 44185
    },
    {
      "entropy": 5.32530632019043,
      "epoch": 4.962992081765598,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002816316078539727,
      "loss": 4.9073,
      "mean_token_accuracy": 0.2305464953184128,
      "num_tokens": 98646469.0,
      "step": 44190
    },
    {
      "entropy": 5.3372574806213375,
      "epoch": 4.963553658673556,
      "grad_norm": 0.8828125,
      "learning_rate": 0.00028159147647208784,
      "loss": 4.8856,
      "mean_token_accuracy": 0.22412094324827195,
      "num_tokens": 98659249.0,
      "step": 44195
    },
    {
      "entropy": 5.230209064483643,
      "epoch": 4.964115235581513,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002815513448803277,
      "loss": 4.7113,
      "mean_token_accuracy": 0.2424386978149414,
      "num_tokens": 98671322.0,
      "step": 44200
    },
    {
      "entropy": 5.268703365325928,
      "epoch": 4.96467681248947,
      "grad_norm": 0.94140625,
      "learning_rate": 0.0002815112130799703,
      "loss": 4.8042,
      "mean_token_accuracy": 0.23851657658815384,
      "num_tokens": 98684236.0,
      "step": 44205
    },
    {
      "entropy": 5.1876908302307125,
      "epoch": 4.965238389397428,
      "grad_norm": 1.09375,
      "learning_rate": 0.00028147108107229324,
      "loss": 4.6814,
      "mean_token_accuracy": 0.2350316807627678,
      "num_tokens": 98694581.0,
      "step": 44210
    },
    {
      "entropy": 5.275788021087647,
      "epoch": 4.965799966305386,
      "grad_norm": 0.984375,
      "learning_rate": 0.00028143094885857455,
      "loss": 4.8103,
      "mean_token_accuracy": 0.23428725600242614,
      "num_tokens": 98705002.0,
      "step": 44215
    },
    {
      "entropy": 5.236483001708985,
      "epoch": 4.966361543213343,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000281390816440092,
      "loss": 4.7713,
      "mean_token_accuracy": 0.23518777191638945,
      "num_tokens": 98715131.0,
      "step": 44220
    },
    {
      "entropy": 5.157729053497315,
      "epoch": 4.966923120121301,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002813506838181233,
      "loss": 4.6625,
      "mean_token_accuracy": 0.24218697398900985,
      "num_tokens": 98724992.0,
      "step": 44225
    },
    {
      "entropy": 5.196441698074341,
      "epoch": 4.967484697029258,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002813105509939464,
      "loss": 4.7479,
      "mean_token_accuracy": 0.2343054160475731,
      "num_tokens": 98736651.0,
      "step": 44230
    },
    {
      "entropy": 5.263960981369019,
      "epoch": 4.968046273937215,
      "grad_norm": 1.046875,
      "learning_rate": 0.000281270417968839,
      "loss": 4.7654,
      "mean_token_accuracy": 0.24268896281719207,
      "num_tokens": 98747365.0,
      "step": 44235
    },
    {
      "entropy": 5.270205640792847,
      "epoch": 4.968607850845173,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002812302847440791,
      "loss": 4.8766,
      "mean_token_accuracy": 0.22574579119682311,
      "num_tokens": 98759610.0,
      "step": 44240
    },
    {
      "entropy": 5.284456443786621,
      "epoch": 4.969169427753131,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028119015132094455,
      "loss": 4.7828,
      "mean_token_accuracy": 0.2444018006324768,
      "num_tokens": 98770166.0,
      "step": 44245
    },
    {
      "entropy": 5.269431829452515,
      "epoch": 4.969731004661089,
      "grad_norm": 0.921875,
      "learning_rate": 0.0002811500177007132,
      "loss": 4.8053,
      "mean_token_accuracy": 0.2383579820394516,
      "num_tokens": 98782416.0,
      "step": 44250
    },
    {
      "entropy": 5.276953268051147,
      "epoch": 4.970292581569046,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00028110988388466287,
      "loss": 4.8099,
      "mean_token_accuracy": 0.2324220523238182,
      "num_tokens": 98793341.0,
      "step": 44255
    },
    {
      "entropy": 5.159477567672729,
      "epoch": 4.970854158477003,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002810697498740715,
      "loss": 4.684,
      "mean_token_accuracy": 0.24132033288478852,
      "num_tokens": 98803642.0,
      "step": 44260
    },
    {
      "entropy": 5.203957223892212,
      "epoch": 4.971415735384961,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002810296156702169,
      "loss": 4.708,
      "mean_token_accuracy": 0.23866866528987885,
      "num_tokens": 98814650.0,
      "step": 44265
    },
    {
      "entropy": 5.2368443489074705,
      "epoch": 4.9719773122929185,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028098948127437695,
      "loss": 4.7478,
      "mean_token_accuracy": 0.23345359414815903,
      "num_tokens": 98825928.0,
      "step": 44270
    },
    {
      "entropy": 5.244530200958252,
      "epoch": 4.972538889200876,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002809493466878297,
      "loss": 4.734,
      "mean_token_accuracy": 0.2366577297449112,
      "num_tokens": 98837447.0,
      "step": 44275
    },
    {
      "entropy": 5.193045520782471,
      "epoch": 4.973100466108834,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00028090921191185295,
      "loss": 4.7432,
      "mean_token_accuracy": 0.2396464005112648,
      "num_tokens": 98847720.0,
      "step": 44280
    },
    {
      "entropy": 5.1384138584136965,
      "epoch": 4.973662043016791,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002808690769477245,
      "loss": 4.706,
      "mean_token_accuracy": 0.24821527004241944,
      "num_tokens": 98859315.0,
      "step": 44285
    },
    {
      "entropy": 5.174493217468262,
      "epoch": 4.974223619924748,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002808289417967225,
      "loss": 4.7994,
      "mean_token_accuracy": 0.23404530733823775,
      "num_tokens": 98869921.0,
      "step": 44290
    },
    {
      "entropy": 5.25439338684082,
      "epoch": 4.974785196832706,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00028078880646012467,
      "loss": 4.7483,
      "mean_token_accuracy": 0.24588350653648378,
      "num_tokens": 98881473.0,
      "step": 44295
    },
    {
      "entropy": 5.2276341915130615,
      "epoch": 4.975346773740664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00028074867093920897,
      "loss": 4.7157,
      "mean_token_accuracy": 0.23915425688028336,
      "num_tokens": 98892601.0,
      "step": 44300
    },
    {
      "entropy": 5.300288391113281,
      "epoch": 4.975908350648622,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00028070853523525337,
      "loss": 4.8334,
      "mean_token_accuracy": 0.23214435577392578,
      "num_tokens": 98903823.0,
      "step": 44305
    },
    {
      "entropy": 5.208245420455933,
      "epoch": 4.976469927556579,
      "grad_norm": 1.046875,
      "learning_rate": 0.00028066839934953573,
      "loss": 4.7914,
      "mean_token_accuracy": 0.23378988802433015,
      "num_tokens": 98914266.0,
      "step": 44310
    },
    {
      "entropy": 5.21807017326355,
      "epoch": 4.977031504464536,
      "grad_norm": 1.015625,
      "learning_rate": 0.00028062826328333415,
      "loss": 4.7264,
      "mean_token_accuracy": 0.2376851737499237,
      "num_tokens": 98924693.0,
      "step": 44315
    },
    {
      "entropy": 5.187321043014526,
      "epoch": 4.977593081372494,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002805881270379265,
      "loss": 4.7979,
      "mean_token_accuracy": 0.23496724665164948,
      "num_tokens": 98936081.0,
      "step": 44320
    },
    {
      "entropy": 5.234492921829224,
      "epoch": 4.9781546582804515,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002805479906145906,
      "loss": 4.7615,
      "mean_token_accuracy": 0.23513184785842894,
      "num_tokens": 98946495.0,
      "step": 44325
    },
    {
      "entropy": 5.226263570785522,
      "epoch": 4.978716235188409,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002805078540146046,
      "loss": 4.7297,
      "mean_token_accuracy": 0.23634301424026488,
      "num_tokens": 98956873.0,
      "step": 44330
    },
    {
      "entropy": 5.162502479553223,
      "epoch": 4.979277812096367,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002804677172392464,
      "loss": 4.7471,
      "mean_token_accuracy": 0.23905959725379944,
      "num_tokens": 98968349.0,
      "step": 44335
    },
    {
      "entropy": 5.169009780883789,
      "epoch": 4.979839389004324,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000280427580289794,
      "loss": 4.7365,
      "mean_token_accuracy": 0.24058184623718262,
      "num_tokens": 98978558.0,
      "step": 44340
    },
    {
      "entropy": 5.31737756729126,
      "epoch": 4.980400965912281,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028038744316752524,
      "loss": 4.8872,
      "mean_token_accuracy": 0.23625736236572265,
      "num_tokens": 98990435.0,
      "step": 44345
    },
    {
      "entropy": 5.302569198608398,
      "epoch": 4.980962542820239,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00028034730587371824,
      "loss": 4.7696,
      "mean_token_accuracy": 0.23250348567962648,
      "num_tokens": 99001516.0,
      "step": 44350
    },
    {
      "entropy": 5.269147205352783,
      "epoch": 4.9815241197281965,
      "grad_norm": 1.078125,
      "learning_rate": 0.00028030716840965096,
      "loss": 4.8332,
      "mean_token_accuracy": 0.23196515440940857,
      "num_tokens": 99012810.0,
      "step": 44355
    },
    {
      "entropy": 5.279753541946411,
      "epoch": 4.982085696636155,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00028026703077660143,
      "loss": 4.8763,
      "mean_token_accuracy": 0.2301155298948288,
      "num_tokens": 99023342.0,
      "step": 44360
    },
    {
      "entropy": 5.165128135681153,
      "epoch": 4.982647273544112,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002802268929758475,
      "loss": 4.7087,
      "mean_token_accuracy": 0.24001695066690446,
      "num_tokens": 99034047.0,
      "step": 44365
    },
    {
      "entropy": 5.282220697402954,
      "epoch": 4.983208850452069,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002801867550086674,
      "loss": 4.8299,
      "mean_token_accuracy": 0.2330919995903969,
      "num_tokens": 99045333.0,
      "step": 44370
    },
    {
      "entropy": 5.252616548538208,
      "epoch": 4.983770427360027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002801466168763389,
      "loss": 4.8355,
      "mean_token_accuracy": 0.23727480620145797,
      "num_tokens": 99057582.0,
      "step": 44375
    },
    {
      "entropy": 5.265471410751343,
      "epoch": 4.984332004267984,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002801064785801401,
      "loss": 4.7678,
      "mean_token_accuracy": 0.22981626838445662,
      "num_tokens": 99067940.0,
      "step": 44380
    },
    {
      "entropy": 5.289868259429932,
      "epoch": 4.9848935811759425,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00028006634012134915,
      "loss": 4.8177,
      "mean_token_accuracy": 0.23590100854635238,
      "num_tokens": 99079408.0,
      "step": 44385
    },
    {
      "entropy": 5.221021556854248,
      "epoch": 4.9854551580839,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00028002620150124386,
      "loss": 4.7047,
      "mean_token_accuracy": 0.2457431137561798,
      "num_tokens": 99090320.0,
      "step": 44390
    },
    {
      "entropy": 5.270478868484497,
      "epoch": 4.986016734991857,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00027998606272110237,
      "loss": 4.8192,
      "mean_token_accuracy": 0.22821950763463975,
      "num_tokens": 99101503.0,
      "step": 44395
    },
    {
      "entropy": 5.325045156478882,
      "epoch": 4.986578311899815,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002799459237822028,
      "loss": 4.8528,
      "mean_token_accuracy": 0.23291056901216506,
      "num_tokens": 99112768.0,
      "step": 44400
    },
    {
      "entropy": 5.229252576828003,
      "epoch": 4.987139888807772,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027990578468582304,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24214767813682556,
      "num_tokens": 99123473.0,
      "step": 44405
    },
    {
      "entropy": 5.270736169815064,
      "epoch": 4.9877014657157295,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002798656454332412,
      "loss": 4.8307,
      "mean_token_accuracy": 0.2380274072289467,
      "num_tokens": 99136005.0,
      "step": 44410
    },
    {
      "entropy": 5.320199012756348,
      "epoch": 4.988263042623688,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027982550602573524,
      "loss": 4.8424,
      "mean_token_accuracy": 0.24204597175121306,
      "num_tokens": 99147494.0,
      "step": 44415
    },
    {
      "entropy": 5.318610143661499,
      "epoch": 4.988824619531645,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002797853664645834,
      "loss": 4.8579,
      "mean_token_accuracy": 0.22937744110822678,
      "num_tokens": 99159255.0,
      "step": 44420
    },
    {
      "entropy": 5.254358863830566,
      "epoch": 4.989386196439602,
      "grad_norm": 1.2109375,
      "learning_rate": 0.0002797452267510637,
      "loss": 4.7587,
      "mean_token_accuracy": 0.23865382075309755,
      "num_tokens": 99169406.0,
      "step": 44425
    },
    {
      "entropy": 5.17222490310669,
      "epoch": 4.98994777334756,
      "grad_norm": 1.0078125,
      "learning_rate": 0.000279705086886454,
      "loss": 4.7561,
      "mean_token_accuracy": 0.2443811848759651,
      "num_tokens": 99180392.0,
      "step": 44430
    },
    {
      "entropy": 5.226733589172364,
      "epoch": 4.990509350255517,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002796649468720326,
      "loss": 4.6927,
      "mean_token_accuracy": 0.23443136513233184,
      "num_tokens": 99190387.0,
      "step": 44435
    },
    {
      "entropy": 5.252085208892822,
      "epoch": 4.9910709271634754,
      "grad_norm": 1.0,
      "learning_rate": 0.00027962480670907743,
      "loss": 4.774,
      "mean_token_accuracy": 0.23955830037593842,
      "num_tokens": 99202070.0,
      "step": 44440
    },
    {
      "entropy": 5.306234550476074,
      "epoch": 4.991632504071433,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00027958466639886663,
      "loss": 4.831,
      "mean_token_accuracy": 0.23104532212018966,
      "num_tokens": 99213323.0,
      "step": 44445
    },
    {
      "entropy": 5.241688871383667,
      "epoch": 4.99219408097939,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027954452594267825,
      "loss": 4.7703,
      "mean_token_accuracy": 0.23505284637212753,
      "num_tokens": 99224178.0,
      "step": 44450
    },
    {
      "entropy": 5.325490427017212,
      "epoch": 4.992755657887348,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00027950438534179046,
      "loss": 4.85,
      "mean_token_accuracy": 0.23338727205991744,
      "num_tokens": 99234538.0,
      "step": 44455
    },
    {
      "entropy": 5.28168625831604,
      "epoch": 4.993317234795305,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027946424459748126,
      "loss": 4.7819,
      "mean_token_accuracy": 0.23511512875556945,
      "num_tokens": 99245737.0,
      "step": 44460
    },
    {
      "entropy": 5.234690237045288,
      "epoch": 4.993878811703262,
      "grad_norm": 0.921875,
      "learning_rate": 0.00027942410371102877,
      "loss": 4.7803,
      "mean_token_accuracy": 0.23797144442796708,
      "num_tokens": 99257047.0,
      "step": 44465
    },
    {
      "entropy": 5.326316976547242,
      "epoch": 4.9944403886112205,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002793839626837111,
      "loss": 4.8849,
      "mean_token_accuracy": 0.2276895672082901,
      "num_tokens": 99267664.0,
      "step": 44470
    },
    {
      "entropy": 5.219992399215698,
      "epoch": 4.995001965519178,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027934382151680644,
      "loss": 4.8085,
      "mean_token_accuracy": 0.23537483364343642,
      "num_tokens": 99279270.0,
      "step": 44475
    },
    {
      "entropy": 5.227445793151856,
      "epoch": 4.995563542427135,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027930368021159273,
      "loss": 4.7258,
      "mean_token_accuracy": 0.235489684343338,
      "num_tokens": 99289724.0,
      "step": 44480
    },
    {
      "entropy": 5.320882749557495,
      "epoch": 4.996125119335093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002792635387693481,
      "loss": 4.8296,
      "mean_token_accuracy": 0.23984315544366835,
      "num_tokens": 99300326.0,
      "step": 44485
    },
    {
      "entropy": 5.2481273174285885,
      "epoch": 4.99668669624305,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00027922339719135093,
      "loss": 4.7282,
      "mean_token_accuracy": 0.23886145353317262,
      "num_tokens": 99312073.0,
      "step": 44490
    },
    {
      "entropy": 5.294343185424805,
      "epoch": 4.997248273151008,
      "grad_norm": 1.0,
      "learning_rate": 0.000279183255478879,
      "loss": 4.8885,
      "mean_token_accuracy": 0.23342855423688888,
      "num_tokens": 99323938.0,
      "step": 44495
    },
    {
      "entropy": 5.203265953063965,
      "epoch": 4.997809850058966,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027914311363321073,
      "loss": 4.7443,
      "mean_token_accuracy": 0.23976293802261353,
      "num_tokens": 99334773.0,
      "step": 44500
    },
    {
      "entropy": 5.215453863143921,
      "epoch": 4.998371426966923,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027910297165562406,
      "loss": 4.728,
      "mean_token_accuracy": 0.24846929609775542,
      "num_tokens": 99345688.0,
      "step": 44505
    },
    {
      "entropy": 5.262765026092529,
      "epoch": 4.998933003874881,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027906282954739714,
      "loss": 4.7766,
      "mean_token_accuracy": 0.23908553272485733,
      "num_tokens": 99356307.0,
      "step": 44510
    },
    {
      "entropy": 5.168824958801269,
      "epoch": 4.999494580782838,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027902268730980816,
      "loss": 4.6786,
      "mean_token_accuracy": 0.24356340020895004,
      "num_tokens": 99367348.0,
      "step": 44515
    },
    {
      "entropy": 5.267103672027588,
      "epoch": 5.0,
      "grad_norm": 6.46875,
      "learning_rate": 0.0002789825449441353,
      "loss": 4.7913,
      "mean_token_accuracy": 0.2410682870282067,
      "num_tokens": 99377525.0,
      "step": 44520
    },
    {
      "entropy": 5.250489187240601,
      "epoch": 5.000561576907957,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002789424024516566,
      "loss": 4.6289,
      "mean_token_accuracy": 0.2483955904841423,
      "num_tokens": 99388342.0,
      "step": 44525
    },
    {
      "entropy": 5.256641578674317,
      "epoch": 5.001123153815915,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027890225983365034,
      "loss": 4.7285,
      "mean_token_accuracy": 0.2440231993794441,
      "num_tokens": 99400387.0,
      "step": 44530
    },
    {
      "entropy": 5.087394142150879,
      "epoch": 5.0016847307238725,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002788621170913946,
      "loss": 4.6155,
      "mean_token_accuracy": 0.24268828630447387,
      "num_tokens": 99411232.0,
      "step": 44535
    },
    {
      "entropy": 5.1565392971038815,
      "epoch": 5.00224630763183,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027882197422616746,
      "loss": 4.689,
      "mean_token_accuracy": 0.24488697350025176,
      "num_tokens": 99422839.0,
      "step": 44540
    },
    {
      "entropy": 5.287832975387573,
      "epoch": 5.002807884539788,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002787818312392473,
      "loss": 4.6775,
      "mean_token_accuracy": 0.24406201094388963,
      "num_tokens": 99434835.0,
      "step": 44545
    },
    {
      "entropy": 5.223873472213745,
      "epoch": 5.003369461447745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027874168813191207,
      "loss": 4.6922,
      "mean_token_accuracy": 0.24038360714912416,
      "num_tokens": 99445705.0,
      "step": 44550
    },
    {
      "entropy": 5.167449855804444,
      "epoch": 5.003931038355703,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002787015449054401,
      "loss": 4.6982,
      "mean_token_accuracy": 0.24947221130132674,
      "num_tokens": 99457127.0,
      "step": 44555
    },
    {
      "entropy": 5.143905973434448,
      "epoch": 5.00449261526366,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002786614015611095,
      "loss": 4.631,
      "mean_token_accuracy": 0.24662221670150758,
      "num_tokens": 99468653.0,
      "step": 44560
    },
    {
      "entropy": 5.322755813598633,
      "epoch": 5.005054192171618,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027862125810019844,
      "loss": 4.8864,
      "mean_token_accuracy": 0.22701707929372789,
      "num_tokens": 99481238.0,
      "step": 44565
    },
    {
      "entropy": 5.228534364700318,
      "epoch": 5.005615769079576,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00027858111452398515,
      "loss": 4.6564,
      "mean_token_accuracy": 0.25132557153701784,
      "num_tokens": 99493025.0,
      "step": 44570
    },
    {
      "entropy": 5.195140933990478,
      "epoch": 5.006177345987533,
      "grad_norm": 0.96875,
      "learning_rate": 0.00027854097083374773,
      "loss": 4.7157,
      "mean_token_accuracy": 0.25063043683767317,
      "num_tokens": 99504675.0,
      "step": 44575
    },
    {
      "entropy": 5.259073114395141,
      "epoch": 5.00673892289549,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002785008270307645,
      "loss": 4.6833,
      "mean_token_accuracy": 0.24807005375623703,
      "num_tokens": 99516005.0,
      "step": 44580
    },
    {
      "entropy": 5.30901575088501,
      "epoch": 5.007300499803448,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002784606831163136,
      "loss": 4.8502,
      "mean_token_accuracy": 0.22814476042985915,
      "num_tokens": 99526917.0,
      "step": 44585
    },
    {
      "entropy": 5.177452039718628,
      "epoch": 5.0078620767114055,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027842053909167314,
      "loss": 4.6692,
      "mean_token_accuracy": 0.24167088270187378,
      "num_tokens": 99538799.0,
      "step": 44590
    },
    {
      "entropy": 5.214388036727906,
      "epoch": 5.008423653619364,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027838039495812133,
      "loss": 4.6862,
      "mean_token_accuracy": 0.24956719130277633,
      "num_tokens": 99550431.0,
      "step": 44595
    },
    {
      "entropy": 5.181206798553466,
      "epoch": 5.008985230527321,
      "grad_norm": 0.9375,
      "learning_rate": 0.00027834025071693655,
      "loss": 4.6408,
      "mean_token_accuracy": 0.24416609704494477,
      "num_tokens": 99561562.0,
      "step": 44600
    },
    {
      "entropy": 5.1754096984863285,
      "epoch": 5.009546807435278,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00027830010636939684,
      "loss": 4.6006,
      "mean_token_accuracy": 0.24765826016664505,
      "num_tokens": 99572101.0,
      "step": 44605
    },
    {
      "entropy": 5.2014000415802,
      "epoch": 5.010108384343236,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027825996191678043,
      "loss": 4.7051,
      "mean_token_accuracy": 0.23983686417341232,
      "num_tokens": 99583186.0,
      "step": 44610
    },
    {
      "entropy": 5.206631517410278,
      "epoch": 5.010669961251193,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027821981736036555,
      "loss": 4.6931,
      "mean_token_accuracy": 0.25272479057312014,
      "num_tokens": 99593716.0,
      "step": 44615
    },
    {
      "entropy": 5.281145095825195,
      "epoch": 5.011231538159151,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002781796727014305,
      "loss": 4.7362,
      "mean_token_accuracy": 0.23875643461942672,
      "num_tokens": 99605798.0,
      "step": 44620
    },
    {
      "entropy": 5.2192682266235355,
      "epoch": 5.011793115067109,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002781395279412535,
      "loss": 4.6638,
      "mean_token_accuracy": 0.2405787080526352,
      "num_tokens": 99617938.0,
      "step": 44625
    },
    {
      "entropy": 5.176753091812134,
      "epoch": 5.012354691975066,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027809938308111254,
      "loss": 4.6243,
      "mean_token_accuracy": 0.24469539523124695,
      "num_tokens": 99628761.0,
      "step": 44630
    },
    {
      "entropy": 5.193847894668579,
      "epoch": 5.012916268883023,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027805923812228615,
      "loss": 4.7296,
      "mean_token_accuracy": 0.23649005144834517,
      "num_tokens": 99640072.0,
      "step": 44635
    },
    {
      "entropy": 5.221174526214599,
      "epoch": 5.013477845790981,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027801909306605243,
      "loss": 4.7426,
      "mean_token_accuracy": 0.24104944616556168,
      "num_tokens": 99651280.0,
      "step": 44640
    },
    {
      "entropy": 5.227485847473145,
      "epoch": 5.014039422698938,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002779789479136896,
      "loss": 4.6993,
      "mean_token_accuracy": 0.2432950645685196,
      "num_tokens": 99661574.0,
      "step": 44645
    },
    {
      "entropy": 5.2262266159057615,
      "epoch": 5.0146009996068965,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002779388026664759,
      "loss": 4.6657,
      "mean_token_accuracy": 0.2464766725897789,
      "num_tokens": 99671092.0,
      "step": 44650
    },
    {
      "entropy": 5.349297332763672,
      "epoch": 5.015162576514854,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002778986573256896,
      "loss": 4.8186,
      "mean_token_accuracy": 0.2375694125890732,
      "num_tokens": 99682104.0,
      "step": 44655
    },
    {
      "entropy": 5.294382953643799,
      "epoch": 5.015724153422811,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002778585118926089,
      "loss": 4.7463,
      "mean_token_accuracy": 0.24176108986139297,
      "num_tokens": 99692152.0,
      "step": 44660
    },
    {
      "entropy": 5.199601411819458,
      "epoch": 5.016285730330769,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027781836636851207,
      "loss": 4.6526,
      "mean_token_accuracy": 0.24698698967695237,
      "num_tokens": 99703807.0,
      "step": 44665
    },
    {
      "entropy": 5.306618690490723,
      "epoch": 5.016847307238726,
      "grad_norm": 1.125,
      "learning_rate": 0.00027777822075467735,
      "loss": 4.7517,
      "mean_token_accuracy": 0.2452497139573097,
      "num_tokens": 99714039.0,
      "step": 44670
    },
    {
      "entropy": 5.18924994468689,
      "epoch": 5.0174088841466835,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000277738075052383,
      "loss": 4.6766,
      "mean_token_accuracy": 0.24286468476057052,
      "num_tokens": 99725215.0,
      "step": 44675
    },
    {
      "entropy": 5.225334119796753,
      "epoch": 5.017970461054642,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002776979292629074,
      "loss": 4.6988,
      "mean_token_accuracy": 0.24618188589811324,
      "num_tokens": 99736281.0,
      "step": 44680
    },
    {
      "entropy": 5.225396203994751,
      "epoch": 5.018532037962599,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002776577833875285,
      "loss": 4.707,
      "mean_token_accuracy": 0.24078089743852615,
      "num_tokens": 99747544.0,
      "step": 44685
    },
    {
      "entropy": 5.178102970123291,
      "epoch": 5.019093614870557,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027761763742752485,
      "loss": 4.6347,
      "mean_token_accuracy": 0.24268873482942582,
      "num_tokens": 99758806.0,
      "step": 44690
    },
    {
      "entropy": 5.234581136703492,
      "epoch": 5.019655191778514,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002775774913841746,
      "loss": 4.6469,
      "mean_token_accuracy": 0.24653190076351167,
      "num_tokens": 99769477.0,
      "step": 44695
    },
    {
      "entropy": 5.185678482055664,
      "epoch": 5.020216768686471,
      "grad_norm": 0.99609375,
      "learning_rate": 0.000277537345258756,
      "loss": 4.6676,
      "mean_token_accuracy": 0.2461527794599533,
      "num_tokens": 99780283.0,
      "step": 44700
    },
    {
      "entropy": 5.305940055847168,
      "epoch": 5.0207783455944295,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027749719905254737,
      "loss": 4.7366,
      "mean_token_accuracy": 0.24683361053466796,
      "num_tokens": 99790119.0,
      "step": 44705
    },
    {
      "entropy": 5.310364532470703,
      "epoch": 5.021339922502387,
      "grad_norm": 1.078125,
      "learning_rate": 0.000277457052766827,
      "loss": 4.7551,
      "mean_token_accuracy": 0.2366758778691292,
      "num_tokens": 99801619.0,
      "step": 44710
    },
    {
      "entropy": 5.198056173324585,
      "epoch": 5.021901499410344,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027741690640287306,
      "loss": 4.6477,
      "mean_token_accuracy": 0.25162058174610136,
      "num_tokens": 99812972.0,
      "step": 44715
    },
    {
      "entropy": 5.1622401714324955,
      "epoch": 5.022463076318302,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002773767599619639,
      "loss": 4.6302,
      "mean_token_accuracy": 0.24351560622453688,
      "num_tokens": 99823667.0,
      "step": 44720
    },
    {
      "entropy": 5.094820880889893,
      "epoch": 5.023024653226259,
      "grad_norm": 1.109375,
      "learning_rate": 0.00027733661344537776,
      "loss": 4.5612,
      "mean_token_accuracy": 0.25490122735500337,
      "num_tokens": 99834654.0,
      "step": 44725
    },
    {
      "entropy": 5.261213874816894,
      "epoch": 5.0235862301342165,
      "grad_norm": 1.0625,
      "learning_rate": 0.000277296466854393,
      "loss": 4.7716,
      "mean_token_accuracy": 0.24860702753067015,
      "num_tokens": 99846393.0,
      "step": 44730
    },
    {
      "entropy": 5.173838663101196,
      "epoch": 5.024147807042175,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002772563201902877,
      "loss": 4.6027,
      "mean_token_accuracy": 0.25806774497032164,
      "num_tokens": 99857206.0,
      "step": 44735
    },
    {
      "entropy": 5.3300597190856935,
      "epoch": 5.024709383950132,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027721617345434035,
      "loss": 4.8378,
      "mean_token_accuracy": 0.24470466375350952,
      "num_tokens": 99871497.0,
      "step": 44740
    },
    {
      "entropy": 5.183823966979981,
      "epoch": 5.02527096085809,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027717602664782925,
      "loss": 4.7591,
      "mean_token_accuracy": 0.233567912876606,
      "num_tokens": 99882337.0,
      "step": 44745
    },
    {
      "entropy": 5.186196374893188,
      "epoch": 5.025832537766047,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002771358797720326,
      "loss": 4.624,
      "mean_token_accuracy": 0.24762722700834275,
      "num_tokens": 99892863.0,
      "step": 44750
    },
    {
      "entropy": 5.198911428451538,
      "epoch": 5.026394114674004,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027709573282822864,
      "loss": 4.6104,
      "mean_token_accuracy": 0.25095388293266296,
      "num_tokens": 99903566.0,
      "step": 44755
    },
    {
      "entropy": 5.254693126678466,
      "epoch": 5.026955691581962,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002770555858176958,
      "loss": 4.6541,
      "mean_token_accuracy": 0.25035949796438217,
      "num_tokens": 99913466.0,
      "step": 44760
    },
    {
      "entropy": 5.217146682739258,
      "epoch": 5.02751726848992,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002770154387417123,
      "loss": 4.772,
      "mean_token_accuracy": 0.23848848342895507,
      "num_tokens": 99923579.0,
      "step": 44765
    },
    {
      "entropy": 5.237278842926026,
      "epoch": 5.028078845397877,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002769752916015565,
      "loss": 4.7394,
      "mean_token_accuracy": 0.23360720425844192,
      "num_tokens": 99935471.0,
      "step": 44770
    },
    {
      "entropy": 5.2650182247161865,
      "epoch": 5.028640422305835,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027693514439850646,
      "loss": 4.7682,
      "mean_token_accuracy": 0.24300145506858825,
      "num_tokens": 99947082.0,
      "step": 44775
    },
    {
      "entropy": 5.286144828796386,
      "epoch": 5.029201999213792,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027689499713384085,
      "loss": 4.7128,
      "mean_token_accuracy": 0.23782586455345153,
      "num_tokens": 99958186.0,
      "step": 44780
    },
    {
      "entropy": 5.275797033309937,
      "epoch": 5.02976357612175,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002768548498088378,
      "loss": 4.7602,
      "mean_token_accuracy": 0.2379780575633049,
      "num_tokens": 99970928.0,
      "step": 44785
    },
    {
      "entropy": 5.146509599685669,
      "epoch": 5.0303251530297075,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002768147024247756,
      "loss": 4.6639,
      "mean_token_accuracy": 0.24968788027763367,
      "num_tokens": 99982549.0,
      "step": 44790
    },
    {
      "entropy": 5.117396688461303,
      "epoch": 5.030886729937665,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002767745549829325,
      "loss": 4.5863,
      "mean_token_accuracy": 0.24404510110616684,
      "num_tokens": 99993835.0,
      "step": 44795
    },
    {
      "entropy": 5.273512601852417,
      "epoch": 5.031448306845623,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00027673440748458687,
      "loss": 4.7147,
      "mean_token_accuracy": 0.2430853232741356,
      "num_tokens": 100004336.0,
      "step": 44800
    },
    {
      "entropy": 5.201387071609497,
      "epoch": 5.03200988375358,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027669425993101716,
      "loss": 4.7191,
      "mean_token_accuracy": 0.24528108388185502,
      "num_tokens": 100015334.0,
      "step": 44805
    },
    {
      "entropy": 5.245095586776733,
      "epoch": 5.032571460661537,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002766541123235014,
      "loss": 4.7587,
      "mean_token_accuracy": 0.24121448546648025,
      "num_tokens": 100026665.0,
      "step": 44810
    },
    {
      "entropy": 5.188733386993408,
      "epoch": 5.033133037569495,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002766139646633182,
      "loss": 4.6575,
      "mean_token_accuracy": 0.24650531262159348,
      "num_tokens": 100037263.0,
      "step": 44815
    },
    {
      "entropy": 5.15772271156311,
      "epoch": 5.033694614477453,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027657381695174563,
      "loss": 4.6475,
      "mean_token_accuracy": 0.2461007684469223,
      "num_tokens": 100048482.0,
      "step": 44820
    },
    {
      "entropy": 5.180821275711059,
      "epoch": 5.03425619138541,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027653366919006217,
      "loss": 4.6912,
      "mean_token_accuracy": 0.24056586921215056,
      "num_tokens": 100059351.0,
      "step": 44825
    },
    {
      "entropy": 5.194101095199585,
      "epoch": 5.034817768293368,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002764935213795461,
      "loss": 4.5561,
      "mean_token_accuracy": 0.252719983458519,
      "num_tokens": 100069924.0,
      "step": 44830
    },
    {
      "entropy": 5.15826301574707,
      "epoch": 5.035379345201325,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002764533735214757,
      "loss": 4.6165,
      "mean_token_accuracy": 0.25024496763944626,
      "num_tokens": 100080733.0,
      "step": 44835
    },
    {
      "entropy": 5.167450952529907,
      "epoch": 5.035940922109283,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027641322561712935,
      "loss": 4.7244,
      "mean_token_accuracy": 0.23997031301259994,
      "num_tokens": 100091834.0,
      "step": 44840
    },
    {
      "entropy": 5.350835657119751,
      "epoch": 5.0365024990172405,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027637307766778524,
      "loss": 4.9014,
      "mean_token_accuracy": 0.23294377624988555,
      "num_tokens": 100103872.0,
      "step": 44845
    },
    {
      "entropy": 5.293009185791016,
      "epoch": 5.037064075925198,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000276332929674722,
      "loss": 4.7545,
      "mean_token_accuracy": 0.2425166830420494,
      "num_tokens": 100114391.0,
      "step": 44850
    },
    {
      "entropy": 5.263408422470093,
      "epoch": 5.037625652833156,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027629278163921763,
      "loss": 4.6685,
      "mean_token_accuracy": 0.24805041253566742,
      "num_tokens": 100125000.0,
      "step": 44855
    },
    {
      "entropy": 5.249198532104492,
      "epoch": 5.038187229741113,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027625263356255067,
      "loss": 4.6861,
      "mean_token_accuracy": 0.2505934491753578,
      "num_tokens": 100135394.0,
      "step": 44860
    },
    {
      "entropy": 5.127843904495239,
      "epoch": 5.03874880664907,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002762124854459992,
      "loss": 4.6135,
      "mean_token_accuracy": 0.2515612319111824,
      "num_tokens": 100146610.0,
      "step": 44865
    },
    {
      "entropy": 5.176906633377075,
      "epoch": 5.039310383557028,
      "grad_norm": 1.171875,
      "learning_rate": 0.00027617233729084194,
      "loss": 4.7278,
      "mean_token_accuracy": 0.24607477635145186,
      "num_tokens": 100156936.0,
      "step": 44870
    },
    {
      "entropy": 5.191626310348511,
      "epoch": 5.0398719604649855,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027613218909835693,
      "loss": 4.6736,
      "mean_token_accuracy": 0.24891327023506166,
      "num_tokens": 100168406.0,
      "step": 44875
    },
    {
      "entropy": 5.187040758132935,
      "epoch": 5.040433537372944,
      "grad_norm": 1.125,
      "learning_rate": 0.0002760920408698225,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2423777088522911,
      "num_tokens": 100179317.0,
      "step": 44880
    },
    {
      "entropy": 5.134545135498047,
      "epoch": 5.040995114280901,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027605189260651716,
      "loss": 4.5964,
      "mean_token_accuracy": 0.2520124718546867,
      "num_tokens": 100190916.0,
      "step": 44885
    },
    {
      "entropy": 5.129185771942138,
      "epoch": 5.041556691188858,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002760117443097191,
      "loss": 4.5807,
      "mean_token_accuracy": 0.25842768102884295,
      "num_tokens": 100201437.0,
      "step": 44890
    },
    {
      "entropy": 5.147569704055786,
      "epoch": 5.042118268096816,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027597159598070675,
      "loss": 4.6376,
      "mean_token_accuracy": 0.2461739882826805,
      "num_tokens": 100212405.0,
      "step": 44895
    },
    {
      "entropy": 5.137987041473389,
      "epoch": 5.042679845004773,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027593144762075837,
      "loss": 4.6086,
      "mean_token_accuracy": 0.25138685703277586,
      "num_tokens": 100223428.0,
      "step": 44900
    },
    {
      "entropy": 5.095826053619385,
      "epoch": 5.043241421912731,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027589129923115234,
      "loss": 4.5801,
      "mean_token_accuracy": 0.25170030891895295,
      "num_tokens": 100234128.0,
      "step": 44905
    },
    {
      "entropy": 5.210123682022095,
      "epoch": 5.043802998820689,
      "grad_norm": 1.078125,
      "learning_rate": 0.000275851150813167,
      "loss": 4.7064,
      "mean_token_accuracy": 0.24095221161842345,
      "num_tokens": 100244920.0,
      "step": 44910
    },
    {
      "entropy": 5.2046161651611325,
      "epoch": 5.044364575728646,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002758110023680807,
      "loss": 4.6708,
      "mean_token_accuracy": 0.24741435647010804,
      "num_tokens": 100256518.0,
      "step": 44915
    },
    {
      "entropy": 5.248053979873657,
      "epoch": 5.044926152636603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002757708538971717,
      "loss": 4.7412,
      "mean_token_accuracy": 0.2442145064473152,
      "num_tokens": 100268203.0,
      "step": 44920
    },
    {
      "entropy": 5.190173864364624,
      "epoch": 5.045487729544561,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002757307054017185,
      "loss": 4.6059,
      "mean_token_accuracy": 0.25077748149633405,
      "num_tokens": 100279440.0,
      "step": 44925
    },
    {
      "entropy": 5.208507776260376,
      "epoch": 5.0460493064525185,
      "grad_norm": 1.125,
      "learning_rate": 0.0002756905568829994,
      "loss": 4.7592,
      "mean_token_accuracy": 0.24439392536878585,
      "num_tokens": 100290616.0,
      "step": 44930
    },
    {
      "entropy": 5.212556171417236,
      "epoch": 5.046610883360477,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027565040834229266,
      "loss": 4.7476,
      "mean_token_accuracy": 0.24303986132144928,
      "num_tokens": 100301602.0,
      "step": 44935
    },
    {
      "entropy": 5.201145792007447,
      "epoch": 5.047172460268434,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00027561025978087663,
      "loss": 4.6613,
      "mean_token_accuracy": 0.2463190048933029,
      "num_tokens": 100313311.0,
      "step": 44940
    },
    {
      "entropy": 5.252402257919312,
      "epoch": 5.047734037176391,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027557011120002975,
      "loss": 4.7616,
      "mean_token_accuracy": 0.2397363692522049,
      "num_tokens": 100325394.0,
      "step": 44945
    },
    {
      "entropy": 5.2303708553314205,
      "epoch": 5.048295614084349,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0002755299626010302,
      "loss": 4.6915,
      "mean_token_accuracy": 0.24616056233644484,
      "num_tokens": 100336643.0,
      "step": 44950
    },
    {
      "entropy": 5.215870761871338,
      "epoch": 5.048857190992306,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002754898139851565,
      "loss": 4.7455,
      "mean_token_accuracy": 0.23877771198749542,
      "num_tokens": 100348409.0,
      "step": 44955
    },
    {
      "entropy": 5.239922761917114,
      "epoch": 5.049418767900264,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027544966535368695,
      "loss": 4.7572,
      "mean_token_accuracy": 0.24707447588443757,
      "num_tokens": 100360115.0,
      "step": 44960
    },
    {
      "entropy": 5.187343120574951,
      "epoch": 5.049980344808222,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002754095167078999,
      "loss": 4.6721,
      "mean_token_accuracy": 0.24528948813676835,
      "num_tokens": 100369618.0,
      "step": 44965
    },
    {
      "entropy": 5.174570989608765,
      "epoch": 5.050541921716179,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002753693680490736,
      "loss": 4.6845,
      "mean_token_accuracy": 0.24504151940345764,
      "num_tokens": 100379470.0,
      "step": 44970
    },
    {
      "entropy": 5.156136131286621,
      "epoch": 5.051103498624137,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002753292193784865,
      "loss": 4.6732,
      "mean_token_accuracy": 0.24129718840122222,
      "num_tokens": 100390490.0,
      "step": 44975
    },
    {
      "entropy": 5.206881952285767,
      "epoch": 5.051665075532094,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027528907069741696,
      "loss": 4.6629,
      "mean_token_accuracy": 0.2455584242939949,
      "num_tokens": 100402817.0,
      "step": 44980
    },
    {
      "entropy": 5.335949993133545,
      "epoch": 5.052226652440051,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002752489220071433,
      "loss": 4.8016,
      "mean_token_accuracy": 0.23998185843229294,
      "num_tokens": 100413411.0,
      "step": 44985
    },
    {
      "entropy": 5.1447960376739506,
      "epoch": 5.0527882293480095,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002752087733089438,
      "loss": 4.6722,
      "mean_token_accuracy": 0.2465070903301239,
      "num_tokens": 100425118.0,
      "step": 44990
    },
    {
      "entropy": 5.303683185577393,
      "epoch": 5.053349806255967,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002751686246040969,
      "loss": 4.7637,
      "mean_token_accuracy": 0.2371656209230423,
      "num_tokens": 100436544.0,
      "step": 44995
    },
    {
      "entropy": 5.249248456954956,
      "epoch": 5.053911383163924,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000275128475893881,
      "loss": 4.7013,
      "mean_token_accuracy": 0.24402351826429367,
      "num_tokens": 100448315.0,
      "step": 45000
    },
    {
      "epoch": 5.053911383163924,
      "eval_entropy": 4.999527418585761,
      "eval_loss": 4.959877014160156,
      "eval_mean_token_accuracy": 0.23503141583111412,
      "eval_num_tokens": 100448315.0,
      "eval_runtime": 24.2264,
      "eval_samples_per_second": 1280.131,
      "eval_steps_per_second": 160.032,
      "step": 45000
    },
    {
      "entropy": 5.126440906524659,
      "epoch": 5.054472960071882,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027508832717957433,
      "loss": 4.6789,
      "mean_token_accuracy": 0.24825461506843566,
      "num_tokens": 100459548.0,
      "step": 45005
    },
    {
      "entropy": 5.251913166046142,
      "epoch": 5.055034536979839,
      "grad_norm": 1.21875,
      "learning_rate": 0.0002750481784624553,
      "loss": 4.6966,
      "mean_token_accuracy": 0.23852693885564805,
      "num_tokens": 100468687.0,
      "step": 45010
    },
    {
      "entropy": 5.221861743927002,
      "epoch": 5.0555961138877965,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027500802974380224,
      "loss": 4.6669,
      "mean_token_accuracy": 0.24514099657535554,
      "num_tokens": 100479607.0,
      "step": 45015
    },
    {
      "entropy": 5.220742511749267,
      "epoch": 5.056157690795755,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002749678810248935,
      "loss": 4.7746,
      "mean_token_accuracy": 0.2427370950579643,
      "num_tokens": 100492245.0,
      "step": 45020
    },
    {
      "entropy": 5.1982269287109375,
      "epoch": 5.056719267703712,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002749277323070075,
      "loss": 4.6267,
      "mean_token_accuracy": 0.244378562271595,
      "num_tokens": 100503973.0,
      "step": 45025
    },
    {
      "entropy": 5.223833513259888,
      "epoch": 5.05728084461167,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027488758359142245,
      "loss": 4.6955,
      "mean_token_accuracy": 0.24776308834552765,
      "num_tokens": 100516003.0,
      "step": 45030
    },
    {
      "entropy": 5.227192115783692,
      "epoch": 5.057842421519627,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002748474348794167,
      "loss": 4.7706,
      "mean_token_accuracy": 0.24209462553262712,
      "num_tokens": 100528130.0,
      "step": 45035
    },
    {
      "entropy": 5.205281209945679,
      "epoch": 5.058403998427584,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027480728617226884,
      "loss": 4.695,
      "mean_token_accuracy": 0.2466667339205742,
      "num_tokens": 100539532.0,
      "step": 45040
    },
    {
      "entropy": 5.257230472564697,
      "epoch": 5.0589655753355425,
      "grad_norm": 1.0625,
      "learning_rate": 0.000274767137471257,
      "loss": 4.7857,
      "mean_token_accuracy": 0.2385192796587944,
      "num_tokens": 100550520.0,
      "step": 45045
    },
    {
      "entropy": 5.171261548995972,
      "epoch": 5.0595271522435,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027472698877765967,
      "loss": 4.5833,
      "mean_token_accuracy": 0.2531109720468521,
      "num_tokens": 100561596.0,
      "step": 45050
    },
    {
      "entropy": 5.213835191726685,
      "epoch": 5.060088729151457,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002746868400927551,
      "loss": 4.6692,
      "mean_token_accuracy": 0.24653631150722505,
      "num_tokens": 100571238.0,
      "step": 45055
    },
    {
      "entropy": 5.219841623306275,
      "epoch": 5.060650306059415,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002746466914178216,
      "loss": 4.7601,
      "mean_token_accuracy": 0.23924721628427506,
      "num_tokens": 100582687.0,
      "step": 45060
    },
    {
      "entropy": 5.202314186096191,
      "epoch": 5.061211882967372,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002746065427541376,
      "loss": 4.7205,
      "mean_token_accuracy": 0.24209873825311662,
      "num_tokens": 100594270.0,
      "step": 45065
    },
    {
      "entropy": 5.2051409721374515,
      "epoch": 5.06177345987533,
      "grad_norm": 0.984375,
      "learning_rate": 0.00027456639410298144,
      "loss": 4.7072,
      "mean_token_accuracy": 0.24255745559930803,
      "num_tokens": 100605449.0,
      "step": 45070
    },
    {
      "entropy": 5.1185715675354,
      "epoch": 5.062335036783288,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002745262454656314,
      "loss": 4.5938,
      "mean_token_accuracy": 0.2512149438261986,
      "num_tokens": 100616985.0,
      "step": 45075
    },
    {
      "entropy": 5.165256786346435,
      "epoch": 5.062896613691245,
      "grad_norm": 1.0859375,
      "learning_rate": 0.000274486096843366,
      "loss": 4.6768,
      "mean_token_accuracy": 0.24877087771892548,
      "num_tokens": 100628088.0,
      "step": 45080
    },
    {
      "entropy": 5.281077480316162,
      "epoch": 5.063458190599203,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027444594823746344,
      "loss": 4.8234,
      "mean_token_accuracy": 0.24527217596769332,
      "num_tokens": 100639805.0,
      "step": 45085
    },
    {
      "entropy": 5.369319581985474,
      "epoch": 5.06401976750716,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002744057996492021,
      "loss": 4.8855,
      "mean_token_accuracy": 0.23004891723394394,
      "num_tokens": 100650437.0,
      "step": 45090
    },
    {
      "entropy": 5.135490894317627,
      "epoch": 5.064581344415117,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002743656510798603,
      "loss": 4.6542,
      "mean_token_accuracy": 0.24632474035024643,
      "num_tokens": 100661466.0,
      "step": 45095
    },
    {
      "entropy": 5.24559907913208,
      "epoch": 5.065142921323075,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002743255025307165,
      "loss": 4.7235,
      "mean_token_accuracy": 0.24524563997983934,
      "num_tokens": 100671706.0,
      "step": 45100
    },
    {
      "entropy": 5.200650405883789,
      "epoch": 5.065704498231033,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00027428535400304886,
      "loss": 4.674,
      "mean_token_accuracy": 0.2429823398590088,
      "num_tokens": 100682592.0,
      "step": 45105
    },
    {
      "entropy": 5.250971555709839,
      "epoch": 5.06626607513899,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002742452054981358,
      "loss": 4.706,
      "mean_token_accuracy": 0.2442106246948242,
      "num_tokens": 100693893.0,
      "step": 45110
    },
    {
      "entropy": 5.1158998012542725,
      "epoch": 5.066827652046948,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002742050570172558,
      "loss": 4.6589,
      "mean_token_accuracy": 0.24817839115858079,
      "num_tokens": 100705455.0,
      "step": 45115
    },
    {
      "entropy": 5.178588628768921,
      "epoch": 5.067389228954905,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000274164908561687,
      "loss": 4.6224,
      "mean_token_accuracy": 0.2471787765622139,
      "num_tokens": 100714945.0,
      "step": 45120
    },
    {
      "entropy": 5.222664308547974,
      "epoch": 5.067950805862863,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027412476013270786,
      "loss": 4.7171,
      "mean_token_accuracy": 0.23711041659116744,
      "num_tokens": 100726299.0,
      "step": 45125
    },
    {
      "entropy": 5.274002742767334,
      "epoch": 5.0685123827708205,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027408461173159676,
      "loss": 4.822,
      "mean_token_accuracy": 0.23725317567586898,
      "num_tokens": 100739415.0,
      "step": 45130
    },
    {
      "entropy": 5.30993685722351,
      "epoch": 5.069073959678778,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00027404446335963195,
      "loss": 4.8195,
      "mean_token_accuracy": 0.23297815024852753,
      "num_tokens": 100751018.0,
      "step": 45135
    },
    {
      "entropy": 5.163495922088623,
      "epoch": 5.069635536586736,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002740043150180918,
      "loss": 4.6393,
      "mean_token_accuracy": 0.24554416984319688,
      "num_tokens": 100761425.0,
      "step": 45140
    },
    {
      "entropy": 5.257727718353271,
      "epoch": 5.070197113494693,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002739641667082546,
      "loss": 4.7333,
      "mean_token_accuracy": 0.23576885759830474,
      "num_tokens": 100772906.0,
      "step": 45145
    },
    {
      "entropy": 5.252061128616333,
      "epoch": 5.07075869040265,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00027392401843139876,
      "loss": 4.7566,
      "mean_token_accuracy": 0.2463841050863266,
      "num_tokens": 100784106.0,
      "step": 45150
    },
    {
      "entropy": 5.255598545074463,
      "epoch": 5.071320267310608,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002738838701888026,
      "loss": 4.8217,
      "mean_token_accuracy": 0.22958529442548753,
      "num_tokens": 100795066.0,
      "step": 45155
    },
    {
      "entropy": 5.185648822784424,
      "epoch": 5.071881844218566,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002738437219817444,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24098818749189377,
      "num_tokens": 100806936.0,
      "step": 45160
    },
    {
      "entropy": 5.0772336483001705,
      "epoch": 5.072443421126524,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027380357381150263,
      "loss": 4.5802,
      "mean_token_accuracy": 0.24759837985038757,
      "num_tokens": 100817483.0,
      "step": 45165
    },
    {
      "entropy": 5.321823024749756,
      "epoch": 5.073004998034481,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002737634256793555,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23667031526565552,
      "num_tokens": 100828482.0,
      "step": 45170
    },
    {
      "entropy": 5.207942247390747,
      "epoch": 5.073566574942438,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027372327758658135,
      "loss": 4.7655,
      "mean_token_accuracy": 0.24307496398687362,
      "num_tokens": 100839745.0,
      "step": 45175
    },
    {
      "entropy": 5.165433502197265,
      "epoch": 5.074128151850396,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027368312953445845,
      "loss": 4.6601,
      "mean_token_accuracy": 0.24432343542575835,
      "num_tokens": 100850441.0,
      "step": 45180
    },
    {
      "entropy": 5.109458541870117,
      "epoch": 5.0746897287583534,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002736429815242653,
      "loss": 4.6627,
      "mean_token_accuracy": 0.24075965732336044,
      "num_tokens": 100861428.0,
      "step": 45185
    },
    {
      "entropy": 5.149024820327758,
      "epoch": 5.075251305666311,
      "grad_norm": 1.125,
      "learning_rate": 0.0002736028335572802,
      "loss": 4.6245,
      "mean_token_accuracy": 0.24779110550880432,
      "num_tokens": 100872874.0,
      "step": 45190
    },
    {
      "entropy": 5.2584902286529545,
      "epoch": 5.075812882574269,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027356268563478135,
      "loss": 4.8053,
      "mean_token_accuracy": 0.22985466718673705,
      "num_tokens": 100884150.0,
      "step": 45195
    },
    {
      "entropy": 5.215853929519653,
      "epoch": 5.076374459482226,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002735225377580472,
      "loss": 4.6825,
      "mean_token_accuracy": 0.24616120755672455,
      "num_tokens": 100896440.0,
      "step": 45200
    },
    {
      "entropy": 5.196907711029053,
      "epoch": 5.076936036390183,
      "grad_norm": 1.078125,
      "learning_rate": 0.000273482389928356,
      "loss": 4.6872,
      "mean_token_accuracy": 0.2512981936335564,
      "num_tokens": 100906854.0,
      "step": 45205
    },
    {
      "entropy": 5.183503818511963,
      "epoch": 5.077497613298141,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002734422421469861,
      "loss": 4.6691,
      "mean_token_accuracy": 0.25517650246620177,
      "num_tokens": 100918644.0,
      "step": 45210
    },
    {
      "entropy": 5.229396915435791,
      "epoch": 5.0780591902060985,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002734020944152157,
      "loss": 4.6935,
      "mean_token_accuracy": 0.2395261585712433,
      "num_tokens": 100929313.0,
      "step": 45215
    },
    {
      "entropy": 5.258406019210815,
      "epoch": 5.078620767114057,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002733619467343234,
      "loss": 4.7061,
      "mean_token_accuracy": 0.24161602556705475,
      "num_tokens": 100940406.0,
      "step": 45220
    },
    {
      "entropy": 5.194463443756104,
      "epoch": 5.079182344022014,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002733217991055873,
      "loss": 4.6259,
      "mean_token_accuracy": 0.24537360072135925,
      "num_tokens": 100951086.0,
      "step": 45225
    },
    {
      "entropy": 5.217798662185669,
      "epoch": 5.079743920929971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002732816515302858,
      "loss": 4.7686,
      "mean_token_accuracy": 0.24190197885036469,
      "num_tokens": 100961417.0,
      "step": 45230
    },
    {
      "entropy": 5.209833097457886,
      "epoch": 5.080305497837929,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002732415040096971,
      "loss": 4.7497,
      "mean_token_accuracy": 0.2401091620326042,
      "num_tokens": 100973667.0,
      "step": 45235
    },
    {
      "entropy": 5.2440464973449705,
      "epoch": 5.080867074745886,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00027320135654509964,
      "loss": 4.7529,
      "mean_token_accuracy": 0.24299568831920623,
      "num_tokens": 100984891.0,
      "step": 45240
    },
    {
      "entropy": 5.209268379211426,
      "epoch": 5.081428651653844,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002731612091377717,
      "loss": 4.7445,
      "mean_token_accuracy": 0.2414698541164398,
      "num_tokens": 100997273.0,
      "step": 45245
    },
    {
      "entropy": 5.097209024429321,
      "epoch": 5.081990228561802,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027312106178899154,
      "loss": 4.5348,
      "mean_token_accuracy": 0.2624175578355789,
      "num_tokens": 101007341.0,
      "step": 45250
    },
    {
      "entropy": 5.252381753921509,
      "epoch": 5.082551805469759,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002730809145000374,
      "loss": 4.7524,
      "mean_token_accuracy": 0.2403867244720459,
      "num_tokens": 101017471.0,
      "step": 45255
    },
    {
      "entropy": 5.146218204498291,
      "epoch": 5.083113382377717,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002730407672721878,
      "loss": 4.6354,
      "mean_token_accuracy": 0.2507937729358673,
      "num_tokens": 101028726.0,
      "step": 45260
    },
    {
      "entropy": 5.409296178817749,
      "epoch": 5.083674959285674,
      "grad_norm": 1.0390625,
      "learning_rate": 0.000273000620106721,
      "loss": 4.8621,
      "mean_token_accuracy": 0.23138701021671296,
      "num_tokens": 101040824.0,
      "step": 45265
    },
    {
      "entropy": 5.285718202590942,
      "epoch": 5.0842365361936315,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002729604730049151,
      "loss": 4.7445,
      "mean_token_accuracy": 0.23959893882274627,
      "num_tokens": 101051849.0,
      "step": 45270
    },
    {
      "entropy": 5.173731136322021,
      "epoch": 5.08479811310159,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002729203259680486,
      "loss": 4.7282,
      "mean_token_accuracy": 0.24122956097126008,
      "num_tokens": 101062667.0,
      "step": 45275
    },
    {
      "entropy": 5.185438013076782,
      "epoch": 5.085359690009547,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002728801789973998,
      "loss": 4.6838,
      "mean_token_accuracy": 0.2441992074251175,
      "num_tokens": 101073679.0,
      "step": 45280
    },
    {
      "entropy": 5.299175882339478,
      "epoch": 5.085921266917504,
      "grad_norm": 1.0,
      "learning_rate": 0.0002728400320942468,
      "loss": 4.7865,
      "mean_token_accuracy": 0.2390756607055664,
      "num_tokens": 101085394.0,
      "step": 45285
    },
    {
      "entropy": 5.318982362747192,
      "epoch": 5.086482843825462,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000272799885259868,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23578206747770308,
      "num_tokens": 101096748.0,
      "step": 45290
    },
    {
      "entropy": 5.146937227249145,
      "epoch": 5.087044420733419,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002727597384955418,
      "loss": 4.5303,
      "mean_token_accuracy": 0.2573366343975067,
      "num_tokens": 101107704.0,
      "step": 45295
    },
    {
      "entropy": 5.19688401222229,
      "epoch": 5.0876059976413766,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027271959180254636,
      "loss": 4.6637,
      "mean_token_accuracy": 0.24668407142162324,
      "num_tokens": 101119890.0,
      "step": 45300
    },
    {
      "entropy": 5.185189628601075,
      "epoch": 5.088167574549335,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027267944518215997,
      "loss": 4.6851,
      "mean_token_accuracy": 0.2439446210861206,
      "num_tokens": 101130172.0,
      "step": 45305
    },
    {
      "entropy": 5.1532995223999025,
      "epoch": 5.088729151457292,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027263929863566104,
      "loss": 4.6826,
      "mean_token_accuracy": 0.25064164102077485,
      "num_tokens": 101141374.0,
      "step": 45310
    },
    {
      "entropy": 5.218146848678589,
      "epoch": 5.08929072836525,
      "grad_norm": 0.93359375,
      "learning_rate": 0.0002725991521643277,
      "loss": 4.7369,
      "mean_token_accuracy": 0.2377629354596138,
      "num_tokens": 101155192.0,
      "step": 45315
    },
    {
      "entropy": 5.255433845520019,
      "epoch": 5.089852305273207,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002725590057694383,
      "loss": 4.7314,
      "mean_token_accuracy": 0.2437425434589386,
      "num_tokens": 101168438.0,
      "step": 45320
    },
    {
      "entropy": 5.270839214324951,
      "epoch": 5.090413882181164,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027251885945227093,
      "loss": 4.7175,
      "mean_token_accuracy": 0.24269421994686127,
      "num_tokens": 101180609.0,
      "step": 45325
    },
    {
      "entropy": 5.233816337585449,
      "epoch": 5.0909754590891225,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027247871321410425,
      "loss": 4.6731,
      "mean_token_accuracy": 0.24594552218914031,
      "num_tokens": 101191211.0,
      "step": 45330
    },
    {
      "entropy": 5.161465263366699,
      "epoch": 5.09153703599708,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00027243856705621623,
      "loss": 4.7316,
      "mean_token_accuracy": 0.240126571059227,
      "num_tokens": 101202706.0,
      "step": 45335
    },
    {
      "entropy": 5.192577695846557,
      "epoch": 5.092098612905037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027239842097988526,
      "loss": 4.727,
      "mean_token_accuracy": 0.24180304110050202,
      "num_tokens": 101213890.0,
      "step": 45340
    },
    {
      "entropy": 5.335529327392578,
      "epoch": 5.092660189812995,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002723582749863896,
      "loss": 4.8195,
      "mean_token_accuracy": 0.2331428661942482,
      "num_tokens": 101225946.0,
      "step": 45345
    },
    {
      "entropy": 5.284725666046143,
      "epoch": 5.093221766720952,
      "grad_norm": 1.125,
      "learning_rate": 0.00027231812907700746,
      "loss": 4.745,
      "mean_token_accuracy": 0.23782570362091066,
      "num_tokens": 101236269.0,
      "step": 45350
    },
    {
      "entropy": 5.2215556621551515,
      "epoch": 5.09378334362891,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002722779832530171,
      "loss": 4.7247,
      "mean_token_accuracy": 0.24864230751991273,
      "num_tokens": 101247641.0,
      "step": 45355
    },
    {
      "entropy": 5.176593446731568,
      "epoch": 5.094344920536868,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00027223783751569687,
      "loss": 4.6368,
      "mean_token_accuracy": 0.23999546319246293,
      "num_tokens": 101259271.0,
      "step": 45360
    },
    {
      "entropy": 5.11399097442627,
      "epoch": 5.094906497444825,
      "grad_norm": 1.15625,
      "learning_rate": 0.00027219769186632493,
      "loss": 4.5892,
      "mean_token_accuracy": 0.2509427011013031,
      "num_tokens": 101269709.0,
      "step": 45365
    },
    {
      "entropy": 5.199513483047485,
      "epoch": 5.095468074352783,
      "grad_norm": 1.125,
      "learning_rate": 0.0002721575463061796,
      "loss": 4.7542,
      "mean_token_accuracy": 0.23011732697486878,
      "num_tokens": 101280559.0,
      "step": 45370
    },
    {
      "entropy": 5.267155694961548,
      "epoch": 5.09602965126074,
      "grad_norm": 1.046875,
      "learning_rate": 0.00027211740083653903,
      "loss": 4.717,
      "mean_token_accuracy": 0.2448016658425331,
      "num_tokens": 101292356.0,
      "step": 45375
    },
    {
      "entropy": 5.238164234161377,
      "epoch": 5.096591228168697,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002720772554586816,
      "loss": 4.6903,
      "mean_token_accuracy": 0.24208771139383317,
      "num_tokens": 101303881.0,
      "step": 45380
    },
    {
      "entropy": 5.173835182189942,
      "epoch": 5.0971528050766555,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002720371101738855,
      "loss": 4.6351,
      "mean_token_accuracy": 0.25052636712789533,
      "num_tokens": 101315310.0,
      "step": 45385
    },
    {
      "entropy": 5.13978214263916,
      "epoch": 5.097714381984613,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027199696498342893,
      "loss": 4.6683,
      "mean_token_accuracy": 0.24831608980894088,
      "num_tokens": 101327085.0,
      "step": 45390
    },
    {
      "entropy": 5.161143493652344,
      "epoch": 5.09827595889257,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002719568198885901,
      "loss": 4.6775,
      "mean_token_accuracy": 0.24514085948467254,
      "num_tokens": 101340473.0,
      "step": 45395
    },
    {
      "entropy": 5.210851430892944,
      "epoch": 5.098837535800528,
      "grad_norm": 0.96875,
      "learning_rate": 0.0002719166748906473,
      "loss": 4.6782,
      "mean_token_accuracy": 0.2536092773079872,
      "num_tokens": 101353434.0,
      "step": 45400
    },
    {
      "entropy": 5.144558382034302,
      "epoch": 5.099399112708485,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027187652999087887,
      "loss": 4.6501,
      "mean_token_accuracy": 0.2420949384570122,
      "num_tokens": 101363782.0,
      "step": 45405
    },
    {
      "entropy": 5.229752159118652,
      "epoch": 5.099960689616443,
      "grad_norm": 1.0,
      "learning_rate": 0.00027183638519056284,
      "loss": 4.748,
      "mean_token_accuracy": 0.23967815935611725,
      "num_tokens": 101374985.0,
      "step": 45410
    },
    {
      "entropy": 5.182032251358033,
      "epoch": 5.1005222665244005,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002717962404909776,
      "loss": 4.6311,
      "mean_token_accuracy": 0.2508788242936134,
      "num_tokens": 101386916.0,
      "step": 45415
    },
    {
      "entropy": 5.14307541847229,
      "epoch": 5.101083843432358,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027175609589340135,
      "loss": 4.6179,
      "mean_token_accuracy": 0.24302704632282257,
      "num_tokens": 101397656.0,
      "step": 45420
    },
    {
      "entropy": 5.116913318634033,
      "epoch": 5.101645420340316,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002717159513991122,
      "loss": 4.6327,
      "mean_token_accuracy": 0.25080459117889403,
      "num_tokens": 101408725.0,
      "step": 45425
    },
    {
      "entropy": 5.101910352706909,
      "epoch": 5.102206997248273,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00027167580700938844,
      "loss": 4.5904,
      "mean_token_accuracy": 0.2555189073085785,
      "num_tokens": 101419580.0,
      "step": 45430
    },
    {
      "entropy": 5.213802146911621,
      "epoch": 5.10276857415623,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027163566272550826,
      "loss": 4.688,
      "mean_token_accuracy": 0.2409444957971573,
      "num_tokens": 101430735.0,
      "step": 45435
    },
    {
      "entropy": 5.210822629928589,
      "epoch": 5.103330151064188,
      "grad_norm": 1.140625,
      "learning_rate": 0.00027159551854874986,
      "loss": 4.6656,
      "mean_token_accuracy": 0.24101384431123735,
      "num_tokens": 101440544.0,
      "step": 45440
    },
    {
      "entropy": 5.223048257827759,
      "epoch": 5.103891727972146,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002715553744803915,
      "loss": 4.7495,
      "mean_token_accuracy": 0.23787965178489684,
      "num_tokens": 101450793.0,
      "step": 45445
    },
    {
      "entropy": 5.2094423294067385,
      "epoch": 5.104453304880104,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002715152305217114,
      "loss": 4.8346,
      "mean_token_accuracy": 0.2357469841837883,
      "num_tokens": 101463593.0,
      "step": 45450
    },
    {
      "entropy": 5.176147413253784,
      "epoch": 5.105014881788061,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002714750866739877,
      "loss": 4.6548,
      "mean_token_accuracy": 0.24629149734973907,
      "num_tokens": 101474334.0,
      "step": 45455
    },
    {
      "entropy": 5.23781452178955,
      "epoch": 5.105576458696018,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002714349429384986,
      "loss": 4.7403,
      "mean_token_accuracy": 0.24439975172281264,
      "num_tokens": 101485356.0,
      "step": 45460
    },
    {
      "entropy": 5.1658305644989015,
      "epoch": 5.106138035603976,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00027139479931652226,
      "loss": 4.6309,
      "mean_token_accuracy": 0.2482680857181549,
      "num_tokens": 101495939.0,
      "step": 45465
    },
    {
      "entropy": 5.199187421798706,
      "epoch": 5.1066996125119335,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002713546558093368,
      "loss": 4.665,
      "mean_token_accuracy": 0.23852687776088716,
      "num_tokens": 101505563.0,
      "step": 45470
    },
    {
      "entropy": 5.087663221359253,
      "epoch": 5.107261189419891,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002713145124182207,
      "loss": 4.5775,
      "mean_token_accuracy": 0.25636262744665145,
      "num_tokens": 101516829.0,
      "step": 45475
    },
    {
      "entropy": 5.242313432693481,
      "epoch": 5.107822766327849,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027127436914445194,
      "loss": 4.7439,
      "mean_token_accuracy": 0.241522915661335,
      "num_tokens": 101527427.0,
      "step": 45480
    },
    {
      "entropy": 5.197877407073975,
      "epoch": 5.108384343235806,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002712342259893086,
      "loss": 4.6478,
      "mean_token_accuracy": 0.2506156384944916,
      "num_tokens": 101537991.0,
      "step": 45485
    },
    {
      "entropy": 5.168023252487183,
      "epoch": 5.108945920143763,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00027119408295406906,
      "loss": 4.701,
      "mean_token_accuracy": 0.2373270332813263,
      "num_tokens": 101549485.0,
      "step": 45490
    },
    {
      "entropy": 5.2421074390411375,
      "epoch": 5.109507497051721,
      "grad_norm": 1.0,
      "learning_rate": 0.0002711539400400114,
      "loss": 4.7743,
      "mean_token_accuracy": 0.23539668321609497,
      "num_tokens": 101560719.0,
      "step": 45495
    },
    {
      "entropy": 5.283961582183838,
      "epoch": 5.110069073959679,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027111379724841373,
      "loss": 4.8128,
      "mean_token_accuracy": 0.23464485108852387,
      "num_tokens": 101572118.0,
      "step": 45500
    },
    {
      "entropy": 5.2985515117645265,
      "epoch": 5.110630650867637,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027107365458055424,
      "loss": 4.744,
      "mean_token_accuracy": 0.24464682787656783,
      "num_tokens": 101582050.0,
      "step": 45505
    },
    {
      "entropy": 5.13871169090271,
      "epoch": 5.111192227775594,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002710335120377112,
      "loss": 4.5517,
      "mean_token_accuracy": 0.2509902730584145,
      "num_tokens": 101592529.0,
      "step": 45510
    },
    {
      "entropy": 5.165092039108276,
      "epoch": 5.111753804683551,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027099336962116253,
      "loss": 4.7116,
      "mean_token_accuracy": 0.24359225779771804,
      "num_tokens": 101603106.0,
      "step": 45515
    },
    {
      "entropy": 5.263983726501465,
      "epoch": 5.112315381591509,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027095322733218665,
      "loss": 4.7321,
      "mean_token_accuracy": 0.23948917984962464,
      "num_tokens": 101614171.0,
      "step": 45520
    },
    {
      "entropy": 5.297670984268189,
      "epoch": 5.112876958499466,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002709130851720616,
      "loss": 4.7793,
      "mean_token_accuracy": 0.23598139137029647,
      "num_tokens": 101626311.0,
      "step": 45525
    },
    {
      "entropy": 5.122270488739014,
      "epoch": 5.113438535407424,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00027087294314206544,
      "loss": 4.6388,
      "mean_token_accuracy": 0.248039548099041,
      "num_tokens": 101637923.0,
      "step": 45530
    },
    {
      "entropy": 5.296495532989502,
      "epoch": 5.114000112315382,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002708328012434764,
      "loss": 4.8107,
      "mean_token_accuracy": 0.23338906317949296,
      "num_tokens": 101647943.0,
      "step": 45535
    },
    {
      "entropy": 5.232623815536499,
      "epoch": 5.114561689223339,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002707926594775725,
      "loss": 4.7227,
      "mean_token_accuracy": 0.24089373648166656,
      "num_tokens": 101660095.0,
      "step": 45540
    },
    {
      "entropy": 5.191632699966431,
      "epoch": 5.115123266131297,
      "grad_norm": 1.09375,
      "learning_rate": 0.00027075251784563196,
      "loss": 4.6504,
      "mean_token_accuracy": 0.2492341861128807,
      "num_tokens": 101670092.0,
      "step": 45545
    },
    {
      "entropy": 5.188868808746338,
      "epoch": 5.115684843039254,
      "grad_norm": 1.078125,
      "learning_rate": 0.000270712376348933,
      "loss": 4.7251,
      "mean_token_accuracy": 0.2376847341656685,
      "num_tokens": 101681692.0,
      "step": 45550
    },
    {
      "entropy": 5.206754493713379,
      "epoch": 5.1162464199472115,
      "grad_norm": 0.9140625,
      "learning_rate": 0.0002706722349887536,
      "loss": 4.6934,
      "mean_token_accuracy": 0.2409126952290535,
      "num_tokens": 101692569.0,
      "step": 45555
    },
    {
      "entropy": 5.269468927383423,
      "epoch": 5.11680799685517,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002706320937663719,
      "loss": 4.7577,
      "mean_token_accuracy": 0.24354901611804963,
      "num_tokens": 101704185.0,
      "step": 45560
    },
    {
      "entropy": 5.236518478393554,
      "epoch": 5.117369573763127,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002705919526830661,
      "loss": 4.7705,
      "mean_token_accuracy": 0.23717406541109085,
      "num_tokens": 101715500.0,
      "step": 45565
    },
    {
      "entropy": 5.2853460788726805,
      "epoch": 5.117931150671084,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027055181174011415,
      "loss": 4.7701,
      "mean_token_accuracy": 0.2473831668496132,
      "num_tokens": 101727562.0,
      "step": 45570
    },
    {
      "entropy": 5.2080193042755125,
      "epoch": 5.118492727579042,
      "grad_norm": 1.3515625,
      "learning_rate": 0.00027051167093879424,
      "loss": 4.6175,
      "mean_token_accuracy": 0.25069788098335266,
      "num_tokens": 101739269.0,
      "step": 45575
    },
    {
      "entropy": 5.231550073623657,
      "epoch": 5.119054304486999,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002704715302803846,
      "loss": 4.7624,
      "mean_token_accuracy": 0.24106201231479646,
      "num_tokens": 101750189.0,
      "step": 45580
    },
    {
      "entropy": 5.237618064880371,
      "epoch": 5.119615881394957,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00027043138976616304,
      "loss": 4.7448,
      "mean_token_accuracy": 0.23959157466888428,
      "num_tokens": 101760824.0,
      "step": 45585
    },
    {
      "entropy": 5.138515138626099,
      "epoch": 5.120177458302915,
      "grad_norm": 1.0625,
      "learning_rate": 0.00027039124939740787,
      "loss": 4.5653,
      "mean_token_accuracy": 0.2571459412574768,
      "num_tokens": 101772647.0,
      "step": 45590
    },
    {
      "entropy": 5.1950867652893065,
      "epoch": 5.120739035210872,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00027035110917539713,
      "loss": 4.7779,
      "mean_token_accuracy": 0.24067502617835998,
      "num_tokens": 101783020.0,
      "step": 45595
    },
    {
      "entropy": 5.23393406867981,
      "epoch": 5.12130061211883,
      "grad_norm": 1.078125,
      "learning_rate": 0.00027031096910140877,
      "loss": 4.6769,
      "mean_token_accuracy": 0.25143408626317976,
      "num_tokens": 101793372.0,
      "step": 45600
    },
    {
      "entropy": 5.200194835662842,
      "epoch": 5.121862189026787,
      "grad_norm": 1.03125,
      "learning_rate": 0.00027027082917672107,
      "loss": 4.7052,
      "mean_token_accuracy": 0.25018942505121233,
      "num_tokens": 101805077.0,
      "step": 45605
    },
    {
      "entropy": 5.174416732788086,
      "epoch": 5.1224237659347445,
      "grad_norm": 1.015625,
      "learning_rate": 0.00027023068940261203,
      "loss": 4.7261,
      "mean_token_accuracy": 0.24519172459840774,
      "num_tokens": 101816334.0,
      "step": 45610
    },
    {
      "entropy": 5.270313405990601,
      "epoch": 5.122985342842703,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002701905497803595,
      "loss": 4.8447,
      "mean_token_accuracy": 0.23802204579114913,
      "num_tokens": 101827812.0,
      "step": 45615
    },
    {
      "entropy": 5.193028402328491,
      "epoch": 5.12354691975066,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002701504103112419,
      "loss": 4.5892,
      "mean_token_accuracy": 0.25053138881921766,
      "num_tokens": 101839400.0,
      "step": 45620
    },
    {
      "entropy": 5.259127521514893,
      "epoch": 5.124108496658617,
      "grad_norm": 1.0,
      "learning_rate": 0.00027011027099653706,
      "loss": 4.7876,
      "mean_token_accuracy": 0.24164507538080215,
      "num_tokens": 101851948.0,
      "step": 45625
    },
    {
      "entropy": 5.207732391357422,
      "epoch": 5.124670073566575,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002700701318375232,
      "loss": 4.7707,
      "mean_token_accuracy": 0.23660111874341966,
      "num_tokens": 101862985.0,
      "step": 45630
    },
    {
      "entropy": 5.2119697570800785,
      "epoch": 5.125231650474532,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002700299928354781,
      "loss": 4.6608,
      "mean_token_accuracy": 0.24236390888690948,
      "num_tokens": 101873247.0,
      "step": 45635
    },
    {
      "entropy": 5.273244428634643,
      "epoch": 5.12579322738249,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026998985399168005,
      "loss": 4.7797,
      "mean_token_accuracy": 0.24448072463274,
      "num_tokens": 101883686.0,
      "step": 45640
    },
    {
      "entropy": 5.288251638412476,
      "epoch": 5.126354804290448,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026994971530740696,
      "loss": 4.7285,
      "mean_token_accuracy": 0.238612100481987,
      "num_tokens": 101894905.0,
      "step": 45645
    },
    {
      "entropy": 5.170753860473633,
      "epoch": 5.126916381198405,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026990957678393684,
      "loss": 4.7342,
      "mean_token_accuracy": 0.2401392638683319,
      "num_tokens": 101906671.0,
      "step": 45650
    },
    {
      "entropy": 5.1206776142120365,
      "epoch": 5.127477958106363,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026986943842254783,
      "loss": 4.6349,
      "mean_token_accuracy": 0.24596176147460938,
      "num_tokens": 101917440.0,
      "step": 45655
    },
    {
      "entropy": 5.21184344291687,
      "epoch": 5.12803953501432,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002698293002245179,
      "loss": 4.7262,
      "mean_token_accuracy": 0.24133458882570266,
      "num_tokens": 101929177.0,
      "step": 45660
    },
    {
      "entropy": 5.266460037231445,
      "epoch": 5.128601111922277,
      "grad_norm": 1.125,
      "learning_rate": 0.00026978916219112505,
      "loss": 4.7438,
      "mean_token_accuracy": 0.2428775057196617,
      "num_tokens": 101940034.0,
      "step": 45665
    },
    {
      "entropy": 5.3378924369812015,
      "epoch": 5.1291626888302355,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026974902432364727,
      "loss": 4.8115,
      "mean_token_accuracy": 0.23502749502658843,
      "num_tokens": 101952108.0,
      "step": 45670
    },
    {
      "entropy": 5.299449110031128,
      "epoch": 5.129724265738193,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026970888662336263,
      "loss": 4.7991,
      "mean_token_accuracy": 0.23536616712808608,
      "num_tokens": 101963617.0,
      "step": 45675
    },
    {
      "entropy": 5.243570280075073,
      "epoch": 5.130285842646151,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002696687490915491,
      "loss": 4.6994,
      "mean_token_accuracy": 0.25103124529123305,
      "num_tokens": 101975565.0,
      "step": 45680
    },
    {
      "entropy": 5.148386096954345,
      "epoch": 5.130847419554108,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026962861172948456,
      "loss": 4.7284,
      "mean_token_accuracy": 0.25118421465158464,
      "num_tokens": 101988048.0,
      "step": 45685
    },
    {
      "entropy": 5.133386898040771,
      "epoch": 5.131408996462065,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026958847453844713,
      "loss": 4.6524,
      "mean_token_accuracy": 0.24645467698574067,
      "num_tokens": 101999045.0,
      "step": 45690
    },
    {
      "entropy": 5.219966220855713,
      "epoch": 5.131970573370023,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026954833751971486,
      "loss": 4.7156,
      "mean_token_accuracy": 0.23830828070640564,
      "num_tokens": 102009679.0,
      "step": 45695
    },
    {
      "entropy": 5.2043304443359375,
      "epoch": 5.132532150277981,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002695082006745656,
      "loss": 4.6833,
      "mean_token_accuracy": 0.24028846472501755,
      "num_tokens": 102021378.0,
      "step": 45700
    },
    {
      "entropy": 5.283108234405518,
      "epoch": 5.133093727185938,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002694680640042774,
      "loss": 4.7246,
      "mean_token_accuracy": 0.2481010675430298,
      "num_tokens": 102031037.0,
      "step": 45705
    },
    {
      "entropy": 5.2323527336120605,
      "epoch": 5.133655304093896,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026942792751012814,
      "loss": 4.6687,
      "mean_token_accuracy": 0.24739885479211807,
      "num_tokens": 102041767.0,
      "step": 45710
    },
    {
      "entropy": 5.168567895889282,
      "epoch": 5.134216881001853,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026938779119339585,
      "loss": 4.7132,
      "mean_token_accuracy": 0.24367951601743698,
      "num_tokens": 102052792.0,
      "step": 45715
    },
    {
      "entropy": 5.173182392120362,
      "epoch": 5.13477845790981,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002693476550553584,
      "loss": 4.6742,
      "mean_token_accuracy": 0.2500928893685341,
      "num_tokens": 102064238.0,
      "step": 45720
    },
    {
      "entropy": 5.225995683670044,
      "epoch": 5.1353400348177685,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026930751909729384,
      "loss": 4.711,
      "mean_token_accuracy": 0.24400717318058013,
      "num_tokens": 102075557.0,
      "step": 45725
    },
    {
      "entropy": 5.162983083724976,
      "epoch": 5.135901611725726,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026926738332048013,
      "loss": 4.6223,
      "mean_token_accuracy": 0.2516258850693703,
      "num_tokens": 102086691.0,
      "step": 45730
    },
    {
      "entropy": 5.223370695114136,
      "epoch": 5.136463188633684,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026922724772619516,
      "loss": 4.6805,
      "mean_token_accuracy": 0.2434005320072174,
      "num_tokens": 102097215.0,
      "step": 45735
    },
    {
      "entropy": 5.236551570892334,
      "epoch": 5.137024765541641,
      "grad_norm": 0.97265625,
      "learning_rate": 0.0002691871123157168,
      "loss": 4.7374,
      "mean_token_accuracy": 0.24072379022836685,
      "num_tokens": 102108408.0,
      "step": 45740
    },
    {
      "entropy": 5.218963623046875,
      "epoch": 5.137586342449598,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002691469770903231,
      "loss": 4.7364,
      "mean_token_accuracy": 0.24614720344543456,
      "num_tokens": 102120594.0,
      "step": 45745
    },
    {
      "entropy": 5.1938084125518795,
      "epoch": 5.138147919357556,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002691068420512919,
      "loss": 4.6985,
      "mean_token_accuracy": 0.2419376477599144,
      "num_tokens": 102132306.0,
      "step": 45750
    },
    {
      "entropy": 5.256942272186279,
      "epoch": 5.1387094962655135,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000269066707199901,
      "loss": 4.7939,
      "mean_token_accuracy": 0.24090695828199388,
      "num_tokens": 102143271.0,
      "step": 45755
    },
    {
      "entropy": 5.258875608444214,
      "epoch": 5.139271073173471,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026902657253742855,
      "loss": 4.7622,
      "mean_token_accuracy": 0.2429187387228012,
      "num_tokens": 102154657.0,
      "step": 45760
    },
    {
      "entropy": 5.211712074279785,
      "epoch": 5.139832650081429,
      "grad_norm": 0.96484375,
      "learning_rate": 0.0002689864380651524,
      "loss": 4.6234,
      "mean_token_accuracy": 0.24632084667682647,
      "num_tokens": 102167135.0,
      "step": 45765
    },
    {
      "entropy": 5.124246883392334,
      "epoch": 5.140394226989386,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002689463037843504,
      "loss": 4.6262,
      "mean_token_accuracy": 0.24999971687793732,
      "num_tokens": 102178485.0,
      "step": 45770
    },
    {
      "entropy": 5.188170146942139,
      "epoch": 5.140955803897343,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002689061696963004,
      "loss": 4.7573,
      "mean_token_accuracy": 0.23842287212610244,
      "num_tokens": 102189378.0,
      "step": 45775
    },
    {
      "entropy": 5.194934368133545,
      "epoch": 5.141517380805301,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002688660358022803,
      "loss": 4.7468,
      "mean_token_accuracy": 0.2394522726535797,
      "num_tokens": 102199864.0,
      "step": 45780
    },
    {
      "entropy": 5.151724004745484,
      "epoch": 5.142078957713259,
      "grad_norm": 1.09375,
      "learning_rate": 0.000268825902103568,
      "loss": 4.6108,
      "mean_token_accuracy": 0.24800993800163268,
      "num_tokens": 102210699.0,
      "step": 45785
    },
    {
      "entropy": 5.222584390640259,
      "epoch": 5.142640534621217,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002687857686014414,
      "loss": 4.7588,
      "mean_token_accuracy": 0.24499393999576569,
      "num_tokens": 102221724.0,
      "step": 45790
    },
    {
      "entropy": 5.126923513412476,
      "epoch": 5.143202111529174,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002687456352971783,
      "loss": 4.5196,
      "mean_token_accuracy": 0.25962757766246797,
      "num_tokens": 102232023.0,
      "step": 45795
    },
    {
      "entropy": 5.129644727706909,
      "epoch": 5.143763688437131,
      "grad_norm": 1.125,
      "learning_rate": 0.0002687055021920566,
      "loss": 4.6549,
      "mean_token_accuracy": 0.24941769540309905,
      "num_tokens": 102242526.0,
      "step": 45800
    },
    {
      "entropy": 5.110924577713012,
      "epoch": 5.144325265345089,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002686653692873542,
      "loss": 4.7133,
      "mean_token_accuracy": 0.24207475632429123,
      "num_tokens": 102253271.0,
      "step": 45805
    },
    {
      "entropy": 5.246065521240235,
      "epoch": 5.1448868422530465,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00026862523658434883,
      "loss": 4.7157,
      "mean_token_accuracy": 0.2382829949259758,
      "num_tokens": 102263089.0,
      "step": 45810
    },
    {
      "entropy": 5.1997551918029785,
      "epoch": 5.145448419161004,
      "grad_norm": 1.125,
      "learning_rate": 0.0002685851040843185,
      "loss": 4.6949,
      "mean_token_accuracy": 0.2497993141412735,
      "num_tokens": 102273312.0,
      "step": 45815
    },
    {
      "entropy": 5.2401350975036625,
      "epoch": 5.146009996068962,
      "grad_norm": 1.1875,
      "learning_rate": 0.00026854497178854087,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23854244500398636,
      "num_tokens": 102283940.0,
      "step": 45820
    },
    {
      "entropy": 5.238513374328614,
      "epoch": 5.146571572976919,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002685048396982939,
      "loss": 4.73,
      "mean_token_accuracy": 0.2396980807185173,
      "num_tokens": 102295666.0,
      "step": 45825
    },
    {
      "entropy": 5.190066146850586,
      "epoch": 5.147133149884877,
      "grad_norm": 0.9140625,
      "learning_rate": 0.00026846470781485517,
      "loss": 4.7064,
      "mean_token_accuracy": 0.24144137948751448,
      "num_tokens": 102306463.0,
      "step": 45830
    },
    {
      "entropy": 5.143121671676636,
      "epoch": 5.147694726792834,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002684245761395028,
      "loss": 4.5891,
      "mean_token_accuracy": 0.25069240033626555,
      "num_tokens": 102317443.0,
      "step": 45835
    },
    {
      "entropy": 5.184104013442993,
      "epoch": 5.148256303700792,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026838444467351447,
      "loss": 4.7147,
      "mean_token_accuracy": 0.2406146228313446,
      "num_tokens": 102327924.0,
      "step": 45840
    },
    {
      "entropy": 5.3696473121643065,
      "epoch": 5.14881788060875,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00026834431341816805,
      "loss": 4.8451,
      "mean_token_accuracy": 0.236037278175354,
      "num_tokens": 102339364.0,
      "step": 45845
    },
    {
      "entropy": 5.239344120025635,
      "epoch": 5.149379457516707,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026830418237474126,
      "loss": 4.759,
      "mean_token_accuracy": 0.24186039417982103,
      "num_tokens": 102350947.0,
      "step": 45850
    },
    {
      "entropy": 5.292926549911499,
      "epoch": 5.149941034424664,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002682640515445118,
      "loss": 4.7898,
      "mean_token_accuracy": 0.23719180673360823,
      "num_tokens": 102362412.0,
      "step": 45855
    },
    {
      "entropy": 5.2033487319946286,
      "epoch": 5.150502611332622,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002682239209287577,
      "loss": 4.6901,
      "mean_token_accuracy": 0.24804866909980774,
      "num_tokens": 102373630.0,
      "step": 45860
    },
    {
      "entropy": 5.1687342643737795,
      "epoch": 5.151064188240579,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00026818379052875646,
      "loss": 4.7355,
      "mean_token_accuracy": 0.23553345501422882,
      "num_tokens": 102385081.0,
      "step": 45865
    },
    {
      "entropy": 5.180597352981567,
      "epoch": 5.1516257651485375,
      "grad_norm": 1.0234375,
      "learning_rate": 0.000268143660345786,
      "loss": 4.6106,
      "mean_token_accuracy": 0.24612665623426438,
      "num_tokens": 102396519.0,
      "step": 45870
    },
    {
      "entropy": 5.234278917312622,
      "epoch": 5.152187342056495,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002681035303811241,
      "loss": 4.7525,
      "mean_token_accuracy": 0.23991066813468934,
      "num_tokens": 102406326.0,
      "step": 45875
    },
    {
      "entropy": 5.307725954055786,
      "epoch": 5.152748918964452,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026806340063604845,
      "loss": 4.7951,
      "mean_token_accuracy": 0.2362878739833832,
      "num_tokens": 102416965.0,
      "step": 45880
    },
    {
      "entropy": 5.2201557636260985,
      "epoch": 5.15331049587241,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026802327111183686,
      "loss": 4.6567,
      "mean_token_accuracy": 0.2432813361287117,
      "num_tokens": 102428414.0,
      "step": 45885
    },
    {
      "entropy": 5.205064678192139,
      "epoch": 5.153872072780367,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026798314180976696,
      "loss": 4.6395,
      "mean_token_accuracy": 0.24917691349983215,
      "num_tokens": 102438665.0,
      "step": 45890
    },
    {
      "entropy": 5.25748119354248,
      "epoch": 5.1544336496883245,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026794301273111654,
      "loss": 4.7667,
      "mean_token_accuracy": 0.23905421495437623,
      "num_tokens": 102449828.0,
      "step": 45895
    },
    {
      "entropy": 5.272504472732544,
      "epoch": 5.154995226596283,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002679028838771633,
      "loss": 4.7577,
      "mean_token_accuracy": 0.23832455277442932,
      "num_tokens": 102461007.0,
      "step": 45900
    },
    {
      "entropy": 5.169960021972656,
      "epoch": 5.15555680350424,
      "grad_norm": 1.125,
      "learning_rate": 0.00026786275524918503,
      "loss": 4.6178,
      "mean_token_accuracy": 0.2499765157699585,
      "num_tokens": 102472372.0,
      "step": 45905
    },
    {
      "entropy": 5.174081516265869,
      "epoch": 5.156118380412197,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002678226268484594,
      "loss": 4.5854,
      "mean_token_accuracy": 0.2492253914475441,
      "num_tokens": 102482361.0,
      "step": 45910
    },
    {
      "entropy": 5.060249900817871,
      "epoch": 5.156679957320155,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002677824986762641,
      "loss": 4.5651,
      "mean_token_accuracy": 0.24764775931835176,
      "num_tokens": 102493058.0,
      "step": 45915
    },
    {
      "entropy": 5.263690614700318,
      "epoch": 5.157241534228112,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002677423707338769,
      "loss": 4.7722,
      "mean_token_accuracy": 0.24433727264404298,
      "num_tokens": 102505242.0,
      "step": 45920
    },
    {
      "entropy": 5.225339412689209,
      "epoch": 5.1578031111360705,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002677022430225754,
      "loss": 4.6904,
      "mean_token_accuracy": 0.24676866978406906,
      "num_tokens": 102517580.0,
      "step": 45925
    },
    {
      "entropy": 5.277764368057251,
      "epoch": 5.158364688044028,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026766211554363727,
      "loss": 4.791,
      "mean_token_accuracy": 0.242368283867836,
      "num_tokens": 102527336.0,
      "step": 45930
    },
    {
      "entropy": 5.246016931533814,
      "epoch": 5.158926264951985,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002676219882983403,
      "loss": 4.8022,
      "mean_token_accuracy": 0.23467244803905488,
      "num_tokens": 102538511.0,
      "step": 45935
    },
    {
      "entropy": 5.263937759399414,
      "epoch": 5.159487841859943,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026758186128796204,
      "loss": 4.6882,
      "mean_token_accuracy": 0.24396356642246247,
      "num_tokens": 102549981.0,
      "step": 45940
    },
    {
      "entropy": 5.318395137786865,
      "epoch": 5.1600494187679,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026754173451378014,
      "loss": 4.8,
      "mean_token_accuracy": 0.23993492871522903,
      "num_tokens": 102560504.0,
      "step": 45945
    },
    {
      "entropy": 5.260799789428711,
      "epoch": 5.1606109956758575,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026750160797707245,
      "loss": 4.755,
      "mean_token_accuracy": 0.24878625720739364,
      "num_tokens": 102571136.0,
      "step": 45950
    },
    {
      "entropy": 5.2349474906921385,
      "epoch": 5.161172572583816,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002674614816791164,
      "loss": 4.667,
      "mean_token_accuracy": 0.24855289459228516,
      "num_tokens": 102582879.0,
      "step": 45955
    },
    {
      "entropy": 5.242766571044922,
      "epoch": 5.161734149491773,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026742135562118967,
      "loss": 4.7361,
      "mean_token_accuracy": 0.24684247076511384,
      "num_tokens": 102593171.0,
      "step": 45960
    },
    {
      "entropy": 5.291109323501587,
      "epoch": 5.16229572639973,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002673812298045699,
      "loss": 4.8325,
      "mean_token_accuracy": 0.23451317548751832,
      "num_tokens": 102603816.0,
      "step": 45965
    },
    {
      "entropy": 5.219229125976563,
      "epoch": 5.162857303307688,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026734110423053483,
      "loss": 4.7154,
      "mean_token_accuracy": 0.2468166396021843,
      "num_tokens": 102614360.0,
      "step": 45970
    },
    {
      "entropy": 5.289366769790649,
      "epoch": 5.163418880215645,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026730097890036175,
      "loss": 4.7766,
      "mean_token_accuracy": 0.2319038599729538,
      "num_tokens": 102625320.0,
      "step": 45975
    },
    {
      "entropy": 5.2075542449951175,
      "epoch": 5.163980457123603,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026726085381532867,
      "loss": 4.7205,
      "mean_token_accuracy": 0.2368597462773323,
      "num_tokens": 102635728.0,
      "step": 45980
    },
    {
      "entropy": 5.154795789718628,
      "epoch": 5.164542034031561,
      "grad_norm": 1.125,
      "learning_rate": 0.000267220728976713,
      "loss": 4.6126,
      "mean_token_accuracy": 0.240798319876194,
      "num_tokens": 102645880.0,
      "step": 45985
    },
    {
      "entropy": 5.223967695236206,
      "epoch": 5.165103610939518,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002671806043857923,
      "loss": 4.7365,
      "mean_token_accuracy": 0.24173278063535691,
      "num_tokens": 102657854.0,
      "step": 45990
    },
    {
      "entropy": 5.246494197845459,
      "epoch": 5.165665187847476,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002671404800438442,
      "loss": 4.7446,
      "mean_token_accuracy": 0.24287466257810592,
      "num_tokens": 102669637.0,
      "step": 45995
    },
    {
      "entropy": 5.224615430831909,
      "epoch": 5.166226764755433,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002671003559521463,
      "loss": 4.6328,
      "mean_token_accuracy": 0.24495477229356766,
      "num_tokens": 102680520.0,
      "step": 46000
    },
    {
      "entropy": 5.172711324691773,
      "epoch": 5.16678834166339,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026706023211197613,
      "loss": 4.6399,
      "mean_token_accuracy": 0.23974231630563736,
      "num_tokens": 102691633.0,
      "step": 46005
    },
    {
      "entropy": 5.256130409240723,
      "epoch": 5.1673499185713485,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002670201085246111,
      "loss": 4.7478,
      "mean_token_accuracy": 0.2381551131606102,
      "num_tokens": 102703171.0,
      "step": 46010
    },
    {
      "entropy": 5.2710737705230715,
      "epoch": 5.167911495479306,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026697998519132905,
      "loss": 4.7553,
      "mean_token_accuracy": 0.24453552961349487,
      "num_tokens": 102714636.0,
      "step": 46015
    },
    {
      "entropy": 5.237752103805542,
      "epoch": 5.168473072387264,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026693986211340736,
      "loss": 4.8596,
      "mean_token_accuracy": 0.23910859376192092,
      "num_tokens": 102725554.0,
      "step": 46020
    },
    {
      "entropy": 5.246327495574951,
      "epoch": 5.169034649295221,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026689973929212363,
      "loss": 4.772,
      "mean_token_accuracy": 0.23806630074977875,
      "num_tokens": 102736705.0,
      "step": 46025
    },
    {
      "entropy": 5.322335290908813,
      "epoch": 5.169596226203178,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026685961672875524,
      "loss": 4.8291,
      "mean_token_accuracy": 0.22937917858362197,
      "num_tokens": 102749187.0,
      "step": 46030
    },
    {
      "entropy": 5.2077916145324705,
      "epoch": 5.170157803111136,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002668194944245799,
      "loss": 4.7061,
      "mean_token_accuracy": 0.23873357623815536,
      "num_tokens": 102760095.0,
      "step": 46035
    },
    {
      "entropy": 5.228973531723023,
      "epoch": 5.170719380019094,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000266779372380875,
      "loss": 4.7207,
      "mean_token_accuracy": 0.23501876294612883,
      "num_tokens": 102771086.0,
      "step": 46040
    },
    {
      "entropy": 5.300273895263672,
      "epoch": 5.171280956927051,
      "grad_norm": 1.09375,
      "learning_rate": 0.000266739250598918,
      "loss": 4.7219,
      "mean_token_accuracy": 0.2340897351503372,
      "num_tokens": 102782024.0,
      "step": 46045
    },
    {
      "entropy": 5.349847364425659,
      "epoch": 5.171842533835009,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00026669912907998657,
      "loss": 4.7897,
      "mean_token_accuracy": 0.23902954459190368,
      "num_tokens": 102793654.0,
      "step": 46050
    },
    {
      "entropy": 5.340445947647095,
      "epoch": 5.172404110742966,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000266659007825358,
      "loss": 4.8798,
      "mean_token_accuracy": 0.2326076775789261,
      "num_tokens": 102804908.0,
      "step": 46055
    },
    {
      "entropy": 5.125009679794312,
      "epoch": 5.172965687650924,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026661888683631,
      "loss": 4.6163,
      "mean_token_accuracy": 0.2508807241916656,
      "num_tokens": 102815399.0,
      "step": 46060
    },
    {
      "entropy": 5.174178695678711,
      "epoch": 5.1735272645588815,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026657876611411974,
      "loss": 4.6421,
      "mean_token_accuracy": 0.24806170612573625,
      "num_tokens": 102825734.0,
      "step": 46065
    },
    {
      "entropy": 5.190004777908325,
      "epoch": 5.174088841466839,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026653864566006494,
      "loss": 4.5307,
      "mean_token_accuracy": 0.2571601688861847,
      "num_tokens": 102836251.0,
      "step": 46070
    },
    {
      "entropy": 5.269220304489136,
      "epoch": 5.174650418374797,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002664985254754229,
      "loss": 4.8071,
      "mean_token_accuracy": 0.230001562833786,
      "num_tokens": 102847626.0,
      "step": 46075
    },
    {
      "entropy": 5.18292760848999,
      "epoch": 5.175211995282754,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026645840556147103,
      "loss": 4.6131,
      "mean_token_accuracy": 0.2518995612859726,
      "num_tokens": 102859424.0,
      "step": 46080
    },
    {
      "entropy": 5.175965404510498,
      "epoch": 5.175773572190711,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002664182859194869,
      "loss": 4.6867,
      "mean_token_accuracy": 0.24348453432321548,
      "num_tokens": 102870466.0,
      "step": 46085
    },
    {
      "entropy": 5.165291500091553,
      "epoch": 5.176335149098669,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002663781665507479,
      "loss": 4.7008,
      "mean_token_accuracy": 0.23934139758348466,
      "num_tokens": 102882181.0,
      "step": 46090
    },
    {
      "entropy": 5.21685528755188,
      "epoch": 5.1768967260066265,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026633804745653146,
      "loss": 4.7546,
      "mean_token_accuracy": 0.2454872652888298,
      "num_tokens": 102892562.0,
      "step": 46095
    },
    {
      "entropy": 5.166879940032959,
      "epoch": 5.177458302914584,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026629792863811484,
      "loss": 4.6369,
      "mean_token_accuracy": 0.2561677560210228,
      "num_tokens": 102904380.0,
      "step": 46100
    },
    {
      "entropy": 5.160814428329468,
      "epoch": 5.178019879822542,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002662578100967756,
      "loss": 4.6818,
      "mean_token_accuracy": 0.23948870599269867,
      "num_tokens": 102916018.0,
      "step": 46105
    },
    {
      "entropy": 5.197942066192627,
      "epoch": 5.178581456730499,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002662176918337911,
      "loss": 4.7498,
      "mean_token_accuracy": 0.23194365352392196,
      "num_tokens": 102927885.0,
      "step": 46110
    },
    {
      "entropy": 5.2150074481964115,
      "epoch": 5.179143033638457,
      "grad_norm": 1.125,
      "learning_rate": 0.0002661775738504387,
      "loss": 4.6671,
      "mean_token_accuracy": 0.2490622490644455,
      "num_tokens": 102938343.0,
      "step": 46115
    },
    {
      "entropy": 5.163728046417236,
      "epoch": 5.179704610546414,
      "grad_norm": 0.953125,
      "learning_rate": 0.00026613745614799574,
      "loss": 4.7054,
      "mean_token_accuracy": 0.2375745341181755,
      "num_tokens": 102949341.0,
      "step": 46120
    },
    {
      "entropy": 5.2719440937042235,
      "epoch": 5.180266187454372,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002660973387277396,
      "loss": 4.7503,
      "mean_token_accuracy": 0.2494233787059784,
      "num_tokens": 102959722.0,
      "step": 46125
    },
    {
      "entropy": 5.263595533370972,
      "epoch": 5.18082776436233,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002660572215909477,
      "loss": 4.7422,
      "mean_token_accuracy": 0.24381681829690932,
      "num_tokens": 102971423.0,
      "step": 46130
    },
    {
      "entropy": 5.216787195205688,
      "epoch": 5.181389341270287,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002660171047388973,
      "loss": 4.6933,
      "mean_token_accuracy": 0.255261093378067,
      "num_tokens": 102983218.0,
      "step": 46135
    },
    {
      "entropy": 5.228110122680664,
      "epoch": 5.181950918178244,
      "grad_norm": 0.96484375,
      "learning_rate": 0.00026597698817286576,
      "loss": 4.7439,
      "mean_token_accuracy": 0.24311191588640213,
      "num_tokens": 102995287.0,
      "step": 46140
    },
    {
      "entropy": 5.191701650619507,
      "epoch": 5.182512495086202,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00026593687189413044,
      "loss": 4.7336,
      "mean_token_accuracy": 0.2452790319919586,
      "num_tokens": 103006578.0,
      "step": 46145
    },
    {
      "entropy": 5.255736780166626,
      "epoch": 5.1830740719941595,
      "grad_norm": 1.125,
      "learning_rate": 0.00026589675590396865,
      "loss": 4.7771,
      "mean_token_accuracy": 0.23479497134685517,
      "num_tokens": 103017959.0,
      "step": 46150
    },
    {
      "entropy": 5.2107177734375,
      "epoch": 5.183635648902117,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002658566402036576,
      "loss": 4.7021,
      "mean_token_accuracy": 0.24039676040410995,
      "num_tokens": 103029931.0,
      "step": 46155
    },
    {
      "entropy": 5.191237878799439,
      "epoch": 5.184197225810075,
      "grad_norm": 1.125,
      "learning_rate": 0.0002658165247944747,
      "loss": 4.6519,
      "mean_token_accuracy": 0.2522425904870033,
      "num_tokens": 103041603.0,
      "step": 46160
    },
    {
      "entropy": 5.174942064285278,
      "epoch": 5.184758802718032,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026577640967769727,
      "loss": 4.6567,
      "mean_token_accuracy": 0.24071923941373824,
      "num_tokens": 103053353.0,
      "step": 46165
    },
    {
      "entropy": 5.141137981414795,
      "epoch": 5.18532037962599,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00026573629485460257,
      "loss": 4.6045,
      "mean_token_accuracy": 0.2526226222515106,
      "num_tokens": 103064295.0,
      "step": 46170
    },
    {
      "entropy": 5.216891574859619,
      "epoch": 5.185881956533947,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026569618032646773,
      "loss": 4.7628,
      "mean_token_accuracy": 0.25110805332660674,
      "num_tokens": 103075582.0,
      "step": 46175
    },
    {
      "entropy": 5.18881402015686,
      "epoch": 5.186443533441905,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026565606609457016,
      "loss": 4.719,
      "mean_token_accuracy": 0.2410867229104042,
      "num_tokens": 103087969.0,
      "step": 46180
    },
    {
      "entropy": 5.231142950057984,
      "epoch": 5.187005110349863,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002656159521601871,
      "loss": 4.7249,
      "mean_token_accuracy": 0.23250288069248198,
      "num_tokens": 103099505.0,
      "step": 46185
    },
    {
      "entropy": 5.298913908004761,
      "epoch": 5.18756668725782,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002655758385245956,
      "loss": 4.7661,
      "mean_token_accuracy": 0.23616468012332917,
      "num_tokens": 103110706.0,
      "step": 46190
    },
    {
      "entropy": 5.189636135101319,
      "epoch": 5.188128264165777,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002655357251890732,
      "loss": 4.6248,
      "mean_token_accuracy": 0.25785592198371887,
      "num_tokens": 103121129.0,
      "step": 46195
    },
    {
      "entropy": 5.159530878067017,
      "epoch": 5.188689841073735,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026549561215489695,
      "loss": 4.6828,
      "mean_token_accuracy": 0.24164042621850967,
      "num_tokens": 103130706.0,
      "step": 46200
    },
    {
      "entropy": 5.242730903625488,
      "epoch": 5.189251417981692,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002654554994233442,
      "loss": 4.7869,
      "mean_token_accuracy": 0.2401082307100296,
      "num_tokens": 103140441.0,
      "step": 46205
    },
    {
      "entropy": 5.139419031143189,
      "epoch": 5.1898129948896505,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002654153869956919,
      "loss": 4.6583,
      "mean_token_accuracy": 0.24545431584119798,
      "num_tokens": 103151245.0,
      "step": 46210
    },
    {
      "entropy": 5.2215477466583256,
      "epoch": 5.190374571797608,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002653752748732175,
      "loss": 4.7072,
      "mean_token_accuracy": 0.2464641109108925,
      "num_tokens": 103161985.0,
      "step": 46215
    },
    {
      "entropy": 5.198649740219116,
      "epoch": 5.190936148705565,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000265335163057198,
      "loss": 4.7306,
      "mean_token_accuracy": 0.24480506628751755,
      "num_tokens": 103174471.0,
      "step": 46220
    },
    {
      "entropy": 5.154529047012329,
      "epoch": 5.191497725613523,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002652950515489107,
      "loss": 4.6862,
      "mean_token_accuracy": 0.24884845912456513,
      "num_tokens": 103184853.0,
      "step": 46225
    },
    {
      "entropy": 5.271325159072876,
      "epoch": 5.19205930252148,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002652549403496327,
      "loss": 4.7383,
      "mean_token_accuracy": 0.23328094333410263,
      "num_tokens": 103195909.0,
      "step": 46230
    },
    {
      "entropy": 5.3172691822052,
      "epoch": 5.1926208794294375,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026521482946064123,
      "loss": 4.7976,
      "mean_token_accuracy": 0.23655613660812377,
      "num_tokens": 103206557.0,
      "step": 46235
    },
    {
      "entropy": 5.17404727935791,
      "epoch": 5.193182456337396,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002651747188832134,
      "loss": 4.6555,
      "mean_token_accuracy": 0.25216352194547653,
      "num_tokens": 103216887.0,
      "step": 46240
    },
    {
      "entropy": 5.182194948196411,
      "epoch": 5.193744033245353,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002651346086186262,
      "loss": 4.7401,
      "mean_token_accuracy": 0.24248856604099273,
      "num_tokens": 103228879.0,
      "step": 46245
    },
    {
      "entropy": 5.214390802383423,
      "epoch": 5.194305610153311,
      "grad_norm": 0.9921875,
      "learning_rate": 0.000265094498668157,
      "loss": 4.6809,
      "mean_token_accuracy": 0.23919543027877807,
      "num_tokens": 103240564.0,
      "step": 46250
    },
    {
      "entropy": 5.225641345977783,
      "epoch": 5.194867187061268,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026505438903308285,
      "loss": 4.6972,
      "mean_token_accuracy": 0.2432300940155983,
      "num_tokens": 103251297.0,
      "step": 46255
    },
    {
      "entropy": 5.227239751815796,
      "epoch": 5.195428763969225,
      "grad_norm": 1.125,
      "learning_rate": 0.00026501427971468056,
      "loss": 4.6483,
      "mean_token_accuracy": 0.24962817281484603,
      "num_tokens": 103262041.0,
      "step": 46260
    },
    {
      "entropy": 5.1623341083526615,
      "epoch": 5.1959903408771835,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002649741707142277,
      "loss": 4.6618,
      "mean_token_accuracy": 0.2474430188536644,
      "num_tokens": 103273024.0,
      "step": 46265
    },
    {
      "entropy": 5.181561088562011,
      "epoch": 5.196551917785141,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026493406203300106,
      "loss": 4.7169,
      "mean_token_accuracy": 0.24054075181484222,
      "num_tokens": 103284511.0,
      "step": 46270
    },
    {
      "entropy": 5.225053167343139,
      "epoch": 5.197113494693098,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026489395367227785,
      "loss": 4.6925,
      "mean_token_accuracy": 0.2452929884195328,
      "num_tokens": 103296787.0,
      "step": 46275
    },
    {
      "entropy": 5.26023588180542,
      "epoch": 5.197675071601056,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000264853845633335,
      "loss": 4.6492,
      "mean_token_accuracy": 0.24700212180614473,
      "num_tokens": 103308422.0,
      "step": 46280
    },
    {
      "entropy": 5.145177364349365,
      "epoch": 5.198236648509013,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002648137379174497,
      "loss": 4.6352,
      "mean_token_accuracy": 0.24321162700653076,
      "num_tokens": 103319592.0,
      "step": 46285
    },
    {
      "entropy": 5.155283975601196,
      "epoch": 5.1987982254169705,
      "grad_norm": 1.0,
      "learning_rate": 0.0002647736305258989,
      "loss": 4.6482,
      "mean_token_accuracy": 0.2442108377814293,
      "num_tokens": 103330702.0,
      "step": 46290
    },
    {
      "entropy": 5.130159139633179,
      "epoch": 5.199359802324929,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002647335234599597,
      "loss": 4.633,
      "mean_token_accuracy": 0.24160416573286056,
      "num_tokens": 103342339.0,
      "step": 46295
    },
    {
      "entropy": 5.138622522354126,
      "epoch": 5.199921379232886,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00026469341672090906,
      "loss": 4.6429,
      "mean_token_accuracy": 0.24672046750783921,
      "num_tokens": 103353604.0,
      "step": 46300
    },
    {
      "entropy": 5.193921184539795,
      "epoch": 5.200482956140844,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026465331031002404,
      "loss": 4.6661,
      "mean_token_accuracy": 0.24574832916259765,
      "num_tokens": 103363713.0,
      "step": 46305
    },
    {
      "entropy": 5.238105010986328,
      "epoch": 5.201044533048801,
      "grad_norm": 1.03125,
      "learning_rate": 0.00026461320422858166,
      "loss": 4.747,
      "mean_token_accuracy": 0.23660600334405898,
      "num_tokens": 103374966.0,
      "step": 46310
    },
    {
      "entropy": 5.219928550720215,
      "epoch": 5.201606109956758,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002645730984778589,
      "loss": 4.7271,
      "mean_token_accuracy": 0.2382276847958565,
      "num_tokens": 103387025.0,
      "step": 46315
    },
    {
      "entropy": 5.270795679092407,
      "epoch": 5.202167686864716,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002645329930591327,
      "loss": 4.7744,
      "mean_token_accuracy": 0.23086159080266952,
      "num_tokens": 103397947.0,
      "step": 46320
    },
    {
      "entropy": 5.219827270507812,
      "epoch": 5.202729263772674,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026449288797368006,
      "loss": 4.7619,
      "mean_token_accuracy": 0.24166792929172515,
      "num_tokens": 103409874.0,
      "step": 46325
    },
    {
      "entropy": 5.204881858825684,
      "epoch": 5.203290840680631,
      "grad_norm": 0.9375,
      "learning_rate": 0.00026445278322277796,
      "loss": 4.7445,
      "mean_token_accuracy": 0.2388128325343132,
      "num_tokens": 103423484.0,
      "step": 46330
    },
    {
      "entropy": 5.2000292301177975,
      "epoch": 5.203852417588589,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002644126788077032,
      "loss": 4.6808,
      "mean_token_accuracy": 0.24636847376823426,
      "num_tokens": 103434933.0,
      "step": 46335
    },
    {
      "entropy": 5.225559902191162,
      "epoch": 5.204413994496546,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026437257472973296,
      "loss": 4.7468,
      "mean_token_accuracy": 0.23432089984416962,
      "num_tokens": 103447111.0,
      "step": 46340
    },
    {
      "entropy": 5.193677997589111,
      "epoch": 5.204975571404503,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026433247099014406,
      "loss": 4.6424,
      "mean_token_accuracy": 0.2552023962140083,
      "num_tokens": 103457627.0,
      "step": 46345
    },
    {
      "entropy": 5.181444835662842,
      "epoch": 5.2055371483124615,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002642923675902135,
      "loss": 4.6989,
      "mean_token_accuracy": 0.24559007585048676,
      "num_tokens": 103468047.0,
      "step": 46350
    },
    {
      "entropy": 5.145259618759155,
      "epoch": 5.206098725220419,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000264252264531218,
      "loss": 4.6558,
      "mean_token_accuracy": 0.24610958993434906,
      "num_tokens": 103479671.0,
      "step": 46355
    },
    {
      "entropy": 5.189709091186524,
      "epoch": 5.206660302128377,
      "grad_norm": 1.046875,
      "learning_rate": 0.00026421216181443457,
      "loss": 4.658,
      "mean_token_accuracy": 0.24082011878490447,
      "num_tokens": 103490565.0,
      "step": 46360
    },
    {
      "entropy": 5.204175138473511,
      "epoch": 5.207221879036334,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002641720594411401,
      "loss": 4.6816,
      "mean_token_accuracy": 0.24321310222148895,
      "num_tokens": 103500915.0,
      "step": 46365
    },
    {
      "entropy": 5.251686954498291,
      "epoch": 5.207783455944291,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026413195741261145,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24384283572435378,
      "num_tokens": 103511739.0,
      "step": 46370
    },
    {
      "entropy": 5.236045598983765,
      "epoch": 5.208345032852249,
      "grad_norm": 1.0,
      "learning_rate": 0.0002640918557301256,
      "loss": 4.777,
      "mean_token_accuracy": 0.2356971174478531,
      "num_tokens": 103521957.0,
      "step": 46375
    },
    {
      "entropy": 5.206235361099243,
      "epoch": 5.208906609760207,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002640517543949592,
      "loss": 4.7338,
      "mean_token_accuracy": 0.24041425734758376,
      "num_tokens": 103532511.0,
      "step": 46380
    },
    {
      "entropy": 5.248655652999878,
      "epoch": 5.209468186668164,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002640116534083892,
      "loss": 4.726,
      "mean_token_accuracy": 0.24433131963014604,
      "num_tokens": 103544484.0,
      "step": 46385
    },
    {
      "entropy": 5.190707111358643,
      "epoch": 5.210029763576122,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026397155277169246,
      "loss": 4.662,
      "mean_token_accuracy": 0.24496670663356782,
      "num_tokens": 103555020.0,
      "step": 46390
    },
    {
      "entropy": 5.197873306274414,
      "epoch": 5.210591340484079,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002639314524861457,
      "loss": 4.7386,
      "mean_token_accuracy": 0.24295653104782106,
      "num_tokens": 103565331.0,
      "step": 46395
    },
    {
      "entropy": 5.246997261047364,
      "epoch": 5.211152917392037,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002638913525530258,
      "loss": 4.7521,
      "mean_token_accuracy": 0.24486444890499115,
      "num_tokens": 103576970.0,
      "step": 46400
    },
    {
      "entropy": 5.200417184829712,
      "epoch": 5.2117144942999944,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026385125297360946,
      "loss": 4.7386,
      "mean_token_accuracy": 0.23509986996650695,
      "num_tokens": 103587981.0,
      "step": 46405
    },
    {
      "entropy": 5.194011163711548,
      "epoch": 5.212276071207952,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026381115374917363,
      "loss": 4.6863,
      "mean_token_accuracy": 0.24407629817724227,
      "num_tokens": 103599508.0,
      "step": 46410
    },
    {
      "entropy": 5.196869945526123,
      "epoch": 5.21283764811591,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026377105488099503,
      "loss": 4.6392,
      "mean_token_accuracy": 0.24822556376457214,
      "num_tokens": 103609982.0,
      "step": 46415
    },
    {
      "entropy": 5.181015300750732,
      "epoch": 5.213399225023867,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026373095637035037,
      "loss": 4.6695,
      "mean_token_accuracy": 0.25057094991207124,
      "num_tokens": 103620419.0,
      "step": 46420
    },
    {
      "entropy": 5.1489481925964355,
      "epoch": 5.213960801931824,
      "grad_norm": 1.125,
      "learning_rate": 0.00026369085821851645,
      "loss": 4.6499,
      "mean_token_accuracy": 0.24970355331897737,
      "num_tokens": 103630512.0,
      "step": 46425
    },
    {
      "entropy": 5.222163915634155,
      "epoch": 5.214522378839782,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026365076042677,
      "loss": 4.7544,
      "mean_token_accuracy": 0.24939658343791962,
      "num_tokens": 103641596.0,
      "step": 46430
    },
    {
      "entropy": 5.192125272750855,
      "epoch": 5.2150839557477395,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002636106629963877,
      "loss": 4.7095,
      "mean_token_accuracy": 0.2417452484369278,
      "num_tokens": 103654291.0,
      "step": 46435
    },
    {
      "entropy": 5.11577582359314,
      "epoch": 5.215645532655698,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002635705659286463,
      "loss": 4.5444,
      "mean_token_accuracy": 0.2563401356339455,
      "num_tokens": 103665843.0,
      "step": 46440
    },
    {
      "entropy": 5.3227911472320555,
      "epoch": 5.216207109563655,
      "grad_norm": 1.125,
      "learning_rate": 0.00026353046922482256,
      "loss": 4.8483,
      "mean_token_accuracy": 0.23471831679344177,
      "num_tokens": 103676682.0,
      "step": 46445
    },
    {
      "entropy": 5.230702304840088,
      "epoch": 5.216768686471612,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002634903728861931,
      "loss": 4.7308,
      "mean_token_accuracy": 0.236387799680233,
      "num_tokens": 103687189.0,
      "step": 46450
    },
    {
      "entropy": 5.211511611938477,
      "epoch": 5.21733026337957,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002634502769140346,
      "loss": 4.7075,
      "mean_token_accuracy": 0.23988042622804642,
      "num_tokens": 103698332.0,
      "step": 46455
    },
    {
      "entropy": 5.17362642288208,
      "epoch": 5.217891840287527,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026341018130962384,
      "loss": 4.6597,
      "mean_token_accuracy": 0.24314991980791092,
      "num_tokens": 103708652.0,
      "step": 46460
    },
    {
      "entropy": 5.2362500667572025,
      "epoch": 5.218453417195485,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00026337008607423733,
      "loss": 4.7312,
      "mean_token_accuracy": 0.2348134532570839,
      "num_tokens": 103719275.0,
      "step": 46465
    },
    {
      "entropy": 5.163981294631958,
      "epoch": 5.219014994103443,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00026332999120915184,
      "loss": 4.6944,
      "mean_token_accuracy": 0.24553764909505843,
      "num_tokens": 103730865.0,
      "step": 46470
    },
    {
      "entropy": 5.240354871749878,
      "epoch": 5.2195765710114,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026328989671564395,
      "loss": 4.7685,
      "mean_token_accuracy": 0.23215217888355255,
      "num_tokens": 103743784.0,
      "step": 46475
    },
    {
      "entropy": 5.273145866394043,
      "epoch": 5.220138147919357,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002632498025949902,
      "loss": 4.807,
      "mean_token_accuracy": 0.242661689221859,
      "num_tokens": 103755580.0,
      "step": 46480
    },
    {
      "entropy": 5.280274868011475,
      "epoch": 5.220699724827315,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002632097088484674,
      "loss": 4.8458,
      "mean_token_accuracy": 0.23203040957450866,
      "num_tokens": 103767296.0,
      "step": 46485
    },
    {
      "entropy": 5.227224111557007,
      "epoch": 5.2212613017352725,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026316961547735203,
      "loss": 4.7104,
      "mean_token_accuracy": 0.2475880891084671,
      "num_tokens": 103778176.0,
      "step": 46490
    },
    {
      "entropy": 5.294278192520141,
      "epoch": 5.221822878643231,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026312952248292067,
      "loss": 4.7915,
      "mean_token_accuracy": 0.24382981806993484,
      "num_tokens": 103791499.0,
      "step": 46495
    },
    {
      "entropy": 5.232629203796387,
      "epoch": 5.222384455551188,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002630894298664499,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24450955241918565,
      "num_tokens": 103803102.0,
      "step": 46500
    },
    {
      "entropy": 5.190628623962402,
      "epoch": 5.222946032459145,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00026304933762921637,
      "loss": 4.7235,
      "mean_token_accuracy": 0.2461373433470726,
      "num_tokens": 103814273.0,
      "step": 46505
    },
    {
      "entropy": 5.1242451667785645,
      "epoch": 5.223507609367103,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002630092457724965,
      "loss": 4.5491,
      "mean_token_accuracy": 0.2563540801405907,
      "num_tokens": 103823713.0,
      "step": 46510
    },
    {
      "entropy": 5.145799016952514,
      "epoch": 5.22406918627506,
      "grad_norm": 1.125,
      "learning_rate": 0.0002629691542975669,
      "loss": 4.6032,
      "mean_token_accuracy": 0.2496541067957878,
      "num_tokens": 103834244.0,
      "step": 46515
    },
    {
      "entropy": 5.063331985473633,
      "epoch": 5.224630763183018,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002629290632057041,
      "loss": 4.54,
      "mean_token_accuracy": 0.2553810477256775,
      "num_tokens": 103844561.0,
      "step": 46520
    },
    {
      "entropy": 5.244437026977539,
      "epoch": 5.225192340090976,
      "grad_norm": 0.9296875,
      "learning_rate": 0.0002628889724981846,
      "loss": 4.7632,
      "mean_token_accuracy": 0.23926591724157334,
      "num_tokens": 103856812.0,
      "step": 46525
    },
    {
      "entropy": 5.2847723960876465,
      "epoch": 5.225753916998933,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002628488821762849,
      "loss": 4.8198,
      "mean_token_accuracy": 0.2319359540939331,
      "num_tokens": 103870062.0,
      "step": 46530
    },
    {
      "entropy": 5.252866268157959,
      "epoch": 5.22631549390689,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00026280879224128165,
      "loss": 4.7301,
      "mean_token_accuracy": 0.2369665265083313,
      "num_tokens": 103880763.0,
      "step": 46535
    },
    {
      "entropy": 5.224569702148438,
      "epoch": 5.226877070814848,
      "grad_norm": 1.015625,
      "learning_rate": 0.000262768702694451,
      "loss": 4.738,
      "mean_token_accuracy": 0.23949181586503981,
      "num_tokens": 103892528.0,
      "step": 46540
    },
    {
      "entropy": 5.189310073852539,
      "epoch": 5.227438647722805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002627286135370697,
      "loss": 4.693,
      "mean_token_accuracy": 0.23873379677534104,
      "num_tokens": 103903589.0,
      "step": 46545
    },
    {
      "entropy": 5.206418037414551,
      "epoch": 5.2280002246307635,
      "grad_norm": 1.0546875,
      "learning_rate": 0.000262688524770414,
      "loss": 4.706,
      "mean_token_accuracy": 0.24297319054603578,
      "num_tokens": 103914192.0,
      "step": 46550
    },
    {
      "entropy": 5.199195146560669,
      "epoch": 5.228561801538721,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026264843639576055,
      "loss": 4.6232,
      "mean_token_accuracy": 0.2449869394302368,
      "num_tokens": 103924649.0,
      "step": 46555
    },
    {
      "entropy": 5.107458829879761,
      "epoch": 5.229123378446678,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002626083484143856,
      "loss": 4.626,
      "mean_token_accuracy": 0.25170962512493134,
      "num_tokens": 103935834.0,
      "step": 46560
    },
    {
      "entropy": 5.080631446838379,
      "epoch": 5.229684955354636,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002625682608275657,
      "loss": 4.6305,
      "mean_token_accuracy": 0.24529508650302886,
      "num_tokens": 103947007.0,
      "step": 46565
    },
    {
      "entropy": 5.213112115859985,
      "epoch": 5.230246532262593,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002625281736365772,
      "loss": 4.7308,
      "mean_token_accuracy": 0.2430671289563179,
      "num_tokens": 103958424.0,
      "step": 46570
    },
    {
      "entropy": 5.24429874420166,
      "epoch": 5.2308081091705505,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026248808684269646,
      "loss": 4.7515,
      "mean_token_accuracy": 0.24831059575080872,
      "num_tokens": 103969289.0,
      "step": 46575
    },
    {
      "entropy": 5.318042230606079,
      "epoch": 5.231369686078509,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002624480004472,
      "loss": 4.768,
      "mean_token_accuracy": 0.24252644926309586,
      "num_tokens": 103981081.0,
      "step": 46580
    },
    {
      "entropy": 5.273030757904053,
      "epoch": 5.231931262986466,
      "grad_norm": 1.140625,
      "learning_rate": 0.00026240791445136384,
      "loss": 4.762,
      "mean_token_accuracy": 0.233057501912117,
      "num_tokens": 103990904.0,
      "step": 46585
    },
    {
      "entropy": 5.203725528717041,
      "epoch": 5.232492839894424,
      "grad_norm": 1.234375,
      "learning_rate": 0.0002623678288564647,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24763914048671723,
      "num_tokens": 104001535.0,
      "step": 46590
    },
    {
      "entropy": 5.1650481700897215,
      "epoch": 5.233054416802381,
      "grad_norm": 0.953125,
      "learning_rate": 0.0002623277436637788,
      "loss": 4.6693,
      "mean_token_accuracy": 0.24211165606975554,
      "num_tokens": 104012840.0,
      "step": 46595
    },
    {
      "entropy": 5.1965210914611815,
      "epoch": 5.233615993710338,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002622876588745825,
      "loss": 4.6025,
      "mean_token_accuracy": 0.24846984893083573,
      "num_tokens": 104024035.0,
      "step": 46600
    },
    {
      "entropy": 5.204262447357178,
      "epoch": 5.2341775706182965,
      "grad_norm": 1.015625,
      "learning_rate": 0.000262247574490152,
      "loss": 4.6779,
      "mean_token_accuracy": 0.24702017158269882,
      "num_tokens": 104035391.0,
      "step": 46605
    },
    {
      "entropy": 5.2309730052948,
      "epoch": 5.234739147526254,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002622074905117637,
      "loss": 4.6902,
      "mean_token_accuracy": 0.24883703291416168,
      "num_tokens": 104046645.0,
      "step": 46610
    },
    {
      "entropy": 5.1791363716125485,
      "epoch": 5.235300724434211,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026216740694069387,
      "loss": 4.6336,
      "mean_token_accuracy": 0.2496343418955803,
      "num_tokens": 104057706.0,
      "step": 46615
    },
    {
      "entropy": 5.175825405120849,
      "epoch": 5.235862301342169,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002621273237782188,
      "loss": 4.6978,
      "mean_token_accuracy": 0.239370396733284,
      "num_tokens": 104068076.0,
      "step": 46620
    },
    {
      "entropy": 5.170342302322387,
      "epoch": 5.236423878250126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002620872410256146,
      "loss": 4.7017,
      "mean_token_accuracy": 0.23800785690546036,
      "num_tokens": 104078116.0,
      "step": 46625
    },
    {
      "entropy": 5.109737777709961,
      "epoch": 5.236985455158084,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002620471586841577,
      "loss": 4.5298,
      "mean_token_accuracy": 0.258828105032444,
      "num_tokens": 104089543.0,
      "step": 46630
    },
    {
      "entropy": 5.172997379302979,
      "epoch": 5.2375470320660416,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002620070767551244,
      "loss": 4.6555,
      "mean_token_accuracy": 0.24629030078649522,
      "num_tokens": 104100824.0,
      "step": 46635
    },
    {
      "entropy": 5.192794466018677,
      "epoch": 5.238108608973999,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002619669952397908,
      "loss": 4.7111,
      "mean_token_accuracy": 0.24201604276895522,
      "num_tokens": 104112410.0,
      "step": 46640
    },
    {
      "entropy": 5.181004858016967,
      "epoch": 5.238670185881957,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002619269141394331,
      "loss": 4.7155,
      "mean_token_accuracy": 0.24262471199035646,
      "num_tokens": 104124299.0,
      "step": 46645
    },
    {
      "entropy": 5.336115503311158,
      "epoch": 5.239231762789914,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002618868334553275,
      "loss": 4.7858,
      "mean_token_accuracy": 0.23626653105020523,
      "num_tokens": 104136266.0,
      "step": 46650
    },
    {
      "entropy": 5.244254922866821,
      "epoch": 5.239793339697871,
      "grad_norm": 0.9765625,
      "learning_rate": 0.00026184675318875023,
      "loss": 4.7129,
      "mean_token_accuracy": 0.23558660298585893,
      "num_tokens": 104146449.0,
      "step": 46655
    },
    {
      "entropy": 5.130504322052002,
      "epoch": 5.240354916605829,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00026180667334097733,
      "loss": 4.5968,
      "mean_token_accuracy": 0.2416308656334877,
      "num_tokens": 104156959.0,
      "step": 46660
    },
    {
      "entropy": 5.160965824127198,
      "epoch": 5.240916493513787,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002617665939132852,
      "loss": 4.6488,
      "mean_token_accuracy": 0.2416581079363823,
      "num_tokens": 104167834.0,
      "step": 46665
    },
    {
      "entropy": 5.279350090026855,
      "epoch": 5.241478070421744,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002617265149069498,
      "loss": 4.7699,
      "mean_token_accuracy": 0.23806743770837785,
      "num_tokens": 104180035.0,
      "step": 46670
    },
    {
      "entropy": 5.335179185867309,
      "epoch": 5.242039647329702,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002616864363232474,
      "loss": 4.9224,
      "mean_token_accuracy": 0.2418665885925293,
      "num_tokens": 104192960.0,
      "step": 46675
    },
    {
      "entropy": 5.307682895660401,
      "epoch": 5.242601224237659,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026164635816345395,
      "loss": 4.7587,
      "mean_token_accuracy": 0.23615774065256118,
      "num_tokens": 104204120.0,
      "step": 46680
    },
    {
      "entropy": 5.200141096115113,
      "epoch": 5.243162801145617,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002616062804288457,
      "loss": 4.6891,
      "mean_token_accuracy": 0.24404560774564743,
      "num_tokens": 104215222.0,
      "step": 46685
    },
    {
      "entropy": 5.181023073196411,
      "epoch": 5.2437243780535745,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002615662031206986,
      "loss": 4.6936,
      "mean_token_accuracy": 0.24651899337768554,
      "num_tokens": 104227208.0,
      "step": 46690
    },
    {
      "entropy": 5.274977970123291,
      "epoch": 5.244285954961532,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002615261262402887,
      "loss": 4.7739,
      "mean_token_accuracy": 0.23471542149782182,
      "num_tokens": 104236836.0,
      "step": 46695
    },
    {
      "entropy": 5.2130101203918455,
      "epoch": 5.24484753186949,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026148604978889237,
      "loss": 4.716,
      "mean_token_accuracy": 0.2449305146932602,
      "num_tokens": 104248950.0,
      "step": 46700
    },
    {
      "entropy": 5.223094892501831,
      "epoch": 5.245409108777447,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026144597376778533,
      "loss": 4.7316,
      "mean_token_accuracy": 0.2420998841524124,
      "num_tokens": 104260179.0,
      "step": 46705
    },
    {
      "entropy": 5.259052562713623,
      "epoch": 5.245970685685404,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002614058981782438,
      "loss": 4.8069,
      "mean_token_accuracy": 0.24350954741239547,
      "num_tokens": 104273704.0,
      "step": 46710
    },
    {
      "entropy": 5.2070502758026125,
      "epoch": 5.246532262593362,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026136582302154377,
      "loss": 4.637,
      "mean_token_accuracy": 0.24501150846481323,
      "num_tokens": 104283851.0,
      "step": 46715
    },
    {
      "entropy": 5.195169830322266,
      "epoch": 5.24709383950132,
      "grad_norm": 1.1640625,
      "learning_rate": 0.00026132574829896123,
      "loss": 4.6741,
      "mean_token_accuracy": 0.2572523280978203,
      "num_tokens": 104296064.0,
      "step": 46720
    },
    {
      "entropy": 5.26165132522583,
      "epoch": 5.247655416409277,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002612856740117721,
      "loss": 4.8196,
      "mean_token_accuracy": 0.23374126553535463,
      "num_tokens": 104309016.0,
      "step": 46725
    },
    {
      "entropy": 5.232041454315185,
      "epoch": 5.248216993317235,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00026124560016125236,
      "loss": 4.7264,
      "mean_token_accuracy": 0.2397608146071434,
      "num_tokens": 104320598.0,
      "step": 46730
    },
    {
      "entropy": 5.299756622314453,
      "epoch": 5.248778570225192,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002612055267486782,
      "loss": 4.752,
      "mean_token_accuracy": 0.23595504462718964,
      "num_tokens": 104332055.0,
      "step": 46735
    },
    {
      "entropy": 5.21627950668335,
      "epoch": 5.24934014713315,
      "grad_norm": 1.078125,
      "learning_rate": 0.00026116545377532536,
      "loss": 4.7169,
      "mean_token_accuracy": 0.2434979036450386,
      "num_tokens": 104343879.0,
      "step": 46740
    },
    {
      "entropy": 5.201332855224609,
      "epoch": 5.249901724041107,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002611253812424698,
      "loss": 4.7145,
      "mean_token_accuracy": 0.2380588859319687,
      "num_tokens": 104354653.0,
      "step": 46745
    },
    {
      "entropy": 5.252913475036621,
      "epoch": 5.250463300949065,
      "grad_norm": 0.9765625,
      "learning_rate": 0.0002610853091513876,
      "loss": 4.7741,
      "mean_token_accuracy": 0.24092668443918228,
      "num_tokens": 104365273.0,
      "step": 46750
    },
    {
      "entropy": 5.283675479888916,
      "epoch": 5.251024877857023,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00026104523750335445,
      "loss": 4.8025,
      "mean_token_accuracy": 0.2326451227068901,
      "num_tokens": 104375990.0,
      "step": 46755
    },
    {
      "entropy": 5.201947164535523,
      "epoch": 5.25158645476498,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002610051662996464,
      "loss": 4.5976,
      "mean_token_accuracy": 0.25279090702533724,
      "num_tokens": 104385975.0,
      "step": 46760
    },
    {
      "entropy": 5.246315622329712,
      "epoch": 5.252148031672937,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026096509554153914,
      "loss": 4.7209,
      "mean_token_accuracy": 0.2461743414402008,
      "num_tokens": 104396053.0,
      "step": 46765
    },
    {
      "entropy": 5.140825414657593,
      "epoch": 5.252709608580895,
      "grad_norm": 1.1875,
      "learning_rate": 0.0002609250252303088,
      "loss": 4.607,
      "mean_token_accuracy": 0.2527851045131683,
      "num_tokens": 104407874.0,
      "step": 46770
    },
    {
      "entropy": 5.167941331863403,
      "epoch": 5.2532711854888525,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002608849553672311,
      "loss": 4.6933,
      "mean_token_accuracy": 0.24899043887853622,
      "num_tokens": 104418527.0,
      "step": 46775
    },
    {
      "entropy": 5.1758466243743895,
      "epoch": 5.253832762396811,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00026084488595358197,
      "loss": 4.7125,
      "mean_token_accuracy": 0.24288785755634307,
      "num_tokens": 104430490.0,
      "step": 46780
    },
    {
      "entropy": 5.224857473373413,
      "epoch": 5.254394339304768,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002608048169906371,
      "loss": 4.6393,
      "mean_token_accuracy": 0.2502657905220985,
      "num_tokens": 104441767.0,
      "step": 46785
    },
    {
      "entropy": 5.18948884010315,
      "epoch": 5.254955916212725,
      "grad_norm": 1.109375,
      "learning_rate": 0.00026076474847967236,
      "loss": 4.7166,
      "mean_token_accuracy": 0.24560293257236482,
      "num_tokens": 104452720.0,
      "step": 46790
    },
    {
      "entropy": 5.257784461975097,
      "epoch": 5.255517493120683,
      "grad_norm": 1.09375,
      "learning_rate": 0.00026072468042196356,
      "loss": 4.7304,
      "mean_token_accuracy": 0.24246390759944916,
      "num_tokens": 104463829.0,
      "step": 46795
    },
    {
      "entropy": 5.238269519805908,
      "epoch": 5.25607907002864,
      "grad_norm": 1.171875,
      "learning_rate": 0.00026068461281878645,
      "loss": 4.6959,
      "mean_token_accuracy": 0.24410975873470306,
      "num_tokens": 104474830.0,
      "step": 46800
    },
    {
      "entropy": 5.199528646469116,
      "epoch": 5.256640646936598,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00026064454567141683,
      "loss": 4.664,
      "mean_token_accuracy": 0.24269889146089554,
      "num_tokens": 104485518.0,
      "step": 46805
    },
    {
      "entropy": 5.270404815673828,
      "epoch": 5.257202223844556,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002606044789811304,
      "loss": 4.7401,
      "mean_token_accuracy": 0.24063756167888642,
      "num_tokens": 104495562.0,
      "step": 46810
    },
    {
      "entropy": 5.215659761428833,
      "epoch": 5.257763800752513,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000260564412749203,
      "loss": 4.6566,
      "mean_token_accuracy": 0.24586663544178008,
      "num_tokens": 104506057.0,
      "step": 46815
    },
    {
      "entropy": 5.169397211074829,
      "epoch": 5.258325377660471,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00026052434697691035,
      "loss": 4.6577,
      "mean_token_accuracy": 0.2557516172528267,
      "num_tokens": 104518269.0,
      "step": 46820
    },
    {
      "entropy": 5.182109689712524,
      "epoch": 5.258886954568428,
      "grad_norm": 1.03125,
      "learning_rate": 0.000260484281665528,
      "loss": 4.6529,
      "mean_token_accuracy": 0.2466243714094162,
      "num_tokens": 104528469.0,
      "step": 46825
    },
    {
      "entropy": 5.249157571792603,
      "epoch": 5.2594485314763855,
      "grad_norm": 1.125,
      "learning_rate": 0.0002604442168163318,
      "loss": 4.7541,
      "mean_token_accuracy": 0.23728677183389663,
      "num_tokens": 104539079.0,
      "step": 46830
    },
    {
      "entropy": 5.2337103366851805,
      "epoch": 5.260010108384344,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002604041524305974,
      "loss": 4.8314,
      "mean_token_accuracy": 0.23888200372457505,
      "num_tokens": 104551235.0,
      "step": 46835
    },
    {
      "entropy": 5.3149041652679445,
      "epoch": 5.260571685292301,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002603640885096004,
      "loss": 4.812,
      "mean_token_accuracy": 0.2361825153231621,
      "num_tokens": 104563911.0,
      "step": 46840
    },
    {
      "entropy": 5.225880193710327,
      "epoch": 5.261133262200258,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002603240250546165,
      "loss": 4.748,
      "mean_token_accuracy": 0.2407286673784256,
      "num_tokens": 104574877.0,
      "step": 46845
    },
    {
      "entropy": 5.306664037704468,
      "epoch": 5.261694839108216,
      "grad_norm": 1.125,
      "learning_rate": 0.00026028396206692134,
      "loss": 4.7637,
      "mean_token_accuracy": 0.24005619585514068,
      "num_tokens": 104585209.0,
      "step": 46850
    },
    {
      "entropy": 5.167034244537353,
      "epoch": 5.262256416016173,
      "grad_norm": 1.0625,
      "learning_rate": 0.00026024389954779054,
      "loss": 4.6728,
      "mean_token_accuracy": 0.25090349167585374,
      "num_tokens": 104596213.0,
      "step": 46855
    },
    {
      "entropy": 5.176692628860474,
      "epoch": 5.2628179929241306,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002602038374984997,
      "loss": 4.6843,
      "mean_token_accuracy": 0.23926090002059935,
      "num_tokens": 104606813.0,
      "step": 46860
    },
    {
      "entropy": 5.226561975479126,
      "epoch": 5.263379569832089,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00026016377592032446,
      "loss": 4.7605,
      "mean_token_accuracy": 0.2448291927576065,
      "num_tokens": 104619040.0,
      "step": 46865
    },
    {
      "entropy": 5.073655843734741,
      "epoch": 5.263941146740046,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002601237148145403,
      "loss": 4.551,
      "mean_token_accuracy": 0.25132429599761963,
      "num_tokens": 104629879.0,
      "step": 46870
    },
    {
      "entropy": 5.143940782546997,
      "epoch": 5.264502723648004,
      "grad_norm": 1.15625,
      "learning_rate": 0.00026008365418242284,
      "loss": 4.6797,
      "mean_token_accuracy": 0.2479760766029358,
      "num_tokens": 104641408.0,
      "step": 46875
    },
    {
      "entropy": 5.311194372177124,
      "epoch": 5.265064300555961,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002600435940252476,
      "loss": 4.8165,
      "mean_token_accuracy": 0.23811911195516586,
      "num_tokens": 104653537.0,
      "step": 46880
    },
    {
      "entropy": 5.161414766311646,
      "epoch": 5.265625877463918,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002600035343442902,
      "loss": 4.6136,
      "mean_token_accuracy": 0.2530601382255554,
      "num_tokens": 104665499.0,
      "step": 46885
    },
    {
      "entropy": 5.234234714508057,
      "epoch": 5.2661874543718765,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00025996347514082603,
      "loss": 4.7515,
      "mean_token_accuracy": 0.23973702937364577,
      "num_tokens": 104676964.0,
      "step": 46890
    },
    {
      "entropy": 5.309612894058228,
      "epoch": 5.266749031279834,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025992341641613065,
      "loss": 4.7659,
      "mean_token_accuracy": 0.23815751820802689,
      "num_tokens": 104688853.0,
      "step": 46895
    },
    {
      "entropy": 5.19920392036438,
      "epoch": 5.267310608187791,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002598833581714795,
      "loss": 4.6593,
      "mean_token_accuracy": 0.24415512084960939,
      "num_tokens": 104699723.0,
      "step": 46900
    },
    {
      "entropy": 5.197403812408448,
      "epoch": 5.267872185095749,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025984330040814817,
      "loss": 4.7059,
      "mean_token_accuracy": 0.24675113409757615,
      "num_tokens": 104711370.0,
      "step": 46905
    },
    {
      "entropy": 5.2657393455505375,
      "epoch": 5.268433762003706,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002598032431274119,
      "loss": 4.8018,
      "mean_token_accuracy": 0.2361813247203827,
      "num_tokens": 104722939.0,
      "step": 46910
    },
    {
      "entropy": 5.288329696655273,
      "epoch": 5.2689953389116635,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00025976318633054636,
      "loss": 4.8178,
      "mean_token_accuracy": 0.23867819905281068,
      "num_tokens": 104735433.0,
      "step": 46915
    },
    {
      "entropy": 5.228055381774903,
      "epoch": 5.269556915819622,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002597231300188269,
      "loss": 4.7108,
      "mean_token_accuracy": 0.2440870851278305,
      "num_tokens": 104748325.0,
      "step": 46920
    },
    {
      "entropy": 5.205322313308716,
      "epoch": 5.270118492727579,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025968307419352884,
      "loss": 4.7512,
      "mean_token_accuracy": 0.24197482019662858,
      "num_tokens": 104758745.0,
      "step": 46925
    },
    {
      "entropy": 5.172630167007446,
      "epoch": 5.270680069635537,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025964301885592766,
      "loss": 4.6182,
      "mean_token_accuracy": 0.24834054857492446,
      "num_tokens": 104769424.0,
      "step": 46930
    },
    {
      "entropy": 5.142558288574219,
      "epoch": 5.271241646543494,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002596029640072988,
      "loss": 4.6409,
      "mean_token_accuracy": 0.24913930296897888,
      "num_tokens": 104779913.0,
      "step": 46935
    },
    {
      "entropy": 5.166873931884766,
      "epoch": 5.271803223451451,
      "grad_norm": 1.0,
      "learning_rate": 0.0002595629096489174,
      "loss": 4.7157,
      "mean_token_accuracy": 0.2359895259141922,
      "num_tokens": 104793066.0,
      "step": 46940
    },
    {
      "entropy": 5.171387004852295,
      "epoch": 5.2723648003594095,
      "grad_norm": 1.0625,
      "learning_rate": 0.000259522855782059,
      "loss": 4.6676,
      "mean_token_accuracy": 0.2433204770088196,
      "num_tokens": 104803641.0,
      "step": 46945
    },
    {
      "entropy": 5.306146860122681,
      "epoch": 5.272926377267367,
      "grad_norm": 1.0,
      "learning_rate": 0.0002594828024079989,
      "loss": 4.7948,
      "mean_token_accuracy": 0.22630584388971328,
      "num_tokens": 104816025.0,
      "step": 46950
    },
    {
      "entropy": 5.281777334213257,
      "epoch": 5.273487954175324,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025944274952801233,
      "loss": 4.8033,
      "mean_token_accuracy": 0.23842816352844237,
      "num_tokens": 104828188.0,
      "step": 46955
    },
    {
      "entropy": 5.231152105331421,
      "epoch": 5.274049531083282,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002594026971433747,
      "loss": 4.7488,
      "mean_token_accuracy": 0.23583829253911973,
      "num_tokens": 104838276.0,
      "step": 46960
    },
    {
      "entropy": 5.232636785507202,
      "epoch": 5.274611107991239,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025936264525536126,
      "loss": 4.7522,
      "mean_token_accuracy": 0.24095114022493364,
      "num_tokens": 104848295.0,
      "step": 46965
    },
    {
      "entropy": 5.290820598602295,
      "epoch": 5.275172684899197,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002593225938652472,
      "loss": 4.8307,
      "mean_token_accuracy": 0.23836130201816558,
      "num_tokens": 104857857.0,
      "step": 46970
    },
    {
      "entropy": 5.199747323989868,
      "epoch": 5.2757342618071545,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002592825429743079,
      "loss": 4.7137,
      "mean_token_accuracy": 0.2412438988685608,
      "num_tokens": 104867819.0,
      "step": 46975
    },
    {
      "entropy": 5.296141862869263,
      "epoch": 5.276295838715112,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002592424925838185,
      "loss": 4.732,
      "mean_token_accuracy": 0.23789745569229126,
      "num_tokens": 104879034.0,
      "step": 46980
    },
    {
      "entropy": 5.297714567184448,
      "epoch": 5.27685741562307,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002592024426950541,
      "loss": 4.828,
      "mean_token_accuracy": 0.23481700718402862,
      "num_tokens": 104890245.0,
      "step": 46985
    },
    {
      "entropy": 5.237699413299561,
      "epoch": 5.277418992531027,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025916239330929026,
      "loss": 4.7479,
      "mean_token_accuracy": 0.23494412153959274,
      "num_tokens": 104900805.0,
      "step": 46990
    },
    {
      "entropy": 5.181954002380371,
      "epoch": 5.277980569438984,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025912234442780185,
      "loss": 4.6299,
      "mean_token_accuracy": 0.2537124440073967,
      "num_tokens": 104911636.0,
      "step": 46995
    },
    {
      "entropy": 5.241587591171265,
      "epoch": 5.278542146346942,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025908229605186423,
      "loss": 4.705,
      "mean_token_accuracy": 0.24387421309947968,
      "num_tokens": 104922253.0,
      "step": 47000
    },
    {
      "entropy": 5.157780933380127,
      "epoch": 5.2791037232549,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025904224818275247,
      "loss": 4.6891,
      "mean_token_accuracy": 0.246525377035141,
      "num_tokens": 104933274.0,
      "step": 47005
    },
    {
      "entropy": 5.243972730636597,
      "epoch": 5.279665300162858,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025900220082174165,
      "loss": 4.771,
      "mean_token_accuracy": 0.23621198087930678,
      "num_tokens": 104943931.0,
      "step": 47010
    },
    {
      "entropy": 5.152469158172607,
      "epoch": 5.280226877070815,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025896215397010696,
      "loss": 4.6684,
      "mean_token_accuracy": 0.2440330743789673,
      "num_tokens": 104955080.0,
      "step": 47015
    },
    {
      "entropy": 5.206324338912964,
      "epoch": 5.280788453978772,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002589221076291236,
      "loss": 4.6542,
      "mean_token_accuracy": 0.2492628961801529,
      "num_tokens": 104967232.0,
      "step": 47020
    },
    {
      "entropy": 5.244699287414551,
      "epoch": 5.28135003088673,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025888206180006647,
      "loss": 4.7632,
      "mean_token_accuracy": 0.2415197730064392,
      "num_tokens": 104979462.0,
      "step": 47025
    },
    {
      "entropy": 5.237729597091675,
      "epoch": 5.2819116077946875,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002588420164842108,
      "loss": 4.7714,
      "mean_token_accuracy": 0.2407162070274353,
      "num_tokens": 104990567.0,
      "step": 47030
    },
    {
      "entropy": 5.120176553726196,
      "epoch": 5.282473184702645,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002588019716828316,
      "loss": 4.5779,
      "mean_token_accuracy": 0.2520717054605484,
      "num_tokens": 105001965.0,
      "step": 47035
    },
    {
      "entropy": 5.158331346511841,
      "epoch": 5.283034761610603,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025876192739720383,
      "loss": 4.5819,
      "mean_token_accuracy": 0.25201545357704164,
      "num_tokens": 105013349.0,
      "step": 47040
    },
    {
      "entropy": 5.196631669998169,
      "epoch": 5.28359633851856,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025872188362860264,
      "loss": 4.7116,
      "mean_token_accuracy": 0.23995840847492217,
      "num_tokens": 105025009.0,
      "step": 47045
    },
    {
      "entropy": 5.134720993041992,
      "epoch": 5.284157915426517,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025868184037830305,
      "loss": 4.6066,
      "mean_token_accuracy": 0.25068851113319396,
      "num_tokens": 105035474.0,
      "step": 47050
    },
    {
      "entropy": 5.282893896102905,
      "epoch": 5.284719492334475,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002586417976475798,
      "loss": 4.7728,
      "mean_token_accuracy": 0.23813675343990326,
      "num_tokens": 105047310.0,
      "step": 47055
    },
    {
      "entropy": 5.167475748062134,
      "epoch": 5.285281069242433,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025860175543770817,
      "loss": 4.5948,
      "mean_token_accuracy": 0.24773095399141312,
      "num_tokens": 105057198.0,
      "step": 47060
    },
    {
      "entropy": 5.2686584949493405,
      "epoch": 5.285842646150391,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002585617137499631,
      "loss": 4.7225,
      "mean_token_accuracy": 0.24303117394447327,
      "num_tokens": 105067885.0,
      "step": 47065
    },
    {
      "entropy": 5.217652320861816,
      "epoch": 5.286404223058348,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002585216725856193,
      "loss": 4.7423,
      "mean_token_accuracy": 0.2398897111415863,
      "num_tokens": 105078917.0,
      "step": 47070
    },
    {
      "entropy": 5.251429033279419,
      "epoch": 5.286965799966305,
      "grad_norm": 1.1796875,
      "learning_rate": 0.0002584816319459519,
      "loss": 4.8026,
      "mean_token_accuracy": 0.2362357571721077,
      "num_tokens": 105089425.0,
      "step": 47075
    },
    {
      "entropy": 5.320439291000366,
      "epoch": 5.287527376874263,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002584415918322358,
      "loss": 4.7142,
      "mean_token_accuracy": 0.24294593334197997,
      "num_tokens": 105100043.0,
      "step": 47080
    },
    {
      "entropy": 5.159366416931152,
      "epoch": 5.28808895378222,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002584015522457458,
      "loss": 4.627,
      "mean_token_accuracy": 0.24560978263616562,
      "num_tokens": 105110317.0,
      "step": 47085
    },
    {
      "entropy": 5.233426475524903,
      "epoch": 5.288650530690178,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00025836151318775675,
      "loss": 4.7332,
      "mean_token_accuracy": 0.23898347616195678,
      "num_tokens": 105120688.0,
      "step": 47090
    },
    {
      "entropy": 5.323741340637207,
      "epoch": 5.289212107598136,
      "grad_norm": 1.140625,
      "learning_rate": 0.00025832147465954365,
      "loss": 4.8223,
      "mean_token_accuracy": 0.23582988679409028,
      "num_tokens": 105130883.0,
      "step": 47095
    },
    {
      "entropy": 5.316677141189575,
      "epoch": 5.289773684506093,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025828143666238124,
      "loss": 4.774,
      "mean_token_accuracy": 0.23687463402748107,
      "num_tokens": 105141203.0,
      "step": 47100
    },
    {
      "entropy": 5.216434907913208,
      "epoch": 5.29033526141405,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025824139919754436,
      "loss": 4.7008,
      "mean_token_accuracy": 0.2462934136390686,
      "num_tokens": 105152419.0,
      "step": 47105
    },
    {
      "entropy": 5.302488803863525,
      "epoch": 5.290896838322008,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002582013622663079,
      "loss": 4.8428,
      "mean_token_accuracy": 0.23688348680734633,
      "num_tokens": 105164242.0,
      "step": 47110
    },
    {
      "entropy": 5.277243375778198,
      "epoch": 5.2914584152299655,
      "grad_norm": 1.125,
      "learning_rate": 0.0002581613258699465,
      "loss": 4.752,
      "mean_token_accuracy": 0.24569794535636902,
      "num_tokens": 105174008.0,
      "step": 47115
    },
    {
      "entropy": 5.170640754699707,
      "epoch": 5.292019992137924,
      "grad_norm": 1.1015625,
      "learning_rate": 0.000258121290009735,
      "loss": 4.6982,
      "mean_token_accuracy": 0.24332326650619507,
      "num_tokens": 105184880.0,
      "step": 47120
    },
    {
      "entropy": 5.231653261184692,
      "epoch": 5.292581569045881,
      "grad_norm": 0.984375,
      "learning_rate": 0.0002580812546869481,
      "loss": 4.6792,
      "mean_token_accuracy": 0.24045203626155853,
      "num_tokens": 105196492.0,
      "step": 47125
    },
    {
      "entropy": 5.153287696838379,
      "epoch": 5.293143145953838,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002580412199028607,
      "loss": 4.669,
      "mean_token_accuracy": 0.24331005811691284,
      "num_tokens": 105206227.0,
      "step": 47130
    },
    {
      "entropy": 5.354058933258057,
      "epoch": 5.293704722861796,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025800118565874754,
      "loss": 4.8823,
      "mean_token_accuracy": 0.23257537186145782,
      "num_tokens": 105217877.0,
      "step": 47135
    },
    {
      "entropy": 5.237616109848022,
      "epoch": 5.294266299769753,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002579611519558831,
      "loss": 4.6555,
      "mean_token_accuracy": 0.24349069148302077,
      "num_tokens": 105229124.0,
      "step": 47140
    },
    {
      "entropy": 5.139599704742432,
      "epoch": 5.294827876677711,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025792111879554224,
      "loss": 4.5764,
      "mean_token_accuracy": 0.2547999292612076,
      "num_tokens": 105240952.0,
      "step": 47145
    },
    {
      "entropy": 5.188442182540894,
      "epoch": 5.295389453585669,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002578810861789996,
      "loss": 4.7056,
      "mean_token_accuracy": 0.24538663178682327,
      "num_tokens": 105252594.0,
      "step": 47150
    },
    {
      "entropy": 5.26041693687439,
      "epoch": 5.295951030493626,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002578410541075298,
      "loss": 4.7943,
      "mean_token_accuracy": 0.2378400593996048,
      "num_tokens": 105263457.0,
      "step": 47155
    },
    {
      "entropy": 5.2333385944366455,
      "epoch": 5.296512607401584,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025780102258240745,
      "loss": 4.7039,
      "mean_token_accuracy": 0.2512252777814865,
      "num_tokens": 105275455.0,
      "step": 47160
    },
    {
      "entropy": 5.29292140007019,
      "epoch": 5.297074184309541,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002577609916049072,
      "loss": 4.7465,
      "mean_token_accuracy": 0.2409933775663376,
      "num_tokens": 105285643.0,
      "step": 47165
    },
    {
      "entropy": 5.21763277053833,
      "epoch": 5.2976357612174985,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002577209611763037,
      "loss": 4.7333,
      "mean_token_accuracy": 0.24290496557950975,
      "num_tokens": 105297322.0,
      "step": 47170
    },
    {
      "entropy": 5.1541650772094725,
      "epoch": 5.298197338125457,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002576809312978715,
      "loss": 4.6124,
      "mean_token_accuracy": 0.2446564868092537,
      "num_tokens": 105307977.0,
      "step": 47175
    },
    {
      "entropy": 5.201925039291382,
      "epoch": 5.298758915033414,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002576409019708852,
      "loss": 4.7012,
      "mean_token_accuracy": 0.24575281888246536,
      "num_tokens": 105318461.0,
      "step": 47180
    },
    {
      "entropy": 5.32963194847107,
      "epoch": 5.299320491941371,
      "grad_norm": 0.99609375,
      "learning_rate": 0.0002576008731966193,
      "loss": 4.8356,
      "mean_token_accuracy": 0.23608970493078232,
      "num_tokens": 105329494.0,
      "step": 47185
    },
    {
      "entropy": 5.100331211090088,
      "epoch": 5.299882068849329,
      "grad_norm": 1.015625,
      "learning_rate": 0.00025756084497634827,
      "loss": 4.5476,
      "mean_token_accuracy": 0.2513082504272461,
      "num_tokens": 105341475.0,
      "step": 47190
    },
    {
      "entropy": 5.175302743911743,
      "epoch": 5.300443645757286,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002575208173113468,
      "loss": 4.598,
      "mean_token_accuracy": 0.25330994576215743,
      "num_tokens": 105352075.0,
      "step": 47195
    },
    {
      "entropy": 5.247788715362549,
      "epoch": 5.301005222665244,
      "grad_norm": 0.9609375,
      "learning_rate": 0.0002574807902028891,
      "loss": 4.7432,
      "mean_token_accuracy": 0.23589333593845369,
      "num_tokens": 105363801.0,
      "step": 47200
    },
    {
      "entropy": 5.273384666442871,
      "epoch": 5.301566799573202,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025744076365225,
      "loss": 4.7298,
      "mean_token_accuracy": 0.24718472808599473,
      "num_tokens": 105374821.0,
      "step": 47205
    },
    {
      "entropy": 5.1826235294342045,
      "epoch": 5.302128376481159,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025740073766070376,
      "loss": 4.6485,
      "mean_token_accuracy": 0.2527484744787216,
      "num_tokens": 105385890.0,
      "step": 47210
    },
    {
      "entropy": 5.266710472106934,
      "epoch": 5.302689953389117,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002573607122295249,
      "loss": 4.7619,
      "mean_token_accuracy": 0.23594847470521926,
      "num_tokens": 105396943.0,
      "step": 47215
    },
    {
      "entropy": 5.238422060012818,
      "epoch": 5.303251530297074,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002573206873599878,
      "loss": 4.771,
      "mean_token_accuracy": 0.24012994021177292,
      "num_tokens": 105407223.0,
      "step": 47220
    },
    {
      "entropy": 5.2577942371368405,
      "epoch": 5.303813107205031,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025728066305336685,
      "loss": 4.7107,
      "mean_token_accuracy": 0.2427218049764633,
      "num_tokens": 105417714.0,
      "step": 47225
    },
    {
      "entropy": 5.3653637886047365,
      "epoch": 5.3043746841129895,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002572406393109364,
      "loss": 4.8664,
      "mean_token_accuracy": 0.24130997508764268,
      "num_tokens": 105429035.0,
      "step": 47230
    },
    {
      "entropy": 5.234266757965088,
      "epoch": 5.304936261020947,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002572006161339709,
      "loss": 4.7123,
      "mean_token_accuracy": 0.23672604858875274,
      "num_tokens": 105440347.0,
      "step": 47235
    },
    {
      "entropy": 5.282614612579346,
      "epoch": 5.305497837928904,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002571605935237448,
      "loss": 4.7943,
      "mean_token_accuracy": 0.24173162132501602,
      "num_tokens": 105451265.0,
      "step": 47240
    },
    {
      "entropy": 5.279127359390259,
      "epoch": 5.306059414836862,
      "grad_norm": 1.0546875,
      "learning_rate": 0.0002571205714815322,
      "loss": 4.7837,
      "mean_token_accuracy": 0.2297774538397789,
      "num_tokens": 105462034.0,
      "step": 47245
    },
    {
      "entropy": 5.215123176574707,
      "epoch": 5.306620991744819,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002570805500086076,
      "loss": 4.684,
      "mean_token_accuracy": 0.2452326938509941,
      "num_tokens": 105472733.0,
      "step": 47250
    },
    {
      "entropy": 5.272306489944458,
      "epoch": 5.307182568652777,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002570405291062452,
      "loss": 4.7918,
      "mean_token_accuracy": 0.2402726098895073,
      "num_tokens": 105483152.0,
      "step": 47255
    },
    {
      "entropy": 5.1699381351470945,
      "epoch": 5.307744145560735,
      "grad_norm": 1.140625,
      "learning_rate": 0.0002570005087757193,
      "loss": 4.6075,
      "mean_token_accuracy": 0.25129174441099167,
      "num_tokens": 105493541.0,
      "step": 47260
    },
    {
      "entropy": 5.218952608108521,
      "epoch": 5.308305722468692,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002569604890183043,
      "loss": 4.7267,
      "mean_token_accuracy": 0.2448549449443817,
      "num_tokens": 105505550.0,
      "step": 47265
    },
    {
      "entropy": 5.143508672714233,
      "epoch": 5.30886729937665,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002569204698352741,
      "loss": 4.6922,
      "mean_token_accuracy": 0.24564552903175355,
      "num_tokens": 105518460.0,
      "step": 47270
    },
    {
      "entropy": 5.197145652770996,
      "epoch": 5.309428876284607,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002568804512279033,
      "loss": 4.6699,
      "mean_token_accuracy": 0.2479478508234024,
      "num_tokens": 105529074.0,
      "step": 47275
    },
    {
      "entropy": 5.24536623954773,
      "epoch": 5.309990453192564,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025684043319746595,
      "loss": 4.7101,
      "mean_token_accuracy": 0.24609422832727432,
      "num_tokens": 105540501.0,
      "step": 47280
    },
    {
      "entropy": 5.189736986160279,
      "epoch": 5.3105520301005225,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025680041574523626,
      "loss": 4.6779,
      "mean_token_accuracy": 0.2425648584961891,
      "num_tokens": 105551158.0,
      "step": 47285
    },
    {
      "entropy": 5.1941954612731935,
      "epoch": 5.31111360700848,
      "grad_norm": 0.95703125,
      "learning_rate": 0.0002567603988724884,
      "loss": 4.6648,
      "mean_token_accuracy": 0.25304243713617325,
      "num_tokens": 105563219.0,
      "step": 47290
    },
    {
      "entropy": 5.197654867172242,
      "epoch": 5.311675183916437,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002567203825804966,
      "loss": 4.7223,
      "mean_token_accuracy": 0.24946884512901307,
      "num_tokens": 105574846.0,
      "step": 47295
    },
    {
      "entropy": 5.196389293670654,
      "epoch": 5.312236760824395,
      "grad_norm": 1.125,
      "learning_rate": 0.0002566803668705347,
      "loss": 4.6602,
      "mean_token_accuracy": 0.2514175161719322,
      "num_tokens": 105585848.0,
      "step": 47300
    },
    {
      "entropy": 5.230868768692017,
      "epoch": 5.312798337732352,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002566403517438772,
      "loss": 4.7378,
      "mean_token_accuracy": 0.24695564061403275,
      "num_tokens": 105596757.0,
      "step": 47305
    },
    {
      "entropy": 5.226493406295776,
      "epoch": 5.31335991464031,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025660033720179795,
      "loss": 4.8492,
      "mean_token_accuracy": 0.23688118308782577,
      "num_tokens": 105609363.0,
      "step": 47310
    },
    {
      "entropy": 5.224488544464111,
      "epoch": 5.3139214915482675,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025656032324557107,
      "loss": 4.6979,
      "mean_token_accuracy": 0.24370907843112946,
      "num_tokens": 105620909.0,
      "step": 47315
    },
    {
      "entropy": 5.160915994644165,
      "epoch": 5.314483068456225,
      "grad_norm": 1.0,
      "learning_rate": 0.00025652030987647066,
      "loss": 4.5698,
      "mean_token_accuracy": 0.2511246994137764,
      "num_tokens": 105631134.0,
      "step": 47320
    },
    {
      "entropy": 5.119993734359741,
      "epoch": 5.315044645364183,
      "grad_norm": 1.0078125,
      "learning_rate": 0.0002564802970957708,
      "loss": 4.6088,
      "mean_token_accuracy": 0.25127764940261843,
      "num_tokens": 105641979.0,
      "step": 47325
    },
    {
      "entropy": 5.172795820236206,
      "epoch": 5.31560622227214,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025644028490474544,
      "loss": 4.6648,
      "mean_token_accuracy": 0.243656162917614,
      "num_tokens": 105652320.0,
      "step": 47330
    },
    {
      "entropy": 5.193128538131714,
      "epoch": 5.316167799180098,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025640027330466866,
      "loss": 4.7047,
      "mean_token_accuracy": 0.23899315148591996,
      "num_tokens": 105663278.0,
      "step": 47335
    },
    {
      "entropy": 5.236804437637329,
      "epoch": 5.316729376088055,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025636026229681433,
      "loss": 4.7528,
      "mean_token_accuracy": 0.24506211578845977,
      "num_tokens": 105673930.0,
      "step": 47340
    },
    {
      "entropy": 5.181037187576294,
      "epoch": 5.317290952996013,
      "grad_norm": 1.171875,
      "learning_rate": 0.0002563202518824564,
      "loss": 4.645,
      "mean_token_accuracy": 0.24768121540546417,
      "num_tokens": 105684617.0,
      "step": 47345
    },
    {
      "entropy": 5.17449688911438,
      "epoch": 5.317852529903971,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000256280242062869,
      "loss": 4.6866,
      "mean_token_accuracy": 0.2446838542819023,
      "num_tokens": 105695278.0,
      "step": 47350
    },
    {
      "entropy": 5.231711721420288,
      "epoch": 5.318414106811928,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002562402328393259,
      "loss": 4.6875,
      "mean_token_accuracy": 0.23930345922708512,
      "num_tokens": 105706026.0,
      "step": 47355
    },
    {
      "entropy": 5.187387371063233,
      "epoch": 5.318975683719885,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025620022421310115,
      "loss": 4.6579,
      "mean_token_accuracy": 0.2463258683681488,
      "num_tokens": 105717141.0,
      "step": 47360
    },
    {
      "entropy": 5.1924882411956785,
      "epoch": 5.319537260627843,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025616021618546847,
      "loss": 4.6962,
      "mean_token_accuracy": 0.24487958401441573,
      "num_tokens": 105729240.0,
      "step": 47365
    },
    {
      "entropy": 5.212883853912354,
      "epoch": 5.3200988375358005,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025612020875770186,
      "loss": 4.6797,
      "mean_token_accuracy": 0.2429031565785408,
      "num_tokens": 105741344.0,
      "step": 47370
    },
    {
      "entropy": 5.2384144306182865,
      "epoch": 5.320660414443758,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002560802019310751,
      "loss": 4.705,
      "mean_token_accuracy": 0.24909189641475676,
      "num_tokens": 105752750.0,
      "step": 47375
    },
    {
      "entropy": 5.228447914123535,
      "epoch": 5.321221991351716,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025604019570686203,
      "loss": 4.7587,
      "mean_token_accuracy": 0.2436950385570526,
      "num_tokens": 105764677.0,
      "step": 47380
    },
    {
      "entropy": 5.257855606079102,
      "epoch": 5.321783568259673,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002560001900863365,
      "loss": 4.7022,
      "mean_token_accuracy": 0.23550109714269638,
      "num_tokens": 105775649.0,
      "step": 47385
    },
    {
      "entropy": 5.2457873821258545,
      "epoch": 5.322345145167631,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002559601850707723,
      "loss": 4.7086,
      "mean_token_accuracy": 0.24829225838184357,
      "num_tokens": 105785605.0,
      "step": 47390
    },
    {
      "entropy": 5.23152642250061,
      "epoch": 5.322906722075588,
      "grad_norm": 1.1953125,
      "learning_rate": 0.0002559201806614432,
      "loss": 4.7244,
      "mean_token_accuracy": 0.24037842750549315,
      "num_tokens": 105795887.0,
      "step": 47395
    },
    {
      "entropy": 5.179065370559693,
      "epoch": 5.323468298983546,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025588017685962297,
      "loss": 4.6149,
      "mean_token_accuracy": 0.2507654964923859,
      "num_tokens": 105806129.0,
      "step": 47400
    },
    {
      "entropy": 5.279801321029663,
      "epoch": 5.324029875891504,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025584017366658525,
      "loss": 4.744,
      "mean_token_accuracy": 0.24741773456335067,
      "num_tokens": 105817048.0,
      "step": 47405
    },
    {
      "entropy": 5.268767261505127,
      "epoch": 5.324591452799461,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025580017108360386,
      "loss": 4.7807,
      "mean_token_accuracy": 0.24443348348140717,
      "num_tokens": 105829847.0,
      "step": 47410
    },
    {
      "entropy": 5.214545011520386,
      "epoch": 5.325153029707418,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002557601691119524,
      "loss": 4.6939,
      "mean_token_accuracy": 0.24858913272619249,
      "num_tokens": 105839797.0,
      "step": 47415
    },
    {
      "entropy": 5.220709276199341,
      "epoch": 5.325714606615376,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00025572016775290475,
      "loss": 4.7283,
      "mean_token_accuracy": 0.23960171788930892,
      "num_tokens": 105851004.0,
      "step": 47420
    },
    {
      "entropy": 5.2611254215240475,
      "epoch": 5.326276183523333,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025568016700773435,
      "loss": 4.7625,
      "mean_token_accuracy": 0.23534238189458848,
      "num_tokens": 105861558.0,
      "step": 47425
    },
    {
      "entropy": 5.197557592391968,
      "epoch": 5.326837760431291,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025564016687771495,
      "loss": 4.626,
      "mean_token_accuracy": 0.24942552894353867,
      "num_tokens": 105872705.0,
      "step": 47430
    },
    {
      "entropy": 5.212981176376343,
      "epoch": 5.327399337339249,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025560016736412015,
      "loss": 4.7473,
      "mean_token_accuracy": 0.23319239318370819,
      "num_tokens": 105884426.0,
      "step": 47435
    },
    {
      "entropy": 5.189580631256104,
      "epoch": 5.327960914247206,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025556016846822353,
      "loss": 4.7118,
      "mean_token_accuracy": 0.24282812923192978,
      "num_tokens": 105896141.0,
      "step": 47440
    },
    {
      "entropy": 5.182676172256469,
      "epoch": 5.328522491155164,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025552017019129874,
      "loss": 4.6338,
      "mean_token_accuracy": 0.25339488983154296,
      "num_tokens": 105907216.0,
      "step": 47445
    },
    {
      "entropy": 5.298295688629151,
      "epoch": 5.329084068063121,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025548017253461923,
      "loss": 4.7736,
      "mean_token_accuracy": 0.24798806607723237,
      "num_tokens": 105918011.0,
      "step": 47450
    },
    {
      "entropy": 5.336629581451416,
      "epoch": 5.3296456449710785,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002554401754994586,
      "loss": 4.7973,
      "mean_token_accuracy": 0.23665598183870315,
      "num_tokens": 105929173.0,
      "step": 47455
    },
    {
      "entropy": 5.258586740493774,
      "epoch": 5.330207221879037,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002554001790870904,
      "loss": 4.7755,
      "mean_token_accuracy": 0.24236849397420884,
      "num_tokens": 105940825.0,
      "step": 47460
    },
    {
      "entropy": 5.192132091522216,
      "epoch": 5.330768798786994,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025536018329878814,
      "loss": 4.6747,
      "mean_token_accuracy": 0.24511311799287797,
      "num_tokens": 105951335.0,
      "step": 47465
    },
    {
      "entropy": 5.218711423873901,
      "epoch": 5.331330375694951,
      "grad_norm": 0.98046875,
      "learning_rate": 0.0002553201881358252,
      "loss": 4.7634,
      "mean_token_accuracy": 0.23261232823133468,
      "num_tokens": 105963529.0,
      "step": 47470
    },
    {
      "entropy": 5.304455852508545,
      "epoch": 5.331891952602909,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025528019359947514,
      "loss": 4.8067,
      "mean_token_accuracy": 0.23470841646194457,
      "num_tokens": 105975386.0,
      "step": 47475
    },
    {
      "entropy": 5.184107112884521,
      "epoch": 5.332453529510866,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002552401996910113,
      "loss": 4.6483,
      "mean_token_accuracy": 0.24499585181474687,
      "num_tokens": 105985710.0,
      "step": 47480
    },
    {
      "entropy": 5.143653345108032,
      "epoch": 5.333015106418824,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002552002064117072,
      "loss": 4.6143,
      "mean_token_accuracy": 0.25234147161245346,
      "num_tokens": 105997216.0,
      "step": 47485
    },
    {
      "entropy": 5.245766115188599,
      "epoch": 5.333576683326782,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002551602137628361,
      "loss": 4.6755,
      "mean_token_accuracy": 0.2487018123269081,
      "num_tokens": 106008111.0,
      "step": 47490
    },
    {
      "entropy": 5.214200592041015,
      "epoch": 5.334138260234739,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025512022174567164,
      "loss": 4.6758,
      "mean_token_accuracy": 0.2515032231807709,
      "num_tokens": 106018732.0,
      "step": 47495
    },
    {
      "entropy": 5.164587926864624,
      "epoch": 5.334699837142697,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025508023036148697,
      "loss": 4.6367,
      "mean_token_accuracy": 0.24499060064554215,
      "num_tokens": 106029740.0,
      "step": 47500
    },
    {
      "entropy": 5.17308611869812,
      "epoch": 5.335261414050654,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025504023961155553,
      "loss": 4.6357,
      "mean_token_accuracy": 0.25273127406835555,
      "num_tokens": 106040553.0,
      "step": 47505
    },
    {
      "entropy": 5.210087013244629,
      "epoch": 5.3358229909586115,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025500024949715055,
      "loss": 4.7352,
      "mean_token_accuracy": 0.24016026109457017,
      "num_tokens": 106051667.0,
      "step": 47510
    },
    {
      "entropy": 5.2711138248443605,
      "epoch": 5.33638456786657,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025496026001954543,
      "loss": 4.7923,
      "mean_token_accuracy": 0.24118512719869614,
      "num_tokens": 106062492.0,
      "step": 47515
    },
    {
      "entropy": 5.279668283462525,
      "epoch": 5.336946144774527,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025492027118001345,
      "loss": 4.7462,
      "mean_token_accuracy": 0.23959942162036896,
      "num_tokens": 106072645.0,
      "step": 47520
    },
    {
      "entropy": 5.152760219573975,
      "epoch": 5.337507721682485,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002548802829798277,
      "loss": 4.6435,
      "mean_token_accuracy": 0.25090509504079817,
      "num_tokens": 106084115.0,
      "step": 47525
    },
    {
      "entropy": 5.172522974014282,
      "epoch": 5.338069298590442,
      "grad_norm": 1.125,
      "learning_rate": 0.0002548402954202616,
      "loss": 4.7241,
      "mean_token_accuracy": 0.24157431423664094,
      "num_tokens": 106094093.0,
      "step": 47530
    },
    {
      "entropy": 5.224715995788574,
      "epoch": 5.338630875498399,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002548003085025883,
      "loss": 4.7293,
      "mean_token_accuracy": 0.24907050430774688,
      "num_tokens": 106104452.0,
      "step": 47535
    },
    {
      "entropy": 5.22742772102356,
      "epoch": 5.339192452406357,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025476032222808106,
      "loss": 4.7014,
      "mean_token_accuracy": 0.24263959527015685,
      "num_tokens": 106114540.0,
      "step": 47540
    },
    {
      "entropy": 5.221941041946411,
      "epoch": 5.339754029314315,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025472033659801297,
      "loss": 4.7373,
      "mean_token_accuracy": 0.23515698611736296,
      "num_tokens": 106124981.0,
      "step": 47545
    },
    {
      "entropy": 5.208458709716797,
      "epoch": 5.340315606222272,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002546803516136572,
      "loss": 4.6975,
      "mean_token_accuracy": 0.2372792184352875,
      "num_tokens": 106135037.0,
      "step": 47550
    },
    {
      "entropy": 5.280726575851441,
      "epoch": 5.34087718313023,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025464036727628695,
      "loss": 4.7761,
      "mean_token_accuracy": 0.24078935384750366,
      "num_tokens": 106145515.0,
      "step": 47555
    },
    {
      "entropy": 5.264743185043335,
      "epoch": 5.341438760038187,
      "grad_norm": 1.171875,
      "learning_rate": 0.00025460038358717527,
      "loss": 4.7151,
      "mean_token_accuracy": 0.240646094083786,
      "num_tokens": 106157246.0,
      "step": 47560
    },
    {
      "entropy": 5.256511688232422,
      "epoch": 5.342000336946144,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002545604005475953,
      "loss": 4.7399,
      "mean_token_accuracy": 0.2421582892537117,
      "num_tokens": 106167889.0,
      "step": 47565
    },
    {
      "entropy": 5.177395296096802,
      "epoch": 5.3425619138541025,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002545204181588201,
      "loss": 4.6892,
      "mean_token_accuracy": 0.24530988335609435,
      "num_tokens": 106178330.0,
      "step": 47570
    },
    {
      "entropy": 5.189733934402466,
      "epoch": 5.34312349076206,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025448043642212274,
      "loss": 4.6806,
      "mean_token_accuracy": 0.2419336512684822,
      "num_tokens": 106188995.0,
      "step": 47575
    },
    {
      "entropy": 5.196035766601563,
      "epoch": 5.343685067670018,
      "grad_norm": 1.2421875,
      "learning_rate": 0.0002544404553387763,
      "loss": 4.7001,
      "mean_token_accuracy": 0.23890968710184096,
      "num_tokens": 106199271.0,
      "step": 47580
    },
    {
      "entropy": 5.184180498123169,
      "epoch": 5.344246644577975,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025440047491005367,
      "loss": 4.7182,
      "mean_token_accuracy": 0.24283102452754973,
      "num_tokens": 106211495.0,
      "step": 47585
    },
    {
      "entropy": 5.206792593002319,
      "epoch": 5.344808221485932,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025436049513722795,
      "loss": 4.6572,
      "mean_token_accuracy": 0.24540145546197892,
      "num_tokens": 106223437.0,
      "step": 47590
    },
    {
      "entropy": 5.253609752655029,
      "epoch": 5.34536979839389,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00025432051602157203,
      "loss": 4.7211,
      "mean_token_accuracy": 0.23461210131645202,
      "num_tokens": 106233456.0,
      "step": 47595
    },
    {
      "entropy": 5.187824630737305,
      "epoch": 5.345931375301848,
      "grad_norm": 1.03125,
      "learning_rate": 0.0002542805375643589,
      "loss": 4.6221,
      "mean_token_accuracy": 0.24453176259994508,
      "num_tokens": 106244559.0,
      "step": 47600
    },
    {
      "entropy": 5.154763984680176,
      "epoch": 5.346492952209805,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002542405597668615,
      "loss": 4.6314,
      "mean_token_accuracy": 0.24645740687847137,
      "num_tokens": 106254427.0,
      "step": 47605
    },
    {
      "entropy": 5.204897165298462,
      "epoch": 5.347054529117763,
      "grad_norm": 1.03125,
      "learning_rate": 0.00025420058263035276,
      "loss": 4.6823,
      "mean_token_accuracy": 0.24702100604772567,
      "num_tokens": 106264616.0,
      "step": 47610
    },
    {
      "entropy": 5.219987297058106,
      "epoch": 5.34761610602572,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002541606061561055,
      "loss": 4.6513,
      "mean_token_accuracy": 0.24938438683748246,
      "num_tokens": 106275103.0,
      "step": 47615
    },
    {
      "entropy": 5.281322574615478,
      "epoch": 5.348177682933677,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002541206303453927,
      "loss": 4.7904,
      "mean_token_accuracy": 0.2420526072382927,
      "num_tokens": 106286004.0,
      "step": 47620
    },
    {
      "entropy": 5.2512462615966795,
      "epoch": 5.3487392598416355,
      "grad_norm": 1.1171875,
      "learning_rate": 0.000254080655199487,
      "loss": 4.7367,
      "mean_token_accuracy": 0.24278880953788756,
      "num_tokens": 106297161.0,
      "step": 47625
    },
    {
      "entropy": 5.25691237449646,
      "epoch": 5.349300836749593,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00025404068071966133,
      "loss": 4.8058,
      "mean_token_accuracy": 0.23678888082504274,
      "num_tokens": 106309330.0,
      "step": 47630
    },
    {
      "entropy": 5.253285074234009,
      "epoch": 5.349862413657551,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025400070690718865,
      "loss": 4.7204,
      "mean_token_accuracy": 0.24053678661584854,
      "num_tokens": 106319801.0,
      "step": 47635
    },
    {
      "entropy": 5.224301195144653,
      "epoch": 5.350423990565508,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025396073376334155,
      "loss": 4.7616,
      "mean_token_accuracy": 0.23499933630228043,
      "num_tokens": 106331328.0,
      "step": 47640
    },
    {
      "entropy": 5.228511381149292,
      "epoch": 5.350985567473465,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025392076128939283,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24810541570186614,
      "num_tokens": 106343740.0,
      "step": 47645
    },
    {
      "entropy": 5.193121433258057,
      "epoch": 5.351547144381423,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0002538807894866153,
      "loss": 4.6521,
      "mean_token_accuracy": 0.25384672582149503,
      "num_tokens": 106354620.0,
      "step": 47650
    },
    {
      "entropy": 5.179651403427124,
      "epoch": 5.3521087212893805,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025384081835628163,
      "loss": 4.6383,
      "mean_token_accuracy": 0.24564062654972077,
      "num_tokens": 106366247.0,
      "step": 47655
    },
    {
      "entropy": 5.211027383804321,
      "epoch": 5.352670298197338,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002538008478996645,
      "loss": 4.7308,
      "mean_token_accuracy": 0.24635204076766967,
      "num_tokens": 106377045.0,
      "step": 47660
    },
    {
      "entropy": 5.16166729927063,
      "epoch": 5.353231875105296,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002537608781180366,
      "loss": 4.6693,
      "mean_token_accuracy": 0.2456437572836876,
      "num_tokens": 106389070.0,
      "step": 47665
    },
    {
      "entropy": 5.187287950515747,
      "epoch": 5.353793452013253,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002537209090126704,
      "loss": 4.6843,
      "mean_token_accuracy": 0.24161513149738312,
      "num_tokens": 106401292.0,
      "step": 47670
    },
    {
      "entropy": 5.2541405200958256,
      "epoch": 5.35435502892121,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025368094058483885,
      "loss": 4.6745,
      "mean_token_accuracy": 0.24745139628648757,
      "num_tokens": 106413045.0,
      "step": 47675
    },
    {
      "entropy": 5.1923199653625485,
      "epoch": 5.354916605829168,
      "grad_norm": 1.0625,
      "learning_rate": 0.0002536409728358145,
      "loss": 4.6509,
      "mean_token_accuracy": 0.2493009701371193,
      "num_tokens": 106424655.0,
      "step": 47680
    },
    {
      "entropy": 5.23819260597229,
      "epoch": 5.355478182737126,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002536010057668698,
      "loss": 4.7659,
      "mean_token_accuracy": 0.2293037950992584,
      "num_tokens": 106436544.0,
      "step": 47685
    },
    {
      "entropy": 5.169367265701294,
      "epoch": 5.356039759645084,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002535610393792773,
      "loss": 4.6983,
      "mean_token_accuracy": 0.24257628619670868,
      "num_tokens": 106446936.0,
      "step": 47690
    },
    {
      "entropy": 5.236230373382568,
      "epoch": 5.356601336553041,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00025352107367430975,
      "loss": 4.7187,
      "mean_token_accuracy": 0.2439270332455635,
      "num_tokens": 106457328.0,
      "step": 47695
    },
    {
      "entropy": 5.212345266342163,
      "epoch": 5.357162913460998,
      "grad_norm": 1.1328125,
      "learning_rate": 0.00025348110865323946,
      "loss": 4.7423,
      "mean_token_accuracy": 0.24503328800201415,
      "num_tokens": 106468037.0,
      "step": 47700
    },
    {
      "entropy": 5.24719090461731,
      "epoch": 5.357724490368956,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002534411443173389,
      "loss": 4.7233,
      "mean_token_accuracy": 0.2462645336985588,
      "num_tokens": 106479406.0,
      "step": 47705
    },
    {
      "entropy": 5.107427310943604,
      "epoch": 5.3582860672769135,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002534011806678809,
      "loss": 4.5944,
      "mean_token_accuracy": 0.24120155721902847,
      "num_tokens": 106491142.0,
      "step": 47710
    },
    {
      "entropy": 5.177434682846069,
      "epoch": 5.358847644184872,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002533612177061376,
      "loss": 4.7258,
      "mean_token_accuracy": 0.24725850224494933,
      "num_tokens": 106502743.0,
      "step": 47715
    },
    {
      "entropy": 5.229650783538818,
      "epoch": 5.359409221092829,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00025332125543338146,
      "loss": 4.7334,
      "mean_token_accuracy": 0.24470157623291017,
      "num_tokens": 106513765.0,
      "step": 47720
    },
    {
      "entropy": 5.158356666564941,
      "epoch": 5.359970798000786,
      "grad_norm": 1.171875,
      "learning_rate": 0.000253281293850885,
      "loss": 4.5824,
      "mean_token_accuracy": 0.25398160964250566,
      "num_tokens": 106524154.0,
      "step": 47725
    },
    {
      "entropy": 5.180279922485352,
      "epoch": 5.360532374908744,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00025324133295992056,
      "loss": 4.7016,
      "mean_token_accuracy": 0.24183404743671416,
      "num_tokens": 106535946.0,
      "step": 47730
    },
    {
      "entropy": 5.287762451171875,
      "epoch": 5.361093951816701,
      "grad_norm": 1.15625,
      "learning_rate": 0.0002532013727617605,
      "loss": 4.8159,
      "mean_token_accuracy": 0.24044443517923356,
      "num_tokens": 106546352.0,
      "step": 47735
    },
    {
      "entropy": 5.229134607315063,
      "epoch": 5.361655528724659,
      "grad_norm": 1.125,
      "learning_rate": 0.0002531614132576772,
      "loss": 4.6255,
      "mean_token_accuracy": 0.24989299923181535,
      "num_tokens": 106557640.0,
      "step": 47740
    },
    {
      "entropy": 5.24246621131897,
      "epoch": 5.362217105632617,
      "grad_norm": 1.1875,
      "learning_rate": 0.000253121454448943,
      "loss": 4.7138,
      "mean_token_accuracy": 0.23983247876167296,
      "num_tokens": 106567762.0,
      "step": 47745
    },
    {
      "entropy": 5.210256958007813,
      "epoch": 5.362778682540574,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025308149633683014,
      "loss": 4.697,
      "mean_token_accuracy": 0.25403257757425307,
      "num_tokens": 106579589.0,
      "step": 47750
    },
    {
      "entropy": 5.197050762176514,
      "epoch": 5.363340259448531,
      "grad_norm": 1.09375,
      "learning_rate": 0.000253041538922611,
      "loss": 4.6683,
      "mean_token_accuracy": 0.2419513925909996,
      "num_tokens": 106590122.0,
      "step": 47755
    },
    {
      "entropy": 5.2822164535522464,
      "epoch": 5.363901836356489,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025300158220755776,
      "loss": 4.7828,
      "mean_token_accuracy": 0.24043029695749282,
      "num_tokens": 106601269.0,
      "step": 47760
    },
    {
      "entropy": 5.32268648147583,
      "epoch": 5.364463413264446,
      "grad_norm": 1.0859375,
      "learning_rate": 0.00025296162619294267,
      "loss": 4.8787,
      "mean_token_accuracy": 0.23098440766334533,
      "num_tokens": 106612462.0,
      "step": 47765
    },
    {
      "entropy": 5.248672914505005,
      "epoch": 5.3650249901724045,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002529216708800379,
      "loss": 4.7054,
      "mean_token_accuracy": 0.24711115062236785,
      "num_tokens": 106623333.0,
      "step": 47770
    },
    {
      "entropy": 5.215450191497803,
      "epoch": 5.365586567080362,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025288171627011564,
      "loss": 4.6576,
      "mean_token_accuracy": 0.2529521718621254,
      "num_tokens": 106634505.0,
      "step": 47775
    },
    {
      "entropy": 5.267479181289673,
      "epoch": 5.366148143988319,
      "grad_norm": 1.1484375,
      "learning_rate": 0.00025284176236444817,
      "loss": 4.7882,
      "mean_token_accuracy": 0.2382896885275841,
      "num_tokens": 106645764.0,
      "step": 47780
    },
    {
      "entropy": 5.148192453384399,
      "epoch": 5.366709720896277,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025280180916430764,
      "loss": 4.6587,
      "mean_token_accuracy": 0.24721285253763198,
      "num_tokens": 106657484.0,
      "step": 47785
    },
    {
      "entropy": 5.169534921646118,
      "epoch": 5.367271297804234,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025276185667096605,
      "loss": 4.6768,
      "mean_token_accuracy": 0.24393104761838913,
      "num_tokens": 106669038.0,
      "step": 47790
    },
    {
      "entropy": 5.276986503601075,
      "epoch": 5.3678328747121915,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025272190488569564,
      "loss": 4.6881,
      "mean_token_accuracy": 0.24738954305648803,
      "num_tokens": 106679447.0,
      "step": 47795
    },
    {
      "entropy": 5.217820310592652,
      "epoch": 5.36839445162015,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025268195380976835,
      "loss": 4.7253,
      "mean_token_accuracy": 0.23748468458652497,
      "num_tokens": 106689596.0,
      "step": 47800
    },
    {
      "entropy": 5.222412014007569,
      "epoch": 5.368956028528107,
      "grad_norm": 1.1015625,
      "learning_rate": 0.0002526420034444564,
      "loss": 4.7032,
      "mean_token_accuracy": 0.25481328964233396,
      "num_tokens": 106701825.0,
      "step": 47805
    },
    {
      "entropy": 5.339792394638062,
      "epoch": 5.369517605436064,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025260205379103157,
      "loss": 4.8811,
      "mean_token_accuracy": 0.23091376423835755,
      "num_tokens": 106713451.0,
      "step": 47810
    },
    {
      "entropy": 5.296330165863037,
      "epoch": 5.370079182344022,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025256210485076616,
      "loss": 4.8199,
      "mean_token_accuracy": 0.23914206475019456,
      "num_tokens": 106724732.0,
      "step": 47815
    },
    {
      "entropy": 5.262114953994751,
      "epoch": 5.370640759251979,
      "grad_norm": 1.078125,
      "learning_rate": 0.000252522156624932,
      "loss": 4.7472,
      "mean_token_accuracy": 0.24680538177490235,
      "num_tokens": 106736533.0,
      "step": 47820
    },
    {
      "entropy": 5.1662249088287355,
      "epoch": 5.3712023361599375,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00025248220911480105,
      "loss": 4.6322,
      "mean_token_accuracy": 0.2526562035083771,
      "num_tokens": 106747454.0,
      "step": 47825
    },
    {
      "entropy": 5.1196861743927,
      "epoch": 5.371763913067895,
      "grad_norm": 1.0703125,
      "learning_rate": 0.00025244226232164533,
      "loss": 4.654,
      "mean_token_accuracy": 0.2478431537747383,
      "num_tokens": 106757833.0,
      "step": 47830
    },
    {
      "entropy": 5.11096568107605,
      "epoch": 5.372325489975852,
      "grad_norm": 1.0234375,
      "learning_rate": 0.00025240231624673674,
      "loss": 4.6577,
      "mean_token_accuracy": 0.24357233047485352,
      "num_tokens": 106768646.0,
      "step": 47835
    },
    {
      "entropy": 5.204378461837768,
      "epoch": 5.37288706688381,
      "grad_norm": 1.0390625,
      "learning_rate": 0.0002523623708913472,
      "loss": 4.7106,
      "mean_token_accuracy": 0.24395862966775894,
      "num_tokens": 106780434.0,
      "step": 47840
    },
    {
      "entropy": 5.227003383636474,
      "epoch": 5.373448643791767,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002523224262567485,
      "loss": 4.7248,
      "mean_token_accuracy": 0.2529944196343422,
      "num_tokens": 106792372.0,
      "step": 47845
    },
    {
      "entropy": 5.2419744491577145,
      "epoch": 5.3740102206997244,
      "grad_norm": 1.1796875,
      "learning_rate": 0.00025228248234421247,
      "loss": 4.6929,
      "mean_token_accuracy": 0.23565352261066436,
      "num_tokens": 106802251.0,
      "step": 47850
    },
    {
      "entropy": 5.208640384674072,
      "epoch": 5.3745717976076826,
      "grad_norm": 1.1640625,
      "learning_rate": 0.0002522425391550111,
      "loss": 4.7555,
      "mean_token_accuracy": 0.24462482184171677,
      "num_tokens": 106813818.0,
      "step": 47855
    },
    {
      "entropy": 5.2676397323608395,
      "epoch": 5.37513337451564,
      "grad_norm": 1.109375,
      "learning_rate": 0.00025220259669041613,
      "loss": 4.7912,
      "mean_token_accuracy": 0.23384694159030914,
      "num_tokens": 106824096.0,
      "step": 47860
    },
    {
      "entropy": 5.317703485488892,
      "epoch": 5.375694951423597,
      "grad_norm": 1.09375,
      "learning_rate": 0.0002521626549516993,
      "loss": 4.8388,
      "mean_token_accuracy": 0.23523468822240828,
      "num_tokens": 106836037.0,
      "step": 47865
    },
    {
      "entropy": 5.305294704437256,
      "epoch": 5.376256528331555,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002521227139401324,
      "loss": 4.8301,
      "mean_token_accuracy": 0.23898794800043105,
      "num_tokens": 106846370.0,
      "step": 47870
    },
    {
      "entropy": 5.2790247917175295,
      "epoch": 5.376818105239512,
      "grad_norm": 0.921875,
      "learning_rate": 0.00025208277365698715,
      "loss": 4.8176,
      "mean_token_accuracy": 0.2386333763599396,
      "num_tokens": 106857945.0,
      "step": 47875
    },
    {
      "entropy": 5.22119288444519,
      "epoch": 5.37737968214747,
      "grad_norm": 1.1171875,
      "learning_rate": 0.0002520428341035353,
      "loss": 4.663,
      "mean_token_accuracy": 0.24797790497541428,
      "num_tokens": 106867899.0,
      "step": 47880
    },
    {
      "entropy": 5.207743787765503,
      "epoch": 5.377941259055428,
      "grad_norm": 0.9921875,
      "learning_rate": 0.0002520028952810485,
      "loss": 4.7162,
      "mean_token_accuracy": 0.23735710084438325,
      "num_tokens": 106879305.0,
      "step": 47885
    },
    {
      "entropy": 5.295254707336426,
      "epoch": 5.378502835963385,
      "grad_norm": 1.0078125,
      "learning_rate": 0.00025196295719079835,
      "loss": 4.7954,
      "mean_token_accuracy": 0.2400704726576805,
      "num_tokens": 106891944.0,
      "step": 47890
    },
    {
      "entropy": 5.230411529541016,
      "epoch": 5.379064412871343,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002519230198340567,
      "loss": 4.7182,
      "mean_token_accuracy": 0.2460768073797226,
      "num_tokens": 106903420.0,
      "step": 47895
    },
    {
      "entropy": 5.198486232757569,
      "epoch": 5.3796259897793,
      "grad_norm": 1.078125,
      "learning_rate": 0.00025188308321209504,
      "loss": 4.7071,
      "mean_token_accuracy": 0.24388057589530945,
      "num_tokens": 106914838.0,
      "step": 47900
    },
    {
      "entropy": 5.0995556831359865,
      "epoch": 5.380187566687258,
      "grad_norm": 1.0625,
      "learning_rate": 0.00025184314732618496,
      "loss": 4.6317,
      "mean_token_accuracy": 0.2503864258527756,
      "num_tokens": 106926519.0,
      "step": 47905
    },
    {
      "entropy": 5.192728710174561,
      "epoch": 5.3807491435952155,
      "grad_norm": 1.0703125,
      "learning_rate": 0.000251803212177598,
      "loss": 4.7421,
      "mean_token_accuracy": 0.24822444915771485,
      "num_tokens": 106938303.0,
      "step": 47910
    },
    {
      "entropy": 5.221099424362182,
      "epoch": 5.381310720503173,
      "grad_norm": 0.98828125,
      "learning_rate": 0.0002517632777676058,
      "loss": 4.7511,
      "mean_token_accuracy": 0.2459801748394966,
      "num_tokens": 106950120.0,
      "step": 47915
    },
    {
      "entropy": 5.26410608291626,
      "epoch": 5.381872297411131,
      "grad_norm": 1.046875,
      "learning_rate": 0.00025172334409747976,
      "loss": 4.7582,
      "mean_token_accuracy": 0.23686647713184356,
      "num_tokens": 106960998.0,
      "step": 47920
    },
    {
      "entropy": 5.214431715011597,
      "epoch": 5.382433874319088,
      "grad_norm": 1.046875,
      "learning_rate": 0.0002516834111684915,
      "loss": 4.705,
      "mean_token_accuracy": 0.24478403776884078,
      "num_tokens": 106972640.0,
      "step": 47925
    },
    {
      "entropy": 5.19734787940979,
      "epoch": 5.382995451227045,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002516434789819126,
      "loss": 4.7148,
      "mean_token_accuracy": 0.24395986497402192,
      "num_tokens": 106985300.0,
      "step": 47930
    },
    {
      "entropy": 5.262094449996948,
      "epoch": 5.383557028135003,
      "grad_norm": 1.109375,
      "learning_rate": 0.0002516035475390143,
      "loss": 4.7538,
      "mean_token_accuracy": 0.23009011298418044,
      "num_tokens": 106996074.0,
      "step": 47935
    },
    {
      "entropy": 5.278050279617309,
      "epoch": 5.384118605042961,
      "grad_norm": 1.0546875,
      "learning_rate": 0.00025156361684106814,
      "loss": 4.7394,
      "mean_token_accuracy": 0.24572164714336395,
      "num_tokens": 107006493.0,
      "step": 47940
    },
    {
      "entropy": 5.199740743637085,
      "epoch": 5.384680181950918,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002515236868893455,
      "loss": 4.6738,
      "mean_token_accuracy": 0.2383109748363495,
      "num_tokens": 107018739.0,
      "step": 47945
    },
    {
      "entropy": 5.133124208450317,
      "epoch": 5.385241758858876,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025148375768511776,
      "loss": 4.5777,
      "mean_token_accuracy": 0.2550569519400597,
      "num_tokens": 107028762.0,
      "step": 47950
    },
    {
      "entropy": 5.231272506713867,
      "epoch": 5.385803335766833,
      "grad_norm": 1.1484375,
      "learning_rate": 0.0002514438292296563,
      "loss": 4.7299,
      "mean_token_accuracy": 0.24014270156621934,
      "num_tokens": 107039609.0,
      "step": 47955
    },
    {
      "entropy": 5.1270795345306395,
      "epoch": 5.386364912674791,
      "grad_norm": 1.09375,
      "learning_rate": 0.00025140390152423247,
      "loss": 4.631,
      "mean_token_accuracy": 0.2515767216682434,
      "num_tokens": 107051871.0,
      "step": 47960
    },
    {
      "entropy": 5.190711832046508,
      "epoch": 5.3869264895827484,
      "grad_norm": 1.015625,
      "learning_rate": 0.0002513639745701176,
      "loss": 4.7196,
      "mean_token_accuracy": 0.2455105558037758,
      "num_tokens": 107063560.0,
      "step": 47965
    },
    {
      "entropy": 5.250558853149414,
      "epoch": 5.387488066490706,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025132404836858294,
      "loss": 4.7373,
      "mean_token_accuracy": 0.24392561614513397,
      "num_tokens": 107074466.0,
      "step": 47970
    },
    {
      "entropy": 5.23563323020935,
      "epoch": 5.388049643398664,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0002512841229208997,
      "loss": 4.7109,
      "mean_token_accuracy": 0.249300517141819,
      "num_tokens": 107085470.0,
      "step": 47975
    },
    {
      "entropy": 5.2430887699127195,
      "epoch": 5.388611220306621,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0002512441982283392,
      "loss": 4.7797,
      "mean_token_accuracy": 0.2445221945643425,
      "num_tokens": 107098179.0,
      "step": 47980
    },
    {
      "entropy": 5.186902570724487,
      "epoch": 5.389172797214578,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002512042742921727,
      "loss": 4.7207,
      "mean_token_accuracy": 0.24212294369935988,
      "num_tokens": 107108941.0,
      "step": 47985
    },
    {
      "entropy": 5.24734058380127,
      "epoch": 5.389734374122536,
      "grad_norm": 1.1328125,
      "learning_rate": 0.0002511643511136712,
      "loss": 4.782,
      "mean_token_accuracy": 0.24119867980480195,
      "num_tokens": 107120019.0,
      "step": 47990
    },
    {
      "entropy": 5.245054626464844,
      "epoch": 5.3902959510304935,
      "grad_norm": 1.1015625,
      "learning_rate": 0.00025112442869410607,
      "loss": 4.7619,
      "mean_token_accuracy": 0.23407867550849915,
      "num_tokens": 107130254.0,
      "step": 47995
    },
    {
      "entropy": 5.143104600906372,
      "epoch": 5.390857527938451,
      "grad_norm": 1.0390625,
      "learning_rate": 0.00025108450703474844,
      "loss": 4.6193,
      "mean_token_accuracy": 0.24926262348890305,
      "num_tokens": 107140625.0,
      "step": 48000
    },
    {
      "epoch": 5.390857527938451,
      "eval_entropy": 4.957840720617743,
      "eval_loss": 4.944662094116211,
      "eval_mean_token_accuracy": 0.23715294285290384,
      "eval_num_tokens": 107140625.0,
      "eval_runtime": 24.1271,
      "eval_samples_per_second": 1285.401,
      "eval_steps_per_second": 160.691,
      "step": 48000
    }
  ],
  "logging_steps": 5,
  "max_steps": 89030,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 3000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.67929657648128e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}